Part A Administrative and basic information - ePractice.eu by yaoyufang

VIEWS: 3 PAGES: 19

									OASIS
eEUROPE AWARDS
Submission Form
Registration number eGov 2003228




Auteur:             Luc Gathy
Date de création:   24 février, 2003
Dernière modif.:    28 février, 2003
Réf.:               -
Version:            2.1 .0




Approbation:
                               Index


                               Part A : Administrative and basic information ................................................................... 1

                               Part B - Background (B.1) ....................................................................................................... 3

                               Specific objectives (B.2) .......................................................................................................... 4

                               Resources (B.3) ........................................................................................................................ 5
                                     Human resources ............................................................................................................. 5
                                     Technical ressources ........................................................................................................ 5
                                     Financial resources .......................................................................................................... 6
                               Implementation (B.4) .............................................................................................................. 8
                                     Models of fraud-process ................................................................................................. 8
                                     Knowledge management ................................................................................................ 9
                                     Quality control and correction system .......................................................................... 9
                                     Respect of the law on privacy ...................................................................................... 10
                                     Collecting data from various institutions ................................................................... 11
                               Results (B.5) ........................................................................................................................... 12
                                     Authorization of the “Comité de Surveillance” ......................................................... 12
                                     Implementation planning ............................................................................................. 12
                                     Functionalities accessible to the inspection services ................................................. 12
                                     Results (partial) reported by the inspectors and controllers ................................... 13
                                     Available statistics ......................................................................................................... 13
                                     Future projects................................................................................................................ 13
                               Learning points and conclusions (B.6) ............................................................................... 15
                                     From the organization point of view ........................................................................... 15
                                     From the control methodology point of view ............................................................ 15
                                     Conclusion ...................................................................................................................... 16
                               References and links (B.7) .................................................................................................... 17




OASIS eEurope Awards (Submission form)                                                                                                                             ii
Part A : Administrative and basic information

                               a) At wichwhich theme the submission is directed
                               European, central and local Government eCooperation


                               b) Name the case and provide an abstract
                               Anti-fraud Datawarehouse for the Belgian Social Security system - OASIS


                               « A datawarehouse OASIS was created within the framework of a joint project anti-
                               fraud of the inspection services of the Federal Public Service Social Security, the
                               National Office of Social Security, the Federal Public Service Employment, Labour
                               and Social Dialogue and the National Office of Employment.

                               The goal of this datawarehouse is to make it possible for the above mentioned
                               services of inspection on one hand to carry out the controls targeted on the basis of
                               indicators of potential fraud and on the other hand to carry out analyses on relational
                               data coming from various sectors of social security and the services of finances. »

                               c) Public eServices involved
                               Public eServices for citizens:

                                   Not applicable to the citizens

                               Public eServices for business:

                                   Other services for business (please describe)

                                   Data sharing and information interchange for improvement of decision-making process
                                   and risk control by Belgian social inspection services


                               d) Types and levels of government organizations involved
                               Ministry


                               e) Types of integration, cooperation and partnerships involved
                               Other: Inter-departmental at the federal level


                               f) Web address of service or case
                               OASIS is not accessible via Internet. It offers a protected front-office and the
                               integration of several back-offices.

                               http://www.bcss.fgov.be/fr/fluxdonnees/fluxdonnees%5F26.htm


                               g) h) Contact person for the submission and the responsible person

                               Jean-Claude HEIRMAN                              Luc Gathy
                               Head of the Inspection Service                   IT project manager
                               Bureau C905                                      Bureau C803
                               3C Rue de la Vierge Noire                        3C Rue de la Vierge Noire
                               B-1000 Bruxelles                                 B-1000 Bruxelles
                               Belgium                                          Belgium
                               Tel : +32-2-509.89.11                            Tel : +32-2-509.80.37

OASIS eEurope Awards (Submission form)                                                                          1
                               Fax : +32-2-509.89.12
                               jeanclaude.heirman@minsoc.fed.be                    luc.gathy@minsoc.fed.be

                               i) In case of selection to attend the conferene, please indicate if you prefer
                               Both exhibit and make a presentation in a Conference session.

                               We will have your exhibition material on the stand of the Crossroads Bank for the
                               Social Security.

