AquaBrowser Library FAQ ADA Compliant

AquaBrowser Library FAQ ADA Compliant ........................................................................... 3 Advanced searching .................................................................... 3 Apache ...................................................................................... 3 Authority Database ..................................................................... 4 Boolean Searches ....................................................................... 5 Breadcrumbs .............................................................................. 5 Browsing ................................................................................... 6 Catalog Enrichment Subscriptions (also see Syndetics)..................... 6 Customer Support....................................................................... 6 Control Panel.............................................................................. 6 Customization ............................................................................ 7 Dictionary and Thesauri ............................................................... 8 Digital Archives........................................................................... 8 Export lists................................................................................. 9 Facets ....................................................................................... 9 Federated Searching.................................................................. 11 Fiction Connection..................................................................... 13 FRBR (Functional Requirements for Bibliographic Records) .............. 13 Google Book Search .................................................................. 14 Hardware specs and browser compatibility.................................... 15 Harvested Websites – examples .................................................. 17 Hosting Service ........................................................................ 17 Icons....................................................................................... 18 ILS Holds and Reserves ............................................................. 18 ILS Integration ......................................................................... 18 ILS Questions ........................................................................... 20 Implementation Process & Tasks ................................................. 21 Installation Method ................................................................... 23 Item Availability........................................................................ 23 Indexing Routine ...................................................................... 23 Maintenance – Library staff skills................................................. 24 Maintenance & Support Services ................................................. 24 Mobile Computing Interface ........................................................ 28 Multilingual .............................................................................. 29 My Discoveries ......................................................................... 31 AquaBrowser Library FAQ Last Updated: 5/15/2008 OAI......................................................................................... 33 Other Data Sources ................................................................... 33 Non-MARC Data ..................................................................... 34 MARC Data............................................................................ 34 Overview of AquaBrowser – Search, Discover, Refine ..................... 35 PDF docs ................................................................................. 38 Persistent URL .......................................................................... 38 Performance............................................................................. 38 References (Academic Libraries) ................................................. 39 References (Public Libraries)....................................................... 39 Relevancy Ranking .................................................................... 40 Reports and Analytics ................................................................ 41 Results Display ......................................................................... 42 RSS Feeds ............................................................................... 43 Save Records ........................................................................... 44 Search Entry Persistence............................................................ 44 Searching ................................................................................ 44 Security................................................................................... 45 Shibboleth ............................................................................... 45 Skins....................................................................................... 46 Sorting .................................................................................... 46 Spelling Variant ........................................................................ 46 Staff Administration .................................................................. 46 Standards ................................................................................ 47 Stemming, Spell Checking and Word Completion ........................... 47 Syndetics ICE ........................................................................... 48 Training................................................................................... 53 Table of Contents in MARC record................................................ 53 Training – additional.................................................................. 53 Updates & Patches .................................................................... 54 W3C’s Web Content Accessibility Guidelines .................................. 54 WebCrawler ............................................................................. 54 Word Cloud .............................................................................. 55 Z39.50 .................................................................................... 57 Miscellaneous ........................................................................... 57 Page 2 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 ADA Compliant ADA compliant? yes, (actually, european libraries have more stringent guidelines for disabilities, so we go beyond what's required in US) Advanced searching Most AquaBrowser implementations operate with one simple se arch box and leverage faceted navigation to accomplish search precision for casual as well as expert users. A few recent implementations (King County Library System, Howard County Library System, and University of Chicago) now implement ―advanced search options or a pull-down menu of search types. Field searching is available in the optional Advanced Search screen. The library can identify what fields are shown in the optional Advanced Search. Apache Question from our IT department regarding whether AquaBrowser can run with Apache rather than IIS (as we understand that there is a version of Apache that runs on Windows). AquaBrowser: The AquaBrowser website is implemented on the asp.net platform in C#. This is only supported in IIS for production environments. If the IT department feels strongly about having an Apache web server facing the outside world, the library can consider using the setup that is used by Columbus Metropolitan Library (http://catalog.columbuslibrary.org). They run two AquaBrowser servers with Windows and IIS. In front of these servers is a Linux server with Apache that receives all requests and does the load balancing. They use the mod_proxy module for this. Page 3 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 Authority Database Does it utilize authority records from the Library catalog? If so,how? Authority records are fully utilized to show see-also references and see references in both the word cloud and in the search results. You may see an example of this by searching http://lens.lib.uchicago.edu/ for 'jaguar' or 'smell'. The search results are automatically expanded with sees. The alternatives receive lower relevancy ranking so that your original query is what you see first in your results. Authority files are usually indexed by using authority MARC exports, comparable to title records exports. Of course, as with other sources, AquaBrowser can also ingest any XML or structured text based format. If you choose to have your authority records loaded into AquaBrowser , see also references are shown as suggestions as well. Benefits of AquaBrowser - how will it help the library reach its goals of providing a keyword-driven and visually-oriented means of viewing materials available at SPLA libraries. AquaBrowser builds on the traditional searching functions of your III Millennium catalog to provide your customers with an intuitive, visual, keyword-driven discovery tool. It provides simple, fast, convenient, and secure access to information about SPLA local materials, as well as to any of your virtual databases and collections you want to include. Summary of AquaBrowser Library Benefits • One interface o Delivers all types and formats of content through a single interface o Searches all sources from a single query o Delivers the first visually-represented and faceted search solution for libraries o Increases the success rate of patrons’ searches to improve patron satisfaction with library services o Increases usage of all resources • Search o Relevant results help patrons find answers fast o FRBR results help unclutter the results (optional ICE Title Link) o Relevance can be customized by the library Page 4 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 • • • • o Advanced search tools (optional) o Intuitive and easy for patrons to use – no training required Discover o Visual word cloud encourage exploration and discovery Refine o Facets help patrons quickly focus the results Social networking features enhance the patron experience (optional annual My Discoveries subscription) o Patron experiences are enhanced using the My Discoveries “Social Library Experience.” o My Discoveries LibraryThing tag content; your patrons can take advantage of more than 21 million tags and 2 million ratings for almost 17 million books. (optional) Rapid implementation o Installs quickly and easily o Integrates with any ILS o Integrates with nearly any digital resource, subscripted electronic resource, physical resource o Requires no change in existing systems Boolean Searches Boolean searching for OR and NOT operators is available in both the basic and optional Advanced Search. In the basic search box, the user can enter an ―OR∥ in capitals : such as jaguar OR cheetah. The user can do a NOT search by using the minus sign: jaguar –cheetah. Is the Boolean search functional? When we use the “or” it doesn’t seem to work. Is it possible to have more information on how it works? The Help page at University of Chicago is incorrect. You can do an “or” with capital OR: jaguar OR cheetah. You can do a NOT search by using the minus sign: jaguar –cheetah Breadcrumbs Page 5 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 Browsing Browse in the sense of alphabetic paging forward and backward through an a-z list of headings is not supported. Browse in the sense of faceted navigation forward (narrower) and backward (broader) is supported. Catalog Enrichment Subscriptions (also see Syndetics) We integrate any current Catalog Enrichment subscription into the AquaBrowser Library result lists. (Please note catalog enrichment from sources other than Syndetics can appear in the full record native display but not in results lists.) Customer Support Customer support is offered through AquaBrowser North America. Support telephone hours are 8 a.m. -5 p.m. Monday through Friday (Eastern Time), Email may be sent 24/7 to our supportusa@aquabrowser.com. The email will automatically create a ticket in our internal ticket system and assign it to AquaBrowser North America Support. This team works closely with the technical experts and developers in the Netherlands for Tier 2 support as needed. Control Panel The Control Panel offers a way to check system status, when the last update to the indexes took place, and general system-health-related parameters. Common configuration options available within AquaBrowser are editable from a password-protected web-based configuration panel. Cosmetic details of the user interface may be customized in terms of “skins”, cascading style sheets, etc. Page 6 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 The AquaBrowser control panel is web based and is in fact part of the AquaBrowser Library® website itself. After logging in, an administrator can change configuration files that control system behaviour. Almost all configuration files are XML files. Familiarity with XML is preferred for some of the functions in the control panel. Please see a screen shot of the Control Panel Home, below. Customization Are we able to customize AquaBrowser ourselves? AquaBrowser provides both parameter driven customizations and those libraries with the skills for design and general web development are free to customize AquaBrowser as they desire, up to a certain level. Additional customization services at implementation