Annotation

Appendify: Bringing Web Annotation to the Masses Ario Jafarzadeh Technical Communication Department University of Washington TC 505 – final paper - 3/10/06 ario@u.washington.edu Abstract Much has been written about web-based annotation in the academic sphere, but little of this work has actually made its way into the hands of average internet users. Several reasons exist for this, but the most glaring of these from the end-user’s perspective is the dauntingly complex nature of current web annotation solutions. This paper provides an overview of these solutions and highlights where they fall short of gaining adoption with the masses. An alternative solution is proposed for an annotation platform named, Appendify, which takes a more user-centered approach to the problem. I. Introduction Information overload and anxiety has gained much attention over the last few years, particularly with the rise in popularity of blogs, RSS technology, and the ever increasing amount of email that the average computer-based knowledge worker encounters. While this attention is well-deserved, the problem of information overload has been around since the 15th century with the advent of the printing press [17]. Only a short 50 years after this paradigm shifting invention debuted, there were 8 million books printed. From that era onwards, man has always had more information in their world than was possible to digest in one reasonable lifetime. While information overload has been with us for quite some time, the phenomenon of information anxiety as coined by Wurman [22] is unique to the computer generation since the data is no longer passive, “but actively inserting itself into our environment, our attentions.” In the print media world, a reader has to take a concrete action to come into contact with data, but in the digital world, data is “pushed” onto the user in a more passive fashion, just by sitting in front of the TV, or keeping an email or IM client running on the computer. The problem only gets worse as time goes on and the user accumulates more friends online, subscribes to more blogs, and as the data pile itself grows exponentially. A 2003 study at the University of California at Berkley states that the “world produces between 1 and 2 exabytes of unique information per year, which is roughly 250 megabytes for every man, woman, and child on earth.” [13] Of this massive pile of data, only .003% is printed documents, the rest being contained on some form of magnetic storage. This ocean of digital data in the form of html pages doubles in size each year with email growing at 500 times that rate. The curve of this growth is also highly nonlinear, as the study between 2003 and 2000 indicates that the total amount of data (both in print and digital form) doubled between their two studies. It is clear to see from these numbers that, as Odlyzko puts it, “communication, not content, is the killer app." [15] Compounding the problem are the cognitive limitations of the human brain and the disruptive environments they are forced to operate within. The average American worker is interrupted 74 times a day and according to flow theory, each of those interruptions can cost up to 15 minutes of recovery time to get back on task [8]. Studies have also shown that the average person can only handle about two or three chunks of information at a time [20]. So where does this put us? According to former Xerox PARC contributor, David Levy, this leaves us in a state where “we’re feeling that we’re not on top of our lives… we’re losing our ability to focus, as well as our sense of our physical bodies… our workload and speed [do] not leave room for thoughtful reflection” [9]. It’s clear from research and anecdotal personal experiences that a change is needed in order to restore balance to the current internet information sharing system. This change can either come from a technological solution, a human-based solution, or a combination of both. This paper explores the latter through the lens of user-friendly web annotations. II. Annotation History Annotations are short notes that explain, describe, or evaluate a particular document and they have been in use ever since the advent of writing. From the era of caveman wall paintings to the graffiti seen in modern city streets today, people have always had a need to alter and append existing documents. In the modern era, annotations provide the following benefits [16]:       provide a trace of use third party commentary information sharing information filtering semantic labeling of document content enhanced search Almost every professional and academic arena utilizes annotations in some form to facilitate one of these features. Of particular note is the field of engineering with its specification documents and the invention of the STOP format document during the 1960s [18]. These documents made annotations the center piece by taking larger, more complex documents and tailoring them down to their essential components based on audience needs. This same desire for audience customization and simplification is seen today with the popularity of blogs and wikis which demonstrates that users want to be involved in content creation (not just consumption). While these technologies have unarguably advanced the progress of the internet, they fall way short of the original visions imagined by two luminaries in the field of hypertext, Vannevar Bush and Ted Nelson. In Bush’s seminal article, “As We May Think,” he describes a technology called the “memex” which acts as an interactive information portal where users can not only access information, but also append documents through the use of “associative trails.” These trails would show other users in the system how documents relate to each other, thereby reducing the amount of duplication in research. While Bush’s vision lacked the concept of automatic search or granular hyperlinks on individual elements within a document, it was decades ahead of its time in predicting how the knowledge worker of the future would operate. Another influential figure in the field of early annotation thinking, Douglas Engelbart, said the following about the innovations that would arise out of Bush’s memex: These illustrate the types of changes in the ways in which people can cooperate intellectually that can emerge from the augmentation of the individuals. This type of change represents a very significant part of the potential value in pursuing research directly on the means for making individuals intellectually more effective. [7] Around the time of Engelbart’s paper, Nelson developed the idea for Xanadu which is credited as the first hypertext project [14]. His vision predicted and influenced the creation of technologies such as the IFRAME which allows web pages to be included inside of other web pages (aka “transclusion”) and RSS which aggregates several sources of information into a single feed. While these aspects of Xanadu have been fulfilled, one of its core tenets remains out of reach to the modern internet user, the concept of providing free access to hypertext documents for content customization and editing [12]. This concept, also referred to as “the writable web,” has eluded the World Wide Web for a myriad of technical reasons and issues around standardization. W3C director and inventor, Tim Berners-Lee, continues to work on initiatives to bring Nelson’s vision to life in future versions of the web, but for now, Xanadu is still the longest running piece of vaporware in history as its benefits stay out of the hands of users. III. Current Solutions and shortcomings While the fulfillment of Xanadu and the memex may be far off, several technologies exist today that attempt to provide some of their benefits. The sections below provide an overview of each major category in providing two-way document interaction (and what the strengths and shortcomings are for each). Standard Webpage The standard webpage is a purely one-way system for providing internet based discourse. While links to pages can be shared amongst individuals, the content on the pages remains static and uneditable by readers. For years, this has been the standard mode by which users interact with the web and its model has emphasized simplicity and copyright protection as its core benefits. Blogs Blogs have taken off over the last six years with estimates claiming there are approximately 100 million in existence, growing at a rate of 100,000 a day [21]. While these sites do not allow direct manipulation of original content (like a Wiki), they do allow readers to post comments in response to the blog author’s posts. Much like message boards, blogs enable conversations to happen around ideas and documents in a public, but highly personalized format. While blogs are certainly a powerful means for information dispersal, their technology does not allow for annotation directly on top of original documents, forcing authors to point to external sites, while keeping their commentary separate and hosted on their own blog pages. This forces the reader to have to refer back and forth between the two documents and also puts the impetus for finding the sections of interest onto the shoulders of the reader. Social Networking Sites An even larger and faster growing phenomenon is that of social networking sites (SNS), the most popular of which being MySpace. With approximately 250,000 new account creations a day [21], MySpace is a formidable arena for exploration in the field of computer mediated communication. While blogging sites like LiveJournal have had the concept of friend lists prior to the SNS explosion, the concept is not exclusive to the world of blogging. What friend lists do that SNS’s thrive off of is the creation of semi-public spaces where users feel freer to express themselves and share information, without the fear of unwanted onlookers. Through functionality that allows the “locking” of posts and privatizing of journals as “friend’s only,” SNS sites provide a layer of privacy control that blogs do not afford, but with the same annotation shortcomings of blogs. Wikis Wikis are perhaps the closest to mass market level annotation that the web has seen thus far. This technology allows users to create and modify publicly editable documents that are tailored to a particular audience. The most popular example of a wiki, Wikipedia.com, contains thousands of individual wiki pages that users from around the world update and modify for the sake of creating an up to date and living encyclopedia. While wikis do bring to life many of the original aspirations for a twodirectional writable web, the modifications that take place on a wiki are not exactly annotations in the classic sense. Users can modify a wiki page to point out a mistake or omission, but these comments do not live separately from the original, but are actually integrated into the original. To see that a change was made, a user can look into a wiki page’s history to see the change evolution, but the end product is always the latest version