and beyond can be purchased from MediaLab. Page 7 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 How much can you configure displays, and behavior (opening new windows)? AquaBrowser is highly configurable by the user. Please see the variants in deployments of our over 150 implementations at: http://www.AquaBrowser .com. The library can implement for example scripting to tune the calculations of values in certain facets. Real life examples of customization done by customers: - At Indianapolis (sherloc.imcpl.org) the library staff implemented the “make request” functionality themselves. They used a combination of javascript and css on the client side and perl scripting with ODBC request on the server side. - At University of Chicago the library staff implemented the full record display by themselves. This required advanced xslt and css. Dictionary and Thesauri Digital Archives Indianapolis Marion County Public Library’s AquaBrowser has the following sources indexed: - Library catalog (SirsiDynix Horizon) - Museum artifacts collection (ContentDM repository) - May Wright Sewall Papers (ContentDM repository) - The library website (focused web crawler The artifacts collection and May Wright Sewall Papers show thumbnails of digitized objects. The URL for this AquaBrowser is http://sherloc.imcpl.org. The refine facet “Found in” shows the sources it found info in. For example search on Beatles and refine on Museum Artifacts Collection. You see this: Page 8 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 In this case AquaBrowser display the brief record. The library chose to link out directly to the original site for the full record view. AquaBrowser can also display the full record view. Another example is the Rochester Museum & Science Center in Rochester, New York. Their AquaBrowser can be accessed via http://www.youseemore.com/rmsc/default.htm. You can search for example on Rochester photo. Note that AquaBrowser displays both the brief and full record view. Export lists AquaBrowser: AquaBrowser can email results or export in common formats such as RefWorks and EndNote. Facets When we use the “refine facets” a breadcrumb trail appears on the top of the page, depending on the catalogue we use we can cancel different “refine facets”. For example, at the University of Chicago (LENS) it is possible to remove a facet in the middle and at Trinity College (Stella) we can only remove the last facet. We would like to know if theses differences depend on the version of Aqua Brower or if it’s an implementing choice? Page 9 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 The University of Chicago has a newer version of AquaBrowser that allows for the remove/keep functionality. For any customer this is now an implementation choice. Can you describe to us the flexibility that we have in determining the parsing of our data, which fields are indexed in which value, determining the facets and the data on which they are built, etc.. ? We understand that most of this information is detailed in the ABL Workbook. Does it is possible to receive a copy of it as it will probably inform us a lot about those questions? There is full flexibility in these aspects. As examples, please refer to the attached documents that show two for a public Library and the facet chart for University of Chicago. This shows you that you can range from a pretty straightforward implementation to a very detailed, flexible implementation. Specifically, from what fields does AquaBrowser derive the facets used to refine searching? Facets, or refine dimensions, are defined by the Library during implementation and can be from any source field, generally the MARC record is used when choosing dimensions for the Library catalog. However, any tagged and programmatically identifiable component of data can be included in the refine dimensions, such as circulation availability. Can facets come from non-marc fields in marc records. Can facets come from non-marc fields in non-marc records. Are the facets pre-defined. Does the Next-Gen Catalog allow LC classification as a facet. Anything in the source data, be it marc or non-marc, controlled or not, can be used as a facet, in very flexible ways. You can also combine multiple data fields into one facet, or apply hierarchical treatment of data in fields. AquaBrowser supports LC classification as a facet -- again, see http://lens.lib.uchicago.edu/ for an example. Other functions of the faceting include custom sorting, field treatment, and smart 'stepping' of facets, as seen by refining on a publication period (e.g., last 10 years), after which specific years appear. AquaBrowser lets users 'sculpt' their results by removing applied facets or by locking them into place. This interaction model makes it easy and intuitive to apply 'prelimiters' by selecting them first, then locking them in place as a user performs multiple queries. May multiple MARC tags be used to identify one facet? Yes, tags, subfields and indicators are all identified by the Library to determine a facet. Below is an example of how one Library configured their Geographic Region facet during implementation. Page 10 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 FACET 4. Geograp hic / Region TAG POSITIONS/SUBFIELDS CONDITIONS LABEL Geographic region 651 a, e 662 a, b, c, d, e, f, g, h 600, z 610, 611, 630, 648, 650, 651, 654, 655 Ind. 2 = 0, 4 Ind. 2 = 0, 4 May databases be used as facets based on what search terms are entered? There are a couple of methods to accomplish this, depending on your preference. We can index your A to Z list of databases and then push the databases forward for the user to search. Or the Library can catalog their databases so they are returned in the results list. For example, at the University of Oklahoma (http://boss.Library.okstate.edu/), a search on “Early English books” brings up the result for the “Early English Books Online” database. The University of Chicago has also implemented database searching in this manner. Another method is to use familiar trigger words to direct the user to search particular databases. For an example of how this was done, visit http://aqua.queensLibrary.org/ and search “African American”. This search triggers a referral to the “African American Experience” online database. The Library can configure any number of triggers in the AquaBrowser Control Panel. The AquaBrowser Implementation team works closely with library staff during implementation to configure the parameters. Authorized staff will receive web-based training for using the AquaBrowser Control panel so the library can continue to tweak parameters after implementation. Alternatively, AquaBrowser is happy to implement the desired parameter changes for the library post-implementation. Federated Searching Is it possible to integrate Metalib to AquaBrowser instead of WebFeat? Page 11 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 AquaBrowser integrates with any federated search engine. AquaBrowser connects to any federated search interface that supports SRU but also any other form of XML-based output. So we can not only connect to Serials Solutions or WebFeat interfaces, but also to homegrown systems. We have the best integration with 360 Search from our sister company Serials Solutions. As for the categories, we can certainly use those. Our Serials Solutions connection will feature the possibility to use categories for databases and group results accordingly. We can use that interface functionality for the categories you use. AquaBrowser is very flexible in constructing display and links for record display. We're doing outlinks to many different systems, ranging from Syndetics viewable content to all kinds of reviews, information, ETC. from other providers. We create the proper URL encoding based on ISBN or other info needed into the BIR system (assuming it supports deeplinking to the extent we need it). Can the Next-Gen Catalog pass a search through to NExpress and MaineCat. Can the Next-Gen Catalog incorporate references to WebBridge resources. Article searching is possible in an indexed way, if the data is available at the Library, or in a federated search way. Federated search options include WebFeat, Serials Solutions, and any SRU based format. Federated search results are included in the indexed search results in a unique way that doesn't freeze the interface or interfere with the user's work flow. In the US, all of our federated search customers make users log in first before being able to do searches. AquaBrowser can be configured to pass any search through to any other system that supports getting a pre-determined query from the URL. All we need is a description of the target URL. AquaBrowser can be configured to always show the pass-through option, or only when there are no search results. Is it possible to group multiple electronic subscription databases into one category for federated searching purposes? AquaBrowser connects to the XML API that the federated search tool provides, so it is certainly possible to leverage the work that the Library has put into grouping the databases. A Federated Database subscription is required and may be purchased from the library’s preferred vendor. This option can be implemented at any time, pre or post implementation. AquaBrowser integrates with any federated search engine that supports SRU but also any other form of XML-based output. So we can not only connect to Page 12 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 popular federated search engines such as Serials Solutions or WebFeat interfaces, but also to homegrown systems. We have the best integration with 360 Search from our sister company Serials Solutions. AquaBrowser displays the results as they come from 360 Search. AquaBrowser does make use of the categories in 360 Search to organise the display. Fiction Connection We didn’t see the possibility to consult a global alphabetical list (browsing list) by author, title or subject. For our institution it’s a critical function. Could it be possible? Yes, this is certainly possible. In our implementations to date the libraries have not yet classified this functionality as critical. But we can implement it for you. One example can be found at http://www.fictionconnection.com/. The tag clouds are actually browsing lists. If you click on the link below that says "click here to view all topics" you go to http://www.fictionconnection.com/browse.asp, which gives you a full browse of six subjects. Note that the layout and design of this example is for a fiction search engine. FRBR (Functional Requirements for Bibliographic Records) Syndetics ICE Title Link as it is now is a very simple algorithm that identifies items in the Books In Print database that have the exact same title, subtitle and author. We use the entire contents of these fields when linking them up. We have about 1 million titles linked in the BIP database and these 1 million titles are linked to about 3.6 million ISBNs. It seems that FRBR principles (ex. Group similar editions of a title) are not used. Is it possible to confirm? FRBR principles are used by some libraries. We have a number of strategies. One strategy is to use the Syndetics ICE Tiitle Link module. In that product the FRBR information that is available in the Books in Print database is used to group editions of the same title together. University of Oklahoma uses this feature. If you go to http://boss.Library.okstate.edu and search for Freakonomics you can see this functionality. We can also work with you to implement grouping rules that you specify. We did this with the public libraries in Belgium. Please go to Page 13 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 http://zoeken.bibliotheek.be/?uilang=en. If you search for “fromage” you will find the book Fromage by Willem Elsschot on top. (He is a classic Dutch author.) As you see from the results there are actually 34 editions in 7 languages of this title. They are all grouped together by rules that the Belgium Library Association has specified for us. They are happy to share those rules for other AquaBrowser implementations. Is it possible to have examples of the FRBR rules used by the Belgian Library? In Belgium we create a FRBR-key for every title. These are the rules in pseudo code: If there is no author (100$a or 110$a) the key is 130$a (uniform title) If there is a 240 field the key is 240$amnrfgpko else if there is an original title (534$a) the key is 534$ and 100$a or 110$a else if there is 730$a the key is 730$a and 100$a or 110$a else the key is the title: 245$anp Notice that they adapted their cataloging rules to make this work to the fullest. But you can use this as a good starting point. There would be an additional customization fee for this kind of custom development FTP and Scripts - ABO We need to know in addition to enabling FTP accecss/function if our servers need to open another port to access data? In addition, we also need to see more of the script that is installed in our local servers such as the language functionality, what actions or steps it takes to conduct the script, etc. Does the script operate like a webservices? The library creates a full MARCout file and ftp’s that file to the AquaBrowser Online server using an ftp account created specifically for them. The AquaBrowser Online server is on our network, no applications or scripts are installed on the local library Horizon system. A web service is created on the ABO server for the library We import the MARC records and create a database on the ABO server for the library. When the initial search is done it is done against the ABO database and the user sees a list of titles and other details about the title (author, call number, etc.). Page 14 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 When a title is selected we do an http request back to their Horizon catalog using the URL of their Horizon HIP server and we add the control number and other details to the expanded URL and present it as a ‘search’ to their catalog. The information is returned to ABO and rendered on the full record screen. Google Book Search Successful integration of "About this book" pages from Google Book Search service into Ex AquaBrowser. Using a new “viewability” application programming interface supported by Google Book Search, Library patrons can now enhance their findings with Google Book Search features such as full text, book previews, cover thumbnails, and a mashup from Google Maps linking pages in a book describing a specific place to its location on the world map. Available now at Okla State Univ. AquaBrowser Comment: Yes, this is available and in production. See integration with Google books API below from the University of Chicago. Google Books API Hardware specs and browser compatibility Page 15 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 AquaBrowser runs on Windows Server operating system and Intel-compatible server hardware. AquaBrowser does not require database management system software licenses. The library does not need to purchase Oracle or any other DBMS for AquaBrowser. AquaBrowser Library requires a dedicated webserver for optimal performance. Library systems with over 750,000 titles and over 5,000,000 annual circulations may wish to implement two identical servers: one as primary and one for indexing, testing, and hot-spare / load balancing. For most libraries and Library systems a dedicated, a server class Windows OS machine with 2-4 GB RAM and 2 mirrored HDD (minimum 150 GB each) is sufficient. Generally a site your size would require a mid range, Windows server class machine with: • CPU: Dual core 4Ghz Intel-based CPU (recommended: 6Ghz, quad core) • 2-4 GB RAM (recommended: 6GB) • In addition to the C: drive with OS and a separate (fast) hard drive of 160GB minimum • - 100MB or 1GB Network Interface Card - Inbound access on port 80 - Outbound access on port 80 - 150Gb free HD space (or more) Software: Windows 2003 server. • Latest Service Packs • Internet Information Services (IIS) v6 installed 4/08 minimum OS has been changed to Windows Server 2003 from Windows 2000. Existing customers running Windows 2000 can still be upgraded and get the latest beauties except for the automatic FRBRization algorithm (a.k.a gabl engine). Sadly that implies that for My Discoveries, libraries need to upgrade to Windows server 2003 PS. Title link from Syndetics ICE can still be used on Windows 2000 machines. Typical system requirements for a large library (around 1-4 Million title records) are: • • • • • • 3Ghz+ quad-core CPU Windows 2003 server Internet Information Services (IIS) v6 100MB Network Interface Card (recommended: 1GB) Inboud and outbound access on port 80 200 GB free HD space Page 16 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 • 4GB RAM (recommended: 8GB) Workstation Requirements: Operating System Windows 98/NT/2000/2003/XP Mac OS X (10.3+) Linux/BSD • Browser Firefox 1.0+ Internet Explorer 5.5+ Safari 1.2+ Firefox 1.0+ Firefox 1.0+ Macromedia Flash Version* 7+ 8+ 7+ Flash is required for word cloud animation. If Flash is not installed, word suggestions are displayed in a text format rather than an animated word cloud. For the control panel (which is not publicly accessible), only Firefox is officially supported. Harvested Websites – examples • Howard County Public Library (http://aquabrowser.hclibrary.org) This sites crawls the library website and Medline Plus Health. • • Santa Cruz Public Libraries (http://aqua.santacruzpl.org) This sites crawls the library website. Toledo-Lucas County Public Library (http://aquabrowser.toledolibrary.org/aquabrowser) This site crawls the library website, Librarian’s Internet Index, Internet Public Library, Discover Ohio, Library of Congress Reference Shelf. You can find these in the bottom refine facet. Public Library of Amsterdam (http://zoeken.oba.nl/?uilang=en) This site crawls the library website and indexes a wealth of other sources. • Hosting Service For a hosted version, the same level of customization is available. Page 17 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 Icons AquaBrowsers supports a highly configurable and flexible mapping from any data element value to one or more icons or texts to displayed along with the record. ILS Holds and Reserves May a student place a hold on an item in Voyager? AquaBrowser has implemented real time holds to all implemented ILS systems except Voyager, since Library policies at the live AquaBrowser sites on Voyager don’t allow users to place holds. However, we are confident we can easily implement this with Voyager. We would accomplish this as we’ve done for other ILS systems -- by deeplinking to Voyager or via ODBC access to the database. ILS Integration AquaBrowser will integrate with any ILS. We are currently working with a few open source examples. Part of the appeal of AquaBrowser as an “independent discovery platform” is our commitment and ability to simply “migrate” your AquaBrowser implementation from current ILS to “whatever” is your next choice or move in ILS. Our business model to the customer includes this service at no additional cost. Thereby giving the library the flexibility to choose any ILS and to maintain the finest in discovery layer tools for your students, faculty and staff. We are currently working with a couple libraries to port their AB installations from traditional ILSs to Open Source ones. Should the library choose to migrate ILSs in the future, there is no charge to move AB over. All the institutions that use AquaBrowser seem to keep their “classic catalogue”. Is it possible to have more information about the integration process? How does it work for the availability of a document (real-time, update)? How does it work for the advanced search? Do you use the NCIP protocols to have the status? Usually we provide a link to the current classic catalogue because there are users that are initially more comfortable with the old interface. Within AquaBrowser, we usually use web services to retrieve the real-time availability of documents. We can also use z39.50 or NCIP if required. At Trinity College, a web service on top of GeoWeb to accomplish this. Page 18 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 In the United States, AquaBrowser has integrated with all the major ILS systems as well as Open Source ones. Below is a chart showing our success with these systems: ILS Export Availability Holds Book bag Example site SirsiDynix Horizon SirsiDynix DRA SirsiDynix Unicorn III Millennium Ex Libris Voyager Ex Libris Aleph TLC Library.Solution TLC CARL.X VTLS Virtua Polaris ILS Follett ILS Sagebrush Winnebago Spectrum Koha Evergreen Infor Advance Infor GeacPlus Infor VubisSmart Infor Vubis Talis BOND Bibliotheca.NET Bibliotheca 2000 Page 19 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 Voyager - deep linking via OpenURL. Works byb referring to the Voyager bib record ID in the 001 field. Using that field we can deeplink directly to the full record page in the Voyager OPAC. We know from our experience with Oklahoma State and Stanford Graduate Business School that this works with Voyager. III - AquaBrowser Library has a proven record integrating with III Millennium systems, including real time holdings extraction, placing holds directly from the result list and integrating into III’s book bag functionality (allowing patrons to place holds on multiple items simultaneously). For an example of the tight integration, please visit AquaBrowser at the King County Library System at http://explorer.kcls.org. Using an established modification to GeoWeb, the integration can be made tight in order to allow the display of real time availability and placing holds directly from the AquaBrowser interface. For an example of this integration please refer to the implementation at Trinity College, Dublin , Ireland on http://stella.tcd.ie) . We will customize operating procedures to support an interface with your ILS in the following way: Set up an automatic nightly import of bibliographic and holdings data in a MARC format. A complete bibliographic file with holdings is preferred, but if your ILS cannot provide this nightly, a file of incremental changes (dependent upon your ILS output) can be used. In this case we recommend a complete database import be performed monthly. Geac – AquaBrowser can and is working successfully with GEAC Advance (GeoWeb). Set up a real-time link to your GEAC OPAC for holdings and item status information. We can support a deep link via OpenURL (or any agreed URL format) to your native full record display ILS Questions How is bibliographic information pulled from the Library catalog? Does it pull the full record or only specific fields? If the full record, does this include only marc fields or does it pull proprietary fields (e.g., Innovative barcodes). Does it pull all bib records (suppressed and unsuppressed). As for the ILS title record exports: AquaBrowser can pull anything that the Library export provides. Detailed questions on exporting records from an ILS will have to be answered by the Library or the ILS provider. What we do know about III millennium is it can export full records in MARC, and you can set it to export specific additional native fields, such as the internal ID, to for example a 949 field. Page 20 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 AquaBrowser can combine multiple holding files from different systems, even from different vendors. It can de-dupe records for display on any combination of fields -- some customers use the ISBN field. An example of a multiple institution AquaBrowser is Harris County at http://newcatalog.hcpl.net/, which combines an academic and two public libraries. Note that individual institutions within such a site may have their own view including data filters, as seen at http://newcatalog.hcpl.net/HCPL/. These views called 'skins' may even include both visual and configuration customizations. By what means or method does AquaBrowser interface with our ILS to get the real time circ status of items? Deeplinking is a choice to the customer: AquaBrowser supports screen scraping pages for real time information; it also supports connecting to Voyager's underlying database directly. Both look the same in the user interface. Also support Z39.50 and NCIP to get item information. Implementation Process & Tasks Prior to actual implementation, AquaBrowser will work with your key staff to review configuration and standard customization options (colors and logo). If customization services are purchased, the AquaBrowser Project Manager will work with you and Medialab staff in The Netherlands to accomplish this. During these discussions and webex sessions, we will identify pre-installation tasks for your Library and us to perform. System integration tasks will be performed by our staff remotely. Implementation – Sample Timeline Week 1: Library confirms order. Library and AquaBrowser sign Contract. AquaBrowser invoices Library for 50% of Purchase Order. Library acquires and prepares server. Library assigns IP address & server location. Library authorizes Remote Desktop (or VPN) access. Library provides copy of bibliographic & holdings data via ftp to AquaBrowser for initial load. Week 2: Schedule implementation activities. Weeks 3-4: AquaBrowser and Library complete action items assigned in weeks one & two. First deliverable is met and 30% is invoiced. Week 5: Final Implementation and Training activities Week 6-7 Library staff review functionality and decide on any parameter changes. AquaBrowser implements any requested parameter changes AquaBrowser addresses any functional issues AquaBrowser completes implementation and invoices Library for balance of order, 20% usually. Page 21 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 This broad outline with actual dates can be changed to suit your calendar constraints (e.g. working around holidays). And, depending on the amount of customization purchased and the number of add-ons ordered, this schedule may be expanded to accommodate more complex project requirements. Your AquaBrowser Project Manager will work with you to reach a mutually agreeable timeline for installation. AquaBrowser will provide the library with the “AquaBrowserLibrary Standard Implementation Workbook”. During a kick-off session, we explain the workbook and the choices you have for the configuration. The first step is for the library to provide a copy of bibliographic and holdings data in MARC21 or similar format from the current library system. Medialab will setup a first prototype of the AquaBrowser Library site with your data for you to verify. Based on this prototype and your requirement, the Library will with our help specify the configuration items that need to be implemented. One of the most time intensive parts for the Library is deciding upon the facet configuration. Often we will have a couple of iterations before we settle on the configuration for go live. For communication purposes we make use of the online project collaboration environment Basecamp (http://www.basecamphq.com/) Integration Services provided by Medialab include: 1. Verify and update your Windows Server OS, IIS Web Server software, and install AquaBrowser. 