containing all prior revisions. This obfusticates the annotations at the most visible level, which is acceptable for the intentions of a wiki. The main downside to Wikis, however, lies in their insularity. Once outside of the wiki ecosphere all of the editing and collaboration functionality becomes unavailable to the user. Email, IM, & Link Sharing A typical link sharing exchange, whether through email, IM, or blogs, resembles something like this: From: ario To: donte Subject: cool article on robots check out this article on robots… the part you’ll really like is on page 8, 3rd sentence down talks about the kind of robots you like. http://www.coolrobots.com/article.htm This method for commenting and sharing web-based documents is by far the most commonly used. Its benefits include familiarity with existing web technologies, ease of use (copy n’ paste), and speed (hitting “send”). This method falls short much like the way blogs do with regards to annotation, but it does afford highly personalized annotation like that found in STOP documents. While the annotations aren’t provided adjacent to the original, they are highly customized to a particular audience by including specific recipients on the To: line of an email. This personalization is diluted as soon as the annotation is presented in a public forum like a blog or message board since the reader no longer knows how tailored the information is him or her. “Web 2.0” sites Sites like del.icio.us and Flickr.com pride themselves on ushering in a new era of web technologies which place emphasis on user created metadata and capitalizing on the “wisdom of crowds.” While these sites certainly have made improvements in the field of annotation, they are exclusively tailored to annotate only certain types of documents. Flickr, for example, focuses solely on annotations around images. Users can label or “tag” an image, place it in a group with similar images, or place notes directly on top of the image. Del.icio.us focuses on the tagging and archiving of URLs, allowing users only a sentence or two worth of description which is intended to describe the entire document. Both of these sites are great first steps in achieving a more Xanadu-esque type of web, but there has yet to be a more generic annotation solution to earn the label of “Web 2.0.” Microsoft Word Perhaps the most familiar and oft-used form of annotation today exists in the popular word-processing software, Microsoft Word. The application provides a “track changes” feature which allows reviewers of documents to annotate and make changes to existing Word files. The original owner is then sent a saved copy of the original containing the revisions and it is up to them whether or not they choose to accept or reject each annotation. While this system is useful, it is relegated only to documents in a proprietary .DOC format, and not to generic documents that live actively on the web. HTML Anchors An under-utilized feature of standard HTML documents is that of the HTML anchor. Anchors provide a way for attention direction by navigating the user to a subelement on a page. They’re also useful since they can be passed around via the URL itself via a target tag appended to the end of the original URL (i.e. http://www.example.com/index.html#top). This feature is also described more accurately as “granular addressability in HTML documents” [11]. While this addressability is found in many web documents, only a small percentage internet pages utilize this functionality. One service aiming to address this deficiency is Purpleslurple.net which derives its name from the concept of “purple numbers.” These were indexing marks that Engelbart described in his proposed hypertext solution [7]. From the Purpleslurple site, a user can copy and paste a URL of a page they want anchors placed on and after clicking “Submit,” get a “slurped up” version of the page in return. The image below illustrates a resulting slurped page using CNN.com: URL: http://www.purpleslurple.net/ps.php?theurl=http%3A%2F%2Fwww.cnn.com%2F The screenshot depicts hyperlinked purple numbers appearing next to discrete sections of the page. This allows a user who wishes to direct another to view a specific part of the page the ability to capture the hyperlink from the section of interest and then send someone else to the same exact spot. While this is a powerful attention directing tool, it does not allow for any other form of annotation, nor does it allow for manual placement of anchor tags. The algorithm used to autoanchor tag the page is also unable to recognize certain parts of the page which may be contained in an unfamiliar table format or IFRAME (as illustrated by the righthand column in the CNN screenshot not containing any purple numbers). General Web-based Annotation Solutions There are several annotation solutions and services available today. To examine each one is out of scope for this paper, but it is useful to look at the various categories they fall into:    Application based – requires installing special software at the OS level Browser based – requires installing a special browser or plug-in Web based – no software installation required These solutions can go one of three ways with regards to where the annotations are stored:    Locally – saved as a file on the user’s machine On a remote server – saved on a server independent of the user’s machine Or using a solution which splits the duties between both client and server – a system like the one proposed by Denoue [4] which stores