2. Set up an automatic nightly import of bibliographic and holdings data in MARC21 or similar format. A complete bibliographic file with holdings is preferred, but if your ILS cannot provide this nightly a file of changes (dependent upon your ILS output) will be used. In this case Medialab recommends a complete database import be performed monthly. AquaBrowser indexes are reconstructed nightly. 3. Set up a real-time deep link to your ILS OPAC for holdings, item status and patron services. We can support a deep link via OpenURL (or any agreed URL format) to your native full record display (which includes links to patron-placed holds and other services). 4. Implement configuration as to the requirements of the library. Please see below the summary of the implementation process at Oklahoma State University. Note that a lot of the lead-time of the implementation project is needed for communication and internal discussion about configuration at the library. The full presentation from Anne Prestamo about Oklahoma State University’s AquaBrowser implementation can be downloaded from http://gslis.simmons.edu/podcasts/index.php?id=48. Page 22 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 Mar 22 ‐ ‐ ‐ loaded bib/holdings records toMedialab test server real-time circ status retrieved by deep link toWebVoyage Opens WebVoyage Brief Record in newwindow Mar 30- Signed letter of intent forAquaBrowser (during ACRL) Apr-June - Continued customization and modification work on test Server - Serials holdings displays – Relevancy algorithm – Refine elements – Voyager data structure (item type ≠ format) – Language issues (Check out) June – signed agreement for Syndetics June 12 Syndetics ICE content loaded to test server July 26- AquaBrower software installed onlocal server Aug Aug Aug Aug Aug 14- Real-time circ status pulling fromVoyager for full record 15 - FTP for daily Voyager exports to local AquaBrower server 17-19 - Final fixes and tweaks 20- Link to BOSS on Library Homepage 21- BOSS Search box added to Homepage Installation Method AquaBrowser requires a dedicated server. The Library provides a server per the recommended specifications. The server should have the operating system installed and a Remote Desktop connection via the internet configured. (AquaBrowser can use a VPN connection if your policies require that method instead of Remote Desktop.) AquaBrowser will perform remote installation of the AquaBrowser application software and configure it using the Library-defined parameters. Item Availability Indexed availability is used for refining on and can be indexed nightly or more often. The availability shown on the full record is always queried in realtime. Indexing Routine Note: during index swap-out replacement (usually once per 24 hour period), the website will be unavailable for a few seconds. Page 23 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 AquaBrowser indexes are rebuilt nightly and usually takes just a few hours. This does not require a server restart. The nightly import does not disrupt the website until the updated database is built and ready to be put into place. At that time the website is stopped, the new database moved into place and the website is restarted. That takes but a few seconds and is part of the automated scheduled tasks that require no manual involvement. One side effect of this is that there are two sets of indexes on disk during most hours of operation—if there is a failure of the active set, operations can roll production back to the previous set. For this reason, there is no need for an extra backup to other media. Furthermore, since AquaBrowser is never maintaining master data, the only files on the AquaBrowser server that need to be backed up occasionally are OS and application software. This is normally done over your LAN to whatever backup array is currently used for other Windows application. In the environment described here, there is not a need for scheduled downtime or reindexing periods as one is accustomed to seeing in traditional systems. Using a control panel, the customer can configure whether individual data sources are included in the regular import schedule. For file-based harvesting, the customer can influence the schedule of imports by selecting when new files for harvesting are provided. Maintenance – Library staff skills No special skills or programming languages are required in order to maintain AquaBrowser. Medialab staff perform upgrades for the Library. Maintenance & Support Services Remote Desktop support for software maintenance and software support is provided by the support organization at AquaBrowser North America. The AquaBrowser North American support team is backed up by Medialab technical and support staff in Amsterdam. AquaBrowser staff performs upgrades and installs patches to the Library’s system as part of your annual maintenance and support agreement. What does the AquaBrowser maintenance fee cover? unlimited customer support (with no restrictions on the type of support) as well as all new releases, upgrades and enhancements to the AquaBrowser software? AquaBrowser Comment: Ruth, yes there are no restrictions on support. It is not limited, for example, two calls a day, or so many hours per month. We want your software to work correctly and we service it accordingly. Page 24 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 There are services where additional charges apply. This is for example if they request for some high customization down the road or after a year. For example, they might request to add a google mashup to the full record view based on certain subject headings. This would require a development calculation with quote from us. Will we ever need to pay additional costs for an enhancement or a software upgrade? AquaBrowser Comment: Maintenance costs are fixed. Upgrades and enhancements are provided at no additional costs for the software modules you purchase. We may developed new modules (such as our MyDiscoveries options) and there is a charge for these optional components. Can you elaborate on how the maintenance fee is determined. Is it a percentage of the cost? Other model? How does MediaLab arrive at this cost? AquaBrowser Comment: Annual, on-going maintenance costs are a percentage of the cost of the software purchased. What exactly are we getting for the maintenance fee? a. How does MediaLab approach technical support? AquaBrowser Comment: There are three avenues (contact points/methods) for our customers: phone support to both our implementation and support team in the US and our headquarters in Amsterdam, email support and our on-line support mechanism, Group Hub (aka Basecamp). Each of these contact methods is designed to quickly and efficiently resolve issues to the customer’s complete satisfaction. Our vast customer base is the best source of your assurance of how we behave toward support. Please feel free to contact any of them again and specifically ask them about our level of support. Does MediaLab have root access to our servers to troubleshoot problems or do we troubleshoot with guidance? Is technical support “hands-on”, where you connect to our system and perform updates, maintenance, etc., or does someone here do this action with support if needed. AquaBrowser Comment: The AquaBrowser architecture is different than what you might have been accustomed to. We will only access your servers with your permission and only in the limited fashion that your IT staff if comfortable with Your AquaBrowser implementation essentially runs on a mid-sized server and at a high degree of reliability. For example, the largest and perhaps busiest libraries in North America, Queens public library has Page 25 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 never suffered a mission critical outage. Occasionally, we will need someone to perform some minor operations at your site ----usually something as simple as rebooting a server. Our goal, (and one that we have been very successful at) is to minimize impact on staff time. As for root access, Medialab needs local administrator access to the server that runs aquabrowser. Medialab does troubleshooting if necessary. During the process the library staff is trained on this troubleshooting for them to troubleshoot in the future. b. Do we have total local control over web pages, and all settings or do we need to contact MediaLab in order to have certain functions performed? AquaBrowser Comment: AquaBrowser is highly customizable and we will work with you to make changes. Many of our customers have expertise on site to make changes and with just a few minor exceptions, we both allow and encourage. If there are things that only MediaLab can do, please specify what types of things would these be. AquaBrowser Comment: We provided parameter driven tools and administrative console that gives the customer nearly full control. Certain proprietary functions within the software are “shielded” from direct user contact. We are happy to deal with this level of change for you and from a matter of practicality, it is simply easier for us to perform this service for you as may be need as part of the maintenance and support that we provide you. c. Are there any other pricing models for maintenance fees, such as a per hour model or a separation between the upgrade and the support costs? AquaBrowser Comment: There are no additional or per hour fees associated with AquaBrowser. Our maintenance (on-going) costs are as stated and are consistent with maintenance fees across the industry. Is there a hosted option for the software? Or is that some other model that does not require us to purchase and support the software? AquaBrowser Comment: We can provided a hosted solution for you that eliminates the need for you to purchase and support hardware locally. The software support is provided regardless of your choice of deployment (local server versus hosted solution). We can provide costing information for a host solution upon request. However mid-sized institutions, that have reasonable access to competent IT support, have usually opted for an on-site server solution. If one or the other party were to discontinue our contract after we have implemented AquaBrowser, what do we have left? a. Do we still have functioning software at the version that installed at the time? b. Do we lose any functionality? Page 26 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 c. Is there anything else about this that we should know? AquaBrowser Comment to 7 a and b above: Should you decide to terminate your license agreement the authorization to use the software expires upon the end data of your service agreement. AquaBrowser Comment to 7c above: Since AquaBrowser is both highly portable and complete integrated library system neutral (i.e. can be used with any ILS), our agreement with our customers is to “move” AquaBrowser to any new ILS you might choose, at no additional cost. Medialab and Bowker – company info Medialab Solutions BV is the creator of AquaBrowser. Medialab and its AquaBrowser North America division are owned by the RR Bowker Company, located in New Jersey, and are part of the Cambridge Information Group (CIG). Collectively these companies have been providing services to libraries and the publishing and information market for over 150 years. Medialab was founded in 1990 by professor M.M. (Thijs) Chanowski. Until 2000, it was a dedicated research facility and part of BSO Origin Philips. In 2000, Medialab entered the market and turned its scientific achievements into successful searching solutions for companies, non-profit/governmental institutions, and libraries. Today Medialab focuses all of its energy in developing AquaBrowser for libraries. Medialab Solutions BV employs 20 individuals in the Amsterdam, The Netherlands headquarters. Four individuals are employed in the North American virtual office. RR Bowker, the owner of Medialab Solutions BV, is a financially sound, privately-held company. Bowker has been in business for 136 years. Cambridge Information Group which is a privately owned company acquired Bowker in 2001. Since then Bowker acquired many businesses including Syndetic Solutions in 2004 and most recent one –MediaLab’s ABL in 2007. Due to the fact that Bowker is privately owned, we are limited to disclose any financial information. Bowker revenue for 2008 is projected at $50M. In 2007 We have offices in US, UK, Australia and Netherlands with 225 employees worldwide.” For more information, you can visit our website: www.bowker.com. Medialab & AB – staff Terry Parks, Manager of Implementation and Support Services, North America. Terry has spent considerable time implementing the large AquaBrowser installations in a project management role. She will schedule implementation resources, and work with Medialab staff in The Netherlands Page 27 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 to perform installation, customization consulting, and training services. She will manage the project, insuring that other AquaBrowser North America and Medialab resources are applied as needed. Terry has just implemented the first AquaBrowser Online hosted academic library in the U.S. Below is an example of some of the recent and ongoing projects Terry has managed: Columbus-Metropolitan Library – heavily involved in project since October 2007, facilitate weekly project conference calls, provide assistance to primary Medialab project lead in Amsterdam. Free Library of Philadelphia – heavily involved with project since October 2007, facilitate regular project conference calls, update design and database change requests for AquaBrowser upgrade site, work with Amsterdam staff with customization requests. Site is live, work is being done on development server to provide enhanced features. Contra Costa County – facilitate regular project conference calls, work with CARL.Solution staff in coordinating efforts, performed Administrative training, with User training scheduled March 18, 2008. Tim Jones, Senior Implementation Engineer, North America. Tim has also worked closely on the AquaBrowser implementations on a very technical level. Most recently, Tim performed most of the implementation tasks at Contra Costa County Public Library in CA. He teams with Terry and Medialab staff in The Netherlands for customization services and integration with the III Millennium ILS. Tim holds a B.A. from California State University and a M.L.S from Indian University. He has worked in OPAC implementation and support since 1996. Jasper Kaizer, Vice President Product Management, Medialab. Jasper has played a central role in the AquaBrowser’s beginnings in the library market, starting with the Dutch Public Library Association. Mr. Kaizer is responsible as the key representative to all library clients and potential clients. Mr. Kaizer is also the bridge between Medialab and the strategic partners around the world. Mr. Kaizer holds an MSc in Technical Information Systems. Taco Ekkel, Director, Application Development, Medialab. Taco has been intimately involved in the creation and growth of the AquaBrowser Library product since its entrance into the library market. Currently, he oversees the development team, ensuring architectural consistency and bridging current and future market needs to development on the core product. Mr. Ekkel holds an MSc in Computer Science. Mobile Computing Interface   (is there one? for what kinds of devices?) Page 28 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 Owing to the unlimited possibilities of hand held devises we have not yet deployed a version of AquaBrowser that can truly be called “compliant” with these devices. Multilingual AquaBrowser Library supports multiple languages in three different ways: 1. User Interface Prompts, Messages, and Help Text can be implemented in any language defined in Unicode. Libraries may opt to implement Catalan, Dutch, English, French, Frisian, German, Maori, Norwegian, Russian, Samoan, Spanish, and Swedish by means of a control panel setting and a URL switch (e.g. http://hhs.medialab.nl/?uilang=en). User interfaces in other languages may be purchased as part of “customization services”. Medialab waives the price of this customization for libraries who do the translation work to add a new language to the current set. 2. Translations (and Synonyms) in the Discover word cloud The word cloud can be configured to include (along with word associations and spelling variants) translations and synonyms from the Ergane multilingual dictionary http://www8.travlang.com/. Most TLC implementations include only Spanish translations and British spelling variants (e.g. gaol = jail). Any library can enable translations in other languages in the Ergane dictionary at no extra charge. 3. Indexing and Searching Text in any Roman script can be indexed and searched. This includes English, Spanish, Polish, Vietnamese, and other languages encoded as MARC-8 (“ALA character set”) typically found in library catalog records. It also includes data from other sources (e.g. web pages) encoded as Western or Unicode UTF-8. AquaBrowser also supports special handling of English, Dutch, and German by “decompounding” combined words into multiple word components. Data in non-Roman scripts such as Chinese, Russian, and Arabic encoded as Unicode UTF-8 can be displayed in “full record” views. At this time nonRoman scripts cannot be indexed or searched. • • • The default UI language can be set in the control panel (global settings) Depending on the customization, different languages can be activated per user using the URL Translation has an effect on all product labels and strings, including every UI element. Page 29 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 What exactly would we get with the various languages? Does this allow a user to change the language of the screen display as they can at some web sites? If not, what would we be getting? AquaBrowser interfaces come “out of the box” in English, Spanish, French, Dutch, German, and Italian. The user can pick the language. Please see this link: http://zoeken.oba.nl/?q=horses&x=47&y=9 where the user can simply choose “English” and the Dutch site is now English. User Interface languages that can also be offered currently include: English, Spanish, Dutch, German, Catalan, French, Frisian, Russian, Swedish, and Norwegian? Is it possible to load a thesaurus? A translation of the LC Subject Headings (in French)? Absolutely. We will need a file with the translation of the LC Subject Headings and we can then load those. The next step is your choice whether you want to automatically expand your search with translations or show translations as suggestions to the user. You can see how we use an authority file at university of Chicago. Search in LENS (http://lens.lib.uchicago.edu/) for architecture, and you can see above the results that the query is automatically expanded with the SEE references. The SEE ALSO references are shown in the word cloud as suggestions. User Interface Prompts, Messages, and Help Text can be implemented in any language defined in Unicode. Libraries may opt to implement Catalan, Dutch, English, French, Frisian, German, Maori, Norwegian, Russian, Samoan, Spanish, and Swedish by means of a control panel setting and a URL switch (e.g. http://hhs.medialab.nl/?uilang=en). User interfaces in other languages may be purchased as part of “customization services”. Medialab waives the price of this customization for libraries who do the translation work to add a new language to the current set. Is it possible to change language in the middle of a search, not only at AquaBrowser 's home page? This is not standard functionality, but we could implement this with some small customization for you. Translations (and Synonyms) in the Discover word cloud The word cloud can be configured to include (along with word associations and spelling variants) translations and synonyms from the Ergane Page 30 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 multilingual dictionary http://www8.travlang.com/. Most implementations include only Spanish translations and British spelling variants (e.g. gaol = jail). Any Library can enable translations in other languages in the Ergane dictionary at no extra charge. The word cloud is capable of displaying non-Roman characters. If references in the loaded authority file or thesaurus include words with non-Roman characters, the word cloud displays these references correctly. The cooccurrence algorithm for the associations only supports the Roman characters. If search terms are entered in simplified Chinese, will AquaBrowser return results in traditional as well as Pinyin for those results that match the simplified Chinese terms? AquaBrowser stores and displays non-Roman scripts. Searching Chinese is in development for the University of Chicago project. Medialab is working on implementing this and will deliver it by the end of March 2008. This is how the Chinese searching will works: In the indexing, all simplified Chinese is translated to Pinyin. If the query is entered in simplified Chinese, the query will be translated to Pinyin, which is actually searched. Results that are in both Pinyin and simplified Chinese are returned. Note that this translation is on character basis. If a Library would like to provide a translated table of simplified & traditional Chinese, we can implement this so a search retrieves both forms, regardless of what form was searched. Indexing and Searching Text in any Roman script can be indexed and searched. This includes English, Spanish, Polish, Vietnamese, and other languages encoded as MARC-8 (“ALA character set”) typically found in Library catalog records. It also includes data from other sources (e.g. web pages) encoded as Western or Unicode UTF-8. AquaBrowser also supports special handling of English, Dutch, and German by “decompounding” combined words into multiple word components. Data in non-Roman scripts such as Chinese, Russian, and Arabic encoded as Unicode UTF-8 can be displayed in “full record” views. At this time nonRoman scripts cannot be indexed or searched. My Discoveries Page 31 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 Is there added functionality for users who log in. If so, is the login an existing campus or Library login, or is this a login specific to the Next-Gen Catalog. Does all user-interaction functionality (e.g., tagging, ratings) require login or could a user interact anonymously. AquaBrowser has a social layer called 'My Discoveries', that asks users to login to it. Anyone can make an account, independent of a campus/patron login. All the actions in which a user can create or contribute content require logging in. My Discoveries allows users to tag any results, or add results to any number of lists that users can create themselves. Please provide more information about “My Discoveries”, all the web 2.0 function (tagging, comments, users profile, etc.)? Do the “My Discoveries” include suggestions for the users (ex. The people who took out this book also like….)? Do “My Discoveries” include a personal profile for the users and is it possible to save a search and keep it in this profile? We love to give you more information about my discoveries. You can start with the presentation of My Discoveries available on Slideshare: http://www.slideshare.net/AquaBrowser /my-discoveries/ It is probably best if we schedule an online demo on My Discoveries for you. Interesting aspects are that the tags and comments are shared with other libraries in order to make a larger pool of available tags. Also you can prepopulate the tags with the tags from LibraryThing.com. The titles tagged in LibraryThing are becoming more and more relevant for academic libraries. Is it possible to select a list of records (concept of "basket"), to save, print or email? Yes. This functionality is standard available in My Discoveries. It is also possible to request a custom implementation of baskets. You can see this at Columbus Metropolitan Library (http://catalog.columbusLibrary.org). On every full record page you can add the title to a book bag. The book bag is accessible from the control bar (next to the search box). Note that this was custom development where they only wanted to give the ability to add to the book bag from the full record page and not the results page. March 2008: There are two libraries live with My Discoveries - Croydon Univ in London and Amsterdam Public. URLs are: http://titan.croydon.ac.uk/ipoint/ (see below) and http://www.oba.nl/ for Amsterdam's site. Arlington County Public in VA will be going live later this month. Page 32 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 User-contributed content & collaborative environments: We understand the My Discoveries module handles this and is a subscription. If the Library stopped its subscription, what happens to the data added via the State Library websites? Who owns this data? This data is owned by the patrons themselves; by using my discoveries, patrons allow medialab to use that data. Of course, a library may access the data added by its patrons. OAI AquaBrowser easily harvests metadata from OAI servers such as CONTENTdm, dSpace, etc. In addition, AquaBrowser is versatile at importing non-standard formats. Direct indexing of visual and audio formats without accompanying metadata is not supported. Other Data Sources If we wanted to search Mura and Digitool inherited search is this an additional cost? AquaBrowser can connect to any additional data source and we will provide integration and customization services for “connection” to any federated search tool. There is a charge for both the customization as well inclusion of added MARC on non-MARC data sets. Is there a limit on item types and MARC formats? We have quite a few of both. Page 33 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 AquaBrowser supports MARC format(s) in its various nuances. We support US MARC and CAN MARC. Generally speaking the MARC format (MARC 21) and the pursuant ISSO standards that govern the processing of MARC records are supported. Since AquaBrowser is an independent discovery platform that connects to any data source, there is no “limit” to either item types or MARC formats. Can we customize the screens to add new information based on data elements in the record? For instance, the way we added the FindIt/SFX information to our detailed serial records required being able to customize the HIP detail to a great extent, and also having access in that customization include logic that ISSN/ISBN of the particular record being displayed. We believe that you mean can AquaBrowser be “rigged” to present different information paths based on the query “term(s)” – Yes. Go University of Chicago and search for --- Chaucer – note that the Oxford