annotations locally and then rebuilds the annotated version of the page on the fly when requested over the web. Regardless of the above classifications, Bieber et al describe these four characteristics as being critical to any annotation environment [19]:  editable browsers – ability to allow for page editing within the same browsing window  storing document content and link anchors separately – this allows for the original document to go untouched  external linkbases – this provides some way for the annotation to be shared with other users  displaying link spans, node and link attributes – this provides a way for the user to retain context of the original document’s location out on the web These attributes and their corresponding characteristics are more fully examined in the sections that follow. Application based These are annotation solutions that require installation of a program on the user’s machine. They afford powerful annotation tools such as rich drawing tools which aim to make the digital annotation experience more like paper. For annotation to feel natural, users expect to have similar capabilities to that found in the paper world such as the ability to take notes in the margin, underlining, highlighting, circling text, or writing between lines [10]. Application-based annotation solutions do this quite well since this form of annotation is by its very nature complex, requiring more powerful client-side software. While this approach certainly affords richer annotation, it contains two huge issues, one involving installation and the other with annotation sharing. For a user to make use of such a solution, he or she is required to install a separate software package onto the machine which for most people is a limiting option. Users who have restricted access to their machines in the office, at school, or at libraries are unable to make use of such annotation solutions. Many users are simply fearful to install programs they are unfamiliar with because of increased inclusion of spy and malware with web-delivered applications. If a user is able to pass this hurdle, the second big obstacle involves saving the annotated document out to a typically proprietary format that can only be read by the same software package. This requires readers of annotations to also have the application installed which adds to further headache for both parties. S Screenshot of application based Onfolio. Depicts content annotation tools within MS Word Browser based Browser based annotation solutions such as Amaya (http://www.w3.org/Amaya) and Annozilla (http://annozilla.mozdev.org) suffer some of the same pitfalls as application based solutions, with the advantage that they are more closely integrated with the web browsing experience. In the case of Amaya, it falls under both camps since it is an actual web browsing application. Amaya allows users to not only browse webpages, but also to share them with others within the same window. This is currently not a feature of any of the popular web browsing clients and therein lays the main issue with its approach. By creating a new browser platform that breaks some of Internet Explorer and Firefox’s standard conventions, it has lacked a following large enough to make use of its capabilities. Solutions like E-Quill and Annozilla skirt this issue by taking a plug-in (aka “extension”) approach within the browser. While Annozilla does bring annotation functionality closer to the browser experience, the barrier of installing a browser plug-in is its greatest obstacle, not to mention being both OS and browser specific. Screenshot of IE-specific annotation plug-in, “E-Quill” Web Based Unlike the aforementioned solutions, web based annotation schemes do not require the installation of any special software apart from the web browser. Instead, an intermediary website does the job of the application by providing a framework around the original document that allows for annotations. A web based solution could technically store the annotations locally, but given their web-centric nature, most of them store the annotations on a remote server. This introduces issues around performance since the system now requires the fetching of data from both the original document location and the annotation server which stores the user generated metadata. As Denoue point out, this is sub-optimal as a personal annotation solution since speed is a valuable factor with regards to ease of use [5]. In his proposed Yawas system (yet another web annotation system), he advocates the blending of both local and remotely stored annotations where metadata is kept locally all the way up until the point of sharing. Another shortcoming of web based annotation solutions is that their annotation tools are quite limited in comparison to application and plug-in based solutions. That Yawas solution for example only supports the highlighting of text for example which for most, would not be appealing enough to integrate into their daily information sharing tasks. To their credit however, web based annotation systems are perceived as the most hassle free since they can be used without any extra software installation and they also afford improved usability benefits due to the annotation tools existing at the browser-level of interaction. IV. Appendify Technology Acceptance Model When considering whether a new technology will be accepted by the majority of users in any particular area, Davis proposed an idea in information systems theory that classifies technology acceptance based