Biography link appears. Now search for.. tanks … note that the Oxford Biography does not appear as there is no entry in that file for tanks. Non-MARC Data Can facets come from non-marc fields in marc records. Can facets come from non-marc fields in non-marc records. Does it pull these records from the Library catalog (either complete records or selected fields): • • • • Resource records (non-marc) Non-marc bibliographic records (reserve records) Item records (non-marc) Check-in records (non-marc) Would loading finding aids (HTML) and Caption list for photographs be seen as an additional non-MARC data set? Any tagged data can be included in the AquaBrowser presentation to the user, including locally developed captions and other added nonMARC data sets. MARC Data Additional MARC datasets – Do we need this to integrate our authority records? Page 34 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 Yes and No. First, to fully incorporate and leverage your authority control file, we do need to treat this as an added MARC database for integration purposes. For example, at the University Chicago, (http://lens.lib.uchicago.edu/), search for “tombstones. Note how AquaBrowser indexed and leveraged the “See and See Also” pointers out of the Authority Control file and linked them to the bibliographic file record. It did something else as well, it expanded the terms and referred me to Sepulchral monuments in the word cloud. This is highly elegant to the user unfamiliar with AC headings. However, since AquaBrowser fetches some of the entries into the word cloud from the bibliographic records (note not the authority control file, the bib file), it does find in the 5xx and 6xx appropriate, related terms to the search term entered. However, powerful this is, it is not as powerful as when the authority file is integrated as an additional MARC dataset. Do we harvest data from our catalogue to AquaBrowser ? Initial implementation includes the exporting your data file(s) to MediaLab for processing. Nightly loads to update AquaBrowser are performed from an FTP site that is established for this purpose. What format is the data held in AquaBrowser. MARC? MARC XML? AquaBrowser maintains a separate index in a SQL database that does not rely on MARC. Overview of AquaBrowser – Search, Discover, Refine Search finds results that are relevant to you Searching is easy: Just enter some words in the text box above and press Enter (or click 'Search'). You will get search results, ranked by relevance. Click on a result to get a more detailed description. AquaBrowser Library allows patrons to search for information using a Page 35 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 standard query box, the way they are used to doing. The results are portrayed in a typical browser search list by relevance to the query terms given by the patrons. Clicking here takes the patron to the exact sources, both inside and outside the library. AquaBrowser Library's capabilities transcend the limits of conventional OPAC searching. Its finds items using associations, context, and spelling alternatives automatically generated from your library's catalog. AquaBrowser Library offers the first progressively interactive searching experience your users have ever encountered. 1. Enter your query in the text box above and press Enter (or click 'Search') 2. In the central panel, you will get search results, ranked by relevance. 3. Click on a result for a full view with holdings information. Discover helps you explore On the left side of the screen, you see a cloud of words based on your query. These words are suggestions for you to discover new information and help you in formulating your query. These suggestions consist of translations (to another language), spelling variations (to correct for spelling mistakes or words that can be spelt in more than one way) and associations (free 'brainstorming' associations). Clicking on any of these terms in the word cloud yields a new search combining that term with the original one. The system thereby ranks relevance by this history, finding the combination on top of your list of results. The words you clicked on in the word cloud are shown in a different color, as a trail of suggestions in the word cloud, thus never losing your original term. On the right side of the screen, you'll see lists of possible options to refine your result set. The magic of AquaBrowser Library's discovery principle is the “word cloud“ which appears with the most relevant associations, translations, spelling variations and synonyms. Clicking takes patrons to related sources, offering a multitude of choice and broadening the scope. In addition to a search list, AquaBrowser Library compares a user's search terms to the metadata in your catalog to create a visual map (a "word cloud") of associations or overviews of areas of interest. These words are suggestions for patrons to discover new information and help them formulate a query. The word cloud includes the Page 36 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 most relevant associations, foreign language translations, spelling variations, and synonyms. Patrons can discover the sources behind these associations and broaden the scope of their search. Conventional catalogs require exact search terms; AquaBrowser Library finds items using associations, and stemming and spelling alternatives. It also answers the needs of users who aren't quite sure how. A constellation of words on the left shows you related terms, spelling variations and translations. Click on words in the constellation to explore the contents of the catalogue. Refine helps you understand what you've found Users can zoom in on the details by picking any search option, such as a specific subject, author, category, or a language, etc. By doing this, users can not only find the exact information they're looking for, but also from the type of resource they want the information to come from. Clicking an option filters your results so only those results having your option in it remain. Refine shows you what your search results contain. Click on any term to focus and narrow your results. Notes • The search processes are always reproducible • The original search term remains part of the search until you decide to start over. • The refine section organizes resources by category, allowing you to distinguish between books, magazines, newspapers, web pages, etc. This allows you to zoom into a specific part of a result that was already found, but you still wish to reduce the number of results. • Refine helps you to understand what you found. • Search results are organized by the most relevant. You are not flooded with endless lists of results merely because your term appears in them. This saves you time, energy, and makes you a more accurate searcher • In the Discover tool, the size and color of the words in the cloud have a meaning: Page 37 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 • words are variations; these terms have something to do with your query, and can help you formulate a direction in your search; click on an association to search for it in the context of your current results. words are spell variations. Sometimes, there are multiple ways to • spell a word, or spelling inside the catalog. Click on a variation to see results for it. words are translations; click on a translation to search with your word in another language. • • words are thesaurus terms; click on a term to search the catalog using that term. words are terms you have already visited. As for the word size: • the larger a word, the more relevant it is deemed by the automatic relevancy algorithm. A word's position in the constellation has no direct meaning. PDF docs Keyword searching of PDF documents is not available. Persistent URL Persistent URL, for record (available?) YES                          Persistent URL, for search   (available?)  YES  Performance 10 million items are easily supported. Proven in production are 5 million bibliographic records and a significant amount of catalog enrichment as well as other source data add up to 16GB of content at University of Chicago. AquaBrowser indexes 9GB of content for Amsterdam Public Library. This includes a catalog of 580k titles, full text data from 10 years worth of 7 journals and 6 newspapers, and other databases tallied here: 2,238,457 krantenbank 578,691 default (catalog) 280,967 muziekweb 109,625 keesings 69,946 financieeldagblad Page 38 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 68,045 telegraaf 55,696 tijdschriftenbank 42,596 davindiplus 16,552 aladin 5,949 act 2,708 prod 2,259 uitburo 2,027 spider 915 fabchannel 71 vandale References (Academic Libraries) University of Chicago James Mouw Assistant Director For Technical and Electronic Services mouw@uchicago.edu +1 (773) 702-8732 Oklahoma State University Anne Prestamo Associate Dean for Collection and Technology Services 216 Library Oklahoma State University Stillwater, OK 74078-1071 Phone: 405-744-9755 Fax: 405-744-7579 Email: anne.prestamo@okstate.edu   References (Public Libraries) King County Library System David Wasserman, Online Services Coordinator mailto:jmoffitt@kcls.org(425) 369-3433 2005 Implementation 3 million volumes III Millennium Toledo-Lucas County Public Library System Margaret Danziger, Deputy Director 419-259-5207 Page 39 of 58 AquaBrowser Library FAQ 2007 Implementation 2.5 million volumes III Millennium Columbus Metropolitan Library System Scott Fothergill, Director of Computer Services 614-479-3007 2006 Implementation 1.2 million volumes Locally built ILS system Last Updated: 5/15/2008 Queens Borough Public Library Contact Name: Fred Fishel, Applications Development Manager 718-990-8505 Implemented for trial use by staff and public in January 2005; launched as the default catalog in August 2005. 6 million volumes DRA Classic Library: Arlington County Department of Libraries Maureen Karl, Associate Director, Arlington Public Library 703-228-5981 Launched for public use along with the library’s migration to the CARL.X ILS in May 2005. 600,000 volumes TLC CARL.X Relevancy Ranking Relevancy can be tweaked per datasource. Tweaking includes weights, and fields used in different weight sets. AquaBrowser has three weight sets ranging from highly important to less important in terms of relevancy ranking. On what rules is the relevancy ranking based? AquaBrowser has extremely powerful relevancy ranking that packs years of experience in the field of search. Relevancy ranks search results according to information about the matching search terms in the results and the fields they occur in. The weights for all occurrences of search terms in a record are combined, and the result is used to rank the results. The weighting of search terms takes into account: which field(s) a term appears in, for example, an occurrence of a search term in the title is considered more significant than a term in a note field; and whether the matching term in an exact match or an Page 40 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 algorithmically/dictionary-determined spelling variations, exact matches being ranked higher. In the past, we've also employed inverse term frequency algorithms. Upon implementation, a Library can choose which data fields to apply for the results ranking, which stop words to use, and how to weigh the various search fields in relation to each other. Reports and Analytics Customizable reports are planned in development for 2008. Of course, current reports can be customized by AquaBrowser for a customization charge. How do web mapping tools integrate with AB reports so we can apply our own analytics? The AquaBrowser website has an open structure in which every element that is displayed on the screen is defined via xslt transformations and xml pages. To integrate with your web mapping tool, endpoints can be integrated in the search result listings, detail page or other places that you would like to track. We have customers that use Google Analytics to do their own analytics on top of AquaBrowser. In addition to this, the statistics that AquaBrowser collects can be converted to an easy XML structure. Detailed usage statistics are available in a visual format, in the Control Panel. These go beyond normal web server statistics in the sense that more information than just page views can be extracted Usage Statistics include: • Number of unique users (sessions) per month-day, weekday, or hour • Number of queries per day • Help and Item Access • Refine and dimension usage • Sort usage • Search method usage • Result page usage • Most frequently used queries overall (Top queries) • Top Orphan Queries (Most frequently used queries without results) • Most popular items • Frequent Visitors • Client Environments (usage of browser, screen sizes, and Java versions) Page 41 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 Results Display Result views can be configured using multiple layers of XSLT files allowing for fine-grained control over display and structure. Is it possible to display MARC records, on request through a button or a link of some sort? Yes. We can show the link to the MARC display in the traditional OPAC. An example of this can be found at Santa Cruz Public Library (http://aqua.santacruzpl.org). Or, AquaBrowser can show and display the MARC itself. An example of this can be seen at Howard County Public Library (http://aqua.hcLibrary.org). In both cases, go to the full record page to find the MARC link for that title. Can we see what the different icons are for formats recognized by AquaBrowser, particularly audiovisual material. For example is there one for MP3 recordings? AquaBrowser will support any provided material format Icons. See a couple of the possibilities in the screen shots below: Page 42 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 RSS Feeds What types of information is available for RSS feeds. The RSS feed by default shows title, authors, summaries, added date, and keywords, but can be configured in detail upon implementation the same way for example a brief record display can be configured. (This applies to all views, from brief citations to full record views -- all of these can be configured during implementation.) Please take a look at the following link as examples: http://www.columbusLibrary.org/ebranch/index.cfm?pageid=208 for their explanation and pre-configured RSS feeds. Page 43 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 The RSS page in AquaBrowser can be found at http://catalog.columbusLibrary.org/rssfeeds.aspx?q=pittsburgh RSS Feeds, for a search (available) You can “rig” an RSS feed to “live” execute any search, such as “organic chemistry petroleum gases” into any RSS fed search. AquaBrowser RSS feature supports news feeds of user searches, effectively saving searches in the users newsreader rather than the ABL server. AquaBrowser publishes RSS feeds of which the update frequency is determined by the client software, with a fine grainedness equal to the frequency of indexing (normally once per day) Save Records Save records  (can non‐authenticated users save records?) Yes, users don't have to be authenticated to the ILS to save records to lists in My Discoveries. Search Entry Persistence The API for search entry into AquaBrowser will be provided and can be designed into the users experience, for example, at the University’s home page. Searching Is it possible to do call number searching? Yes, it is possible to do call number searching. We can configure this the way you want to. An example can be seen at Columbus Metropolitan Library (http://catalog.columbusLibrary.org). A specific call number field is shown in the advanced search. Page 44 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 Security Provide a summary of the proposed solution‘s overall security approach and design The AquaBrowser front end is public and does not use authentication for security. The control panel is secured using Windows NTLM authentication that applies Windows account information to logging in. AquaBrowser allows for assigning three different levels of control panel access (view statitistics only; configure AquaBrowser functionality; configure AquaBrowser systemrelated parameters) to specific user groups in the Windows Network Domain. AquaBrowser itself as been screened to prevent malicious attacks from successfully extracting information from the server. The server itself stores only title information, so sensitive patron or financial information can never be found on it. Also, AquaBrowser has read-only access to the systems it gets its information from, preventing it from doing possible damage. AquaBrowser runs very stable. Should anything cause the application to become unstable,the IIS webserver is configured to automatically shutdown unhealthy processes and replace them with new ones. Since AquaBrowser runs only indexes of data that exists elsewhere, a full backup is not needed to preserve functionality; a customer may however choose to backup the index to speed up the time needed to get a system back online should a storage device fail. Provide details of the proposed solution‘s process for secure data transfers, including protocols used. AquaBrowser data transfer exists purely on the local domain as far as index and other source data is concerned. Only the public search interface traffic is projected to the outside network, in an unsecured (http, port 80) way. It is possible to run AquaBrowser over an SSL layer but we do not recommend this for day-to-day usage. Provide details of any security and privacy measures for user-contributed content, such as tags, comments etc. User-contributed content is secured using a handshake protocol between client and server. Shibboleth Page 45 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 Skins AquaBrowser Library can support additional, specialized indexes and sets of parameters. This feature might be used to create a staff user interface, with more elaborate search options than the public interface, a children’s interface, or an institutional website, with content limited to the website itself. AquaBrowser can implement multiple ―skins∥ to support a variety of specific user profiles (e.g. staff, adult, juvenile), but not user-specific views by individual.Skins can be displayed as different ―tabs∥ at the top of the site, so users may easily move between different views. Scoped views are called ‘skins’ in AquaBrowser terminology. They comprise custom filters on the dataset, custom settings, and custom visual configuration. Sorting Relevancy ranking is the default. Title, author and date are options. Other “sort” options, based on library customizations are also available. Spelling Variant Can features of the world cloud get turned off; such as the variant spelling?      The word cloud's variant spelling helps users find materials in other languages as well as other accepted spellings... Think of how many ways there are to spell Tchaichovsky (that's probably not one of them!). It's not something we've done in the past, and we would encourage the library to not do this. Did they give a reason why they didn't want this to display? Of course, the customer is always right, so if they insist, we could customize this for them for a fee. But really, we want to discourage eliminating help in the word cloud. It's what distinguishes AB from other products!  ambiguous prefixes, for e.g., "Mc" and "Mac"; If a search results in a ‗not found‘ result, the user will be directed to the other prefix. Staff Administration Page 46 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 Are there screen shots available that show the staff administration tool? Please see attached. Standards AquaBrowser has a wide variety of world class support for open standards. On the ingest side, AquaBrowser can process anything that is proper XML, including standards such as OAI, MARC, MARCXML, ONIX For data output, AquaBrowser supports RSS using either Dublin Core (DC) or MarcXML metadata schemas. Also, search results can be retrieved in SRU/SRW with either Dublin Core (DC) or MarcXML schemas. Communication to the client browser supports XML and JSON formats All of AquaBrowser‘s output is XHTML 1.0 compliant. Internally, all of the display and data configuration is done using XSLT. AquaBrowser supports SRW/SRU with CQL, and can be configured to support OpenSearch. Stemming, Spell Checking and Word Completion We are slightly confused by the functions of spellchecking, word completion and stemming. How does it work? Which services do you offer? Functions such as "Did you mean?" only seem to kick in when there are no result at all; is that right and why? Stemming is used in the searching. That means that if you search for a word (like car), other forms of that word are also found in your results. For example the word: cars. For stemming we use language rules to make sure we are searching on different forms of the same words. In addition to this, AquaBrowser also searches for compounds. So if you search for car, it will also find motorcar (if written as one word). Words that consist of multiple words are automatically detected. Both stemmed variants and compound terms get a lower relevance in the result ranking. Spelling variants are used in two ways. First of all we find a number (configurable, usually 5) of spelling variants of the word that you are searching. These words are guaranteed to occur in the index dataset. These Page 47 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 words are shown in the word cloud as suggestions. The reason is that we want to notify the user of occurrences of slight spelling variants in the catalog. Often this helps the user to find the correct term, especially if there are multiple spellings for the same word in the catalog. This happens a lot, for example, with a composer like Chaikofski. Also this reveals spelling mistakes in the catalog or older spellings from decades ago. We show these spelling variants as suggestions to the user, because it is very dangerous to automatically search with a spelling variant. Only 1 letter difference can give a word with a completely different meaning. We choose to have the user decide which word to search on, but we do offer him the options. If there are no results found, we do some extra processing on the spelling variants to suggest the best spelling variation as a "Did you mean". The algorithm comes down to checking for the spelling variants which variant occurs most often in the catalog. That is probably the "correct" one. Still arbitrary of course and that is why we give the choice and let the user know that there are no results for what he typed in. Functionality for ‘did you mean’ is provided using soundex algorithms. A large number of potential variations is generated, then real time de-duped against the index to provide only links that yield results. Syndetics ICE Enriched record: Is Syndetics ICE your only provider or do you work with BTLF as well? We currently only work with Bowker Syndetics for enriching, but they license Blackwell Table of Contents for use in ICE. If you have other datasets, we can use those as well to enrich records. For example at the University of Chicago the SFX knowledge base is exported and used by AquaBrowser to enrich catalog records for the serials that are in SFX as well. We noticed that the catalogue at the University of Chicago (LENS) seems to retrieve information that can only be found in the table of content. Are the tables of contents indexed? Is it possible to have more information about theses possibilities? Certainly. University of Chicago also bought Syndetics ICE. You can find more information about that on http://www.bowker.com/syndetics/ICE/. ICE allows you to subscribe to extra content (table of contents and/or annotations). This information is then also indexed in your AquaBrowser without the need to load the information into your MARC records. The enriching is done at the index level. Page 48 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 Are we correct that tables of contents, etc. are searchable only if we purchase ICE? Searchable TOCs are part of the ICE subscription and searchable by AquaBrowser . NOTE: If you have already cataloged TOCs in your MARC data, or have in some other way appropriately linked the TOCs to the bib records in some other fashion, we’d certainly be happy to consider making those searchable in AquaBrowser as well. With ICE, we are assuming that the additional content resides on your server and is offered up as needed. Is this correct? Yes, the ICE content is hosted and maintained by AquaBrowser. automatically provide updates, links, loading, verifying files, etc. We At which points can the content be offered? For example, a Michigan State University, links to reviews, etc. did not become an option until we had clicked into the full record. Can these links appear in a browse display? Since AquaBrowser is so flexible and customizable, the enriched content can be viewed at any layer. Compare the University of Chicago to the Oklahoma State implementation. They deploy different content at different layers. NOTE: Usually this is a function reliant to some degree on aesthetics, screen real-estate available and Library choice. The AquaBrowser implementation team will work with you and your team for best, most appealing and efficient arrangement for your content and your users. Is pricing contingent upon purchasing a certain amount of content? If we were to only choose two or three options, will the pricing be significantly higher than in our quote? The pricing for content will fluctuate depending on the display and index elements you choose. There are discounts for choices of higher numbers of content elements. Are there any Syndetics products that are not available in the ICE interface? The Syndetics content has two levels: view or display elements only and searchable (ICE). Some elements lend themselves to being leveraged by AquaBrowser searches, (e.g. TOCs and Summaries). Others such as book jackets lend themselves to display. All available Syndetics elements that enhance the user’s discovery experience are searchable. There are multiple Syndetics elements that are available above those index content elements in ICE. Does the Syndetics ICE title link module also include Books Out of Print, or only Books In Print? Page 49 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 Yes it also includes Out of Print. Does it include recommendations? (E.g.: Readers who have borrowed this title also borrowed these ones...) If it does, on which data is it based? We cannot include the recommendations based on circulation information. This is simply due to the fact that we don't have access to that data. We do receive the recommendations from LibraryThing. So in the LibraryThing package you can have these recommendations. Does it include a search history, valid throughout a full session? AquaBrowser keeps your session valid as long as you don't close the