on two factors [3]. These include:   Perceived usefulness (PU) - the degree to which a person believes that using a particular system will enhance his or her performance Perceived ease-of-use (EOU) - the degree to which a person believes that using a particular system would be free from effort Having considered the current web annotation landscape, I propose a system called “Appendify” which incorporates thinking from solutions proposed during the Bush/Nelson era, as well as modern day solutions. The Appendify solution aims to improve upon existing PU and EOU perceptions found in existing web annotation systems in order to increase adoption and usage of web annotation. For the purposes of this paper, I focus on the computer mediated communication benefits of the proposed system, semi-free of technical limitations and intellectual property issues. While the proposed system is not totally unaware of these issues, the intent is not to focus on those areas, but to explore the system’s CMC impact. How it works The most glaring issue to solve before annotation can gain widespread acceptance is addressing installation phobia. Since users have valid trepidations around installing software, Appendify exists solely in web form. This also frees the tool from platform compatibility issues. To use Appendify, the user navigates to http://www.appendify.com after picking the source document they wish to annotate. Upon arrival to the site, the user pastes the URL of the page they wish to annotate into a simple form and clicks submit. From there, the document loads within an IFRAME that holds an annotation tool set above it. The mockup below illustrates the overall architecture of this page. example mockup of a web document loaded within the Appendify.com site The latin text displayed in the main window represents the original document while the surrounding elements depict Appendify’s parent frame. From here, the user is able to create free standing text bubbles, highlights, and drawings all via the annotation tool set. Once the document has been annotated to the user’s liking, they hit “Save” which generates an “appendified” URL, similar to the one proposed by Denoue’s Yawas system [5], but with one key difference. Denoe’s Yawas URL: http://localhost/URL?annotation=selection&title=title_current_frame Appendify URL: http://appendify.com// Denoe’s version seeks to maintain the URL structure of the original document whereas the Appendify one allows the user to create a unique name for their appendified URL. While this does break the last of Bieber’s four key annotation tenets, the benefit of this scheme is best illustrated with a real world example. Consider the following URL from mapquest.com which contains driving directions: http://www.mapquest.com/maps/map.adp?ovi=1&mqmap.x=300&mqmap.y=75&m apdata=%252bKZmeiIh6N%252bIgpXRP3bylMaN0O4z8OOUkZWYe7NRH6ldDN96YFT IUmSH3Q6OzE5XVqcuc5zb%252fY5wy1MZwTnT2pu%252bNMjOjsHjvNlygTRMzqazP StrN%252f1YzA0oWEWLwkHdhVHeG9sG6cMrfXNJKHY6fML4o6Nb0SeQm75ET9jAjKelr mqBCNta%252bsKC9n8jslz%252fo188N4g3BvAJYuzx8J8r%252f1fPFWkPYg%252bT9 Su5KoQ9YpNSj%252bmo0h0aEK%252bofj3f6vCP In the Denoue system, this URL would be a monstrosity to handle, especially for the purpose of emailing. In the Appendify world, the user is able to create a link for the same annotated page that instead resembles something like this: http://appendify.com/ario/directions_to_joes The user would be prompted to enter this user friendly name at the time of saving the annotations. Now, the annotation creator is able to create a much cleaner URL that encapsulates both the original, along with all of the associated metadata. The URL can then be posted on a blog, sent via IM, or emailed as the user sees fit for sharing the annotated document. When a recipient views the URL, the original URL for the source document is shown in Appendify’s interface to provide the context that Bieber identifies as crucial. Manual anchoring To address granular addressability, Appendify also supports the manual creation of anchor tags within an unanchored document. A prototype of this exists at http://www.cerealrobots.com/tc/stop and is depicted in the screenshot below. Screenshot from an IQSTOP document supporting manual anchoring. Within the Appendify tool set, there exists an anchor marker which allows the user to indicate a point they would like the appendified URL recipient to automatically be taken to upon page load. CMC Implications Given the backdrop of the current information overload situation, Appendify offers a way to tailor and filter the web-based information we share with one another. The CMC implications of an Appendify-style service for annotation include the following:  Allows users a way to provide their own content in and around web-based documents in a way that does not require any special software or computer skills beyond basic web-browsing  Shifts the paradigm of links and emails being seen as an information burden, and more as carriers of useful information  Encourages conversations to happen around documents, which will better facilitate the refinement and exchange of ideas. Annotators can provide the appendified URL to whomever they want to control privacy Provides assurance to document creators that their original documents will be left unaltered   The net effect of all of this will lead to a less information overloaded web where users pre-filter and direct each other to the most relevant areas within documents. Users will no longer have to guess (particularly in larger documents) what the most important parts are or what the sender intended to spotlight. Users creating annotations will also be freed from fumbling around awkward “see paragraph 5, line 3, on page 8” types of clarifications when sending links. While there are still many large technical and legal issues to solve, in the ideal Appendify world, there is great promise for a web where informational noise becomes converted into knowledge. Annotated Bibliography 1. Bieber M, Vitali F. (1997). Toward Support for Hypermedia on the World Wide Web. IEEE Computer (0018-9162). http://www.cs.unibo.it/~fabio/bio/papers/1997/IEEEC97/January/IEEEC0197.ht ml Outlines some of the major technical issues around implementing web-based annotation solutions. 