browser. Syndetics ICE (overview) Syndetics ICE (Indexed Content Enrichment) expands your AquaBrowser Library’s searching beyond the traditional MARC record by indexing Syndetics value-added content such as Tables of Contents, Summaries and Fiction Profiles (detailed descriptors for genre, topic, setting, timeframe, character names and character traits on works of Fiction). The Syndetics ICE content increases the relevancy of your search results and helps your patrons discover more items within your collection that were previously unfound when searching traditional MARC data. Also available is Syndetics Title Link which allows you to group multiple records for the same title on your search results. Searchable Table of Contents – ICE (requires Table of Contents Display Element) Keyword search of Table of Contents reveals highly specific content often missed by traditional MARC record searches. • Database of nearly 1 million Table of Contents (Bowker and Blackwell combined) • 1,600 new Table of Contents added weekly – 80,000 annually • Exclusive content Searchable Summary & Annotations – ICE Page 50 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 (requires Summary & Annotations Display Element) Keyword search of annotation contents quickly locates descriptive information found on book jackets, in content summaries or from publisher-provided copy. • Database of 2 million annotations • 150,000 new annotations added annually • Exclusive content Searchable Fiction Profiles – ICE (requires Fiction & Biography Profiles Display Element) Keyword search of Fiction Profiles content finds books with practically any search term. • Provides extra title descriptors, such as genre and sub-genre, topic, main characters and character traits, geographic and nongeographic settings, time periods, series/sequel information • Includes fiction, biographies, autobiographies and memoirs • Fuels award-winning Fiction Connection • Exclusive content developed by specialized editors Searchable Customized Fiction Connection – ICE (Requires BIP subscription) Delivers readers' advisory functionality customized with your holdings data. • Fiction Connection suggests titles based on readers’ interests or based on a book title they already enjoy • A customized version will provide readers' advisory using the titles in your collection • A patron never has to leave your OPAC Title Link – ICE Groups multiple ISBNs for the same title into one record on your search results page. • Unclutters the page and makes results more understandable • Links variant ISBNs for the same title, subtitle and author (hardcover, softcover, audio, e-book, large print, multiple editions) Title link is a simple algorithm that identifies items in the Books In Print database that have the exact same title, subtitle and author. It uses the entire contents of these fields when linking them up (not just the first x Page 51 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 number of characters in the title, author, subtitle fields). There are about 1 million titles linked in the BIP database and these 1 million titles are linked to about 3.6 million ISBNs. In the “Moby Dick” example, the first one is grouped, the second one is not because it has a subtitle that the first one does not have. The third record does not have the same author (videos rarely do have the same as the book), the fourth and fifth title do look as if they should be linked since they share the same title, subtitle and author – my suspicion is that we do not have these records in the BIP database (I say that because our ebook coverage is not as strong as we would like. If the record is not in the BIP database, Title Link does not put them together in the results list.) Page 52 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 Training The following staff training is offered remotely. Understanding AquaBrowser is a one hour webex session for key public services and technical services staff answers questions about how AquaBrowser parameters operate with the Library's MARC bibliographic and holdings records. AquaBrowser Operations and Control Panel is an one-hour webex session for IT Help Desk and Systems staff explains configuration options that can be managed locally, and configuration options that TLC will manage for the Library. We’ll discuss standard operating procedures, system diagnostics, and common questions the Help Desk might hear from Library staff. Table of Contents in MARC record If a library has loaded Table of Contents information into the MARC record for their own collection, would ABO display the TOCs? Yes! The TOCs from 505 field will display. (The MARC 505 field is the default place for TOCs, aka 'formatted content note'.) Training – additional Medialab provides services to import new data sources. In some cases library staff can be trained to configure import processes. This requires additional training charges. AquaBrowser is a very open environment that comes with the source code for the website and import processes. The library is allowed to change import processes as long as the usage stays within the purchased options. The library can implement for example scripting to tune the calculations of values in certain facets. Page 53 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 Real life examples of customization done by customers: - At Indianapolis (sherloc.imcpl.org) the library staff implemented the “make request” functionality themselves. They used a combination of javascript and css on the client side and perl scripting with ODBC request on the server side. - At University of Chicago the library staff implemented the full record display by themselves. This required advanced xslt and css. Training for this has to be carried out on-site. Either a Medialab developer comes to the library or Library staff come to the Medialab offices. Charges for a Medialab developer on site is $2500.- first day, $1000.- every consecutive day, minimum 3 days plus travel expenses. Updates & Patches Regular AquaBrowser updates are implemented remotely via Remote Desktop (or VPN) by Medialab. This service is included in your annual support and maintenance contract. Typically updates with new features are available several times a year, and libraries can choose to upgrade as often as they desire to take advantage of the new features. W3C’s Web Content Accessibility Guidelines 1.0 http://www.w3.org/TR/WAI-WEBCONTENT Aqua Browser: Level of compliance with the W3C WAI’s WCAG 1.0 is complete for all checkpoints 1.1 through 14.3, with the exception of some of the priority 3 points. WebCrawler Web Crawler – Can you explain this more? We know “AquaBrowser Library can provide integrated searching of Library specific websites.” But what does this mean? How and why would we want this? The Web Crawler is a way for a Library to provide users a “best of the web” as determined by professional librarians. Look at it like a good reference librarians’ bookmarks, all in one place and all accessible along with other resources. This example, albeit from a public Library, demonstrates the power of the web crawler: http://www.toledoLibrary.org/. For example, search for “shipwrecks” and scroll to the bottom on the refines on the right hand side, and you’ll see… Web Sites • • Catalog (1036) Librarians' Internet Index (30) Page 54 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 • • • • Internet Public Library (13) Discover Ohio (2) Toledo PL website (2) LOC Virtual Reference Shelf (1) The Librarians’ Internet Index and Internet Public Library results are a web crawl of sites, juried by professional librarians to be the best of the web. Essentially, we are applying good reference librarianship to the Google problem of authoritative websites vs. those that result first due to advertising. This “crawl” is designed and determined by you and your staff and can continually be added to. AquaBrowser Library can provide integrated searching of library-specified websites. The websites are indexed and updated nightly. Word Cloud Can we switch off tag clouds and just have faceted searching or have it appear as an option rather than a default? Yes, the user may turn the word cloud off or on. Please see: http://lens.lib.uchicago.edu/, where the cloud can be toggled on or off, as indicated by the arrow in the red circle in this screen shot: How would our faceted subject thesaurus terms be displayed in the Word cloud? Can we see an example of how they would look? Subject thesaurus can be searched and are also displayed in the word cloud and in other presentations to the user. See: http://know2.Library.phila.gov/ for a public example and http://lens.lib.uchicago.edu/ for an academic example. In the Discover Word cloud tool, how is the relevancy algorithm calculated? The word cloud displays suggestions that are coming from different sources. The most notable source is associations that are automatically constructed based on analysis of the indexed data. The sources are described below: Associations Associations are related words that are found by means of a co-occurrence Page 55 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 analysis. All texts in the database are analyzed and the numbers of times words occur near each other in are calculated. The closer two words occur together and the more often, the higher the link between these words. In the process consideration is given to the type of word (noun, verb) as well. The original algorithm of our co-occurrence algorithm is described in: Veling and Van der Weerd (1999), Conceptual grouping in word co-occurrence networks, in: Proceedings of the IJCAI ’99, Pages 694-699, Volume 2, Morgan Kaufmann Publishers, San Francisco, USA. The algorithm has been improved over the years by adding support for handling different word types (verbs vs. nouns vs. adverbs) and knowledge of different languages. The main parameters of the algorithm can be configured via an XML file to which Library staff has access. These are the following parameters: • • • • • • Wordclassrelevance. Per word type the relevance of such a word can be indicated. For example, nouns have a relevance of 1.0 and verbs of 0.5. Minwordcount. Minimum number of times the word should occur to become a candidate for inclusion in the associations network. Minwordlength. Minimum number of characters a word should have to become a candidate for inclusion in the association’s network. Minlinkcount. Minimum number of times a link between two words should occur. Windowsize. The window (=number of words) with which the text should be analyzed. Lexiconorder. Order in which languages should be treated. Different forms of the same word are mapped to each other. If the lexicon order is English, Dutch; we first try to map a word to an English base and if that fails to a Dutch base. Blacklist. Words that are not allowed to show up in the association network. Whitelist. Does not care about the rules specified, but try to get these words into the association’s network. Fields. Specifies the fields to consider for the analysis. Usually we use title (245$a), summary (520$a) and topical subjects (650$a). • • • Spelling variants Spelling variants are words that are spelled very similarly to the query and do occur in the catalog. Synonyms or translations Synonyms or translations that are found in the built-in dictionaries. Only those words that occur in the catalog are shown to the user, not bothering them with options that lead to a dead end. In this example the synonym quicksilver is shown. AquaBrowser Library makes use of the Ergane multilingual dictionaries. Any other dictionary can be connected as well. Page 56 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 Authority references If you choose to have your authority records loaded into AquaBrowser, see also references are shown as suggestions as well and could be made complete transparent to the user. Note here in the University of Chicago, the search for “tombstones” simply adds the preferred term to the results for the user, eliminating the need for the user to follow a long trail of “see” and “see also” references. Alternatively, the Library may choose a more traditional approach in the AquaBrowser user interface to reveal and leverage the full authority file. Z39.50 Can ABL can search other Z39.50 compliant libraries catalogs instead using the Z39.50 gateway? AquaBrowser supports connecting to z39.50 for availability only; not for searching other libraries. The standard that is replacing Z39.50, SRU/SRW, is supported in the federated search module. To search other library catalogs, add them as additional databases to ABL. ABL indexes other libraries' catalogs which become 'additional MARC datasets' on the quote. Miscellaneous Is LDAP or Active Directory used so students only login once at the beginning of their search session? We will work with the Library to integrate with your authentication protocols. Experience has proven that this needs some special attention to make sure there is a fluid login process and seamless integration. Is there a limit on concurrent users? No. The End User License agreement defines the terms of use and the customer (Library) is authorized to use one AquaBrowser -deployed web site containing Library data for the public use. Page 57 of 58 AquaBrowser Library FAQ Last Updated: 5/15/2008 Is there an automated script/process? Yes, there is an automated script process. Does any other Library have an automated script/ process for updating and adding records? Yes. Page 58 of 58