2. Bush, V. (1945). As We May Think, The Atlantic Monthly. http://www.ps.uni-sb.de/~duchier/pub/vbush/vbush.shtml Seminal paper which predicates many elements of the current web, along with some ideals which have yet to be realized. 3. Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319340. Provides a useful framework for referring to perceived usefulness and perceived ease of use which are Appendify aims to improve upon in the world of webannotation. 4. Denoue, L. (1999). Adding Metadata to improve retrieval: Yet Another Web Annotation System. (Technical Report). http://www.fxpal.com/people/denoue/publications/TR1999-01.pdf First paper to introduce the idea of annotations being encapsulated and shared via a URL. 5. Denoue, L., Vignollet, L. (2000). New ways of using Web annotations. (WWW9 poster). http://www9.org/final-posters/poster46.html Discusses the benefits of having platform independent annotation solutions. 6. Di Iorio, A & Vitali, F. (2004). Writing the Web. Journal of Digital Information, Volume 5 Issue 1 Article No. 251, 2004-05-27. http://jodi.tamu.edu/Articles/v05/i01/DiIorio/ Discusses Ted Nelson’s Xanadu Project and how it relates to web annotation. 7. Engelbart, D.C. (1962). Augmenting Human Intellect: A Conceptual Framework. Summary Report AFOSR-3223 under Contract AF 49(638)1024, SRI Project 3578 for Air Force Office of Scientific Research, Stanford Research Institute, Menlo Park, Ca. Describes how Bush’s memex can assist in addressing information overload and lead to the augmentation of human intellect. 8. Gupta, A., Sharda, R., Greve, R., Kamath, M., Chinnaswamy, M. (2005). How often should we check our email? Balancing interruptions and quick response times. Presented at Big XII IS Research Symposium, University of Oklahoma, Norman http://iris.okstate.edu/rems/pubs/ICS05.ppt Presentation which provides useful stats around distractions and email volume. 9. Hart, J. (2005). Technoskeptic Techie, Utne magazine. http://www.utne.com/pub/2005_127/promo/11499-1.html Article which covers David Levy’s thoughts on how current digital culture is harmful to society. 10. Heck, R.M., S. M. Luebke, & C. H. Obermark (1999). A Survey of Web Annotation Systems. http://www.math.grin.edu/~luebke/Research/Summer1999/survey\_paper.html Provides an overview of current application and web based annotation models. 11. Kim, E.E. (2001) An Introduction to Purple. Version 1.3. Webpage. http://www.eekim.com/software/purple/purple.html Overview of the Purpleslurp.com anchoring system which relates to some of Engelbart’s early hypertext ideas. 12. Linden, G. (2005). Organizing chaos and information overload. Website. http://glinden.blogspot.com/2005/12/organizing-chaos-and-information.html Provides an aggregation of recent commentary on the information overload problem from several prominent bloggers. 13. Lyman, P. & Varian, H.R. (2000). How Much Information? Website. http://www.sims.berkeley.edu:8000/research/projects/how-muchinfo/summary.html Useful stats to illustrate the information overload problem. 14. Nelson, T.H. (1987) Literary Machines. Mindful Press Nelson’s seminal article in which he outlines the Xanadu Project. 15. Odlyzko, A. (2000). Content is not king. Technical report, AT&T Labs. http://www.research.att.com/~amo/doc/networks.html Paper that emphasizes the need for tools to manage the glut of information available on the web. 16. Perry, P. (2001). Web Annotations. Website. http://www.paulperry.net/notes/annotations.asp Provides useful definitions around web annotations. 17. Postamn, N. (1990). Informing Ourselves to Death. German Informatics Society, Stuttgart. http://www.eff.org/Net_culture/Criticisms/informing_ourselves_to_death.paper Provides an overview the information anxiety problem in a historical context. Challenges the current “technology will fix all” mentality. 18. Starkey, W. S. (2000). The Beginnings of STOP Storyboarding and the Modular Proposal. Proposal Management (Association of Proposal Management Professionals), 41-48. http://www.apmp.org/pdf/41-48.storyboardinghistory.pdf Paper which describes the origins of the STOP document. 19. Vitali, F. & Bieber M. (1999). Hypermedia on the Web: what will it take?. ACM Computing Surveys, 31(4es). http://web.njit.edu/~bieber/pub/acmcs/cs-vb.html Identifies the core components of an effective web annotation solution. 20. Yntema, D. B. (1963). Keeping track of several things at once. Human Factors, 5, 7-17. Study that describes the limitations of short-term human memory. 21. Wright, J (2006). MySpace is the New Blogosphere. Blog post. http://www.ensight.org/archives/2006/02/21/myspace-is-the-new-blogosphere/ Recent blog posting that provides stats around blog and MySpace account creation. 22. Wurman, R.S. (2001). Information Anxiety 2. Que, 15-16.