                               The proposed demonstration is the following;

                                        Screens at the disposal of participants and visitors: Live simulation of data
                                         exchanges and utilization of the platform for detection of the frauds.

                                        Visuals: A visual to present stage by stage the business process re-
                                         engineering within and between the partners of this project.

                                        Documents at the disposal of participants and visitors: Brochures
                                         “Organization Anti-fraud of the Services of Social Inspection”, “Overview of
                                         the project OASIS (technical point of view)”

                                        Contact and interaction: Possibility to meet the conceptors of the applications
                                         and to organize debates on fraud detection.


                               j) Please indicate whether your submission may be available in a publically-
                               accessible database on good practice in eGovernment
                               “You will be contacted again to finally approve the database entry before going live”
                               checked.




OASIS eEurope Awards (Submission form)                                                                              2
Part B - Background (B.1)
                               The four services of inspection of the Federal Public Services Social security and
                               Employment, of the ONSS and the ONEM are committed to enforce the respect of the
                               legislation related to the social security of the workers.

                               Within this framework, they wished to obtain new data-processing tools to fight in a
                               systematic and structural way the principal phenomena of social fraud, in particular
                               the phenomenon of the labour broker, the false free-lancer and alternative forms of
                               remuneration eluding the social security contributions (and tax).

                               Conscious of the rich amount of data currently held by various administrations, the
                               OASIS (Organization Anti-fraud of the Services of Social Inspection) project was born
                               to build a platform of detection of the frauds and to make it possible to identify,
                               collect, clean, combine, organize and exploit data in order to be able to discover
                               convincing tracks of frauds.

                               It was naturally implemented in the form of a data warehouse connected to the
                               network BCSS (Crossroads Bank) of the Belgian Social Security.

                               The goal is to have a more efficient determination of the priorities in the objectives
                               and the immediate operational actions, classified according to an estimation of the
                               amount of the discounted regularizations.




OASIS eEurope Awards (Submission form)                                                                         3
Specific objectives (B.2)

                               Expectations of the users were formulated as follows:

                                        Analysis and research work on the basis of consolidated data, allowing
                                         research preliminary to control on the field;
                                        Short reaction time (data are recent);
                                        Pre-defined and ad hoc reports (a person per region will be trained for this
                                         purpose);
                                        Enrichment of the indicators by getting information from various separated
                                         sources;
                                        A transfer of knowledge between the original suppliers of data and the
                                         inspection services (they can be different in many cases).These suppliers hold
                                         the identification of the treated case.

                               The measuring criteria of the success of OASIS were as follows:
                                        Reduction of the search time to identify the interesting cases;
                                        To enrich the base of already known cases;
                                        Reduction of the reaction time (particularly to arrive before the bankruptcy);
                                        Generation of new cases;
                                        Increase the amounts of regularizations (weak rate when resulting from self-
                                         promoted investigations);
                                        Reduction in the processing time of a case;
                                        Reduction of the number of visits necessary to conclude the investigation;
                                        More precise orientation of the investigations (the result reached corresponds
                                         to the nature of the investigation request).

                               Finally the data-processing platform asked for the following functionalities:

                                        Management of the whole data supply chain arising from various sources:
                                         ONSS, taxation, LATG, Dimona, the employers’ and the workers’ Directory...
                                         This management includes the detailed follow-up of the reception, validation,
                                         correction and the quality control of the data;
                                        The data storage in a historized and organized environment (data
                                         warehouse) allowing the necessary exploitations;
                                        Analysis of the data with powerful tools having advanced statistical functions
                                         and datamining features. This is to meet the undefined specific and the
                                         frequent recurring needs;
                                        A strict management of the data confidentiality;
                                        Assistance to the organization of the inspection services and the coordination
                                         of their activities.




OASIS eEurope Awards (Submission form)                                                                             4
Resources (B.3)
                               The project mobilized human, technical and financial resources, which will be
                               described in this section.

Human resources

                               The partnership established for the success of the project is composed of the following
                               administrations:

                               The project leader is the Federal Public Service Social Security . It ensures
                               coordination and provides knowledge via its social inspectors. They take part in the
                               analysis of the needs and in the various activities resulting from this. The SPF Social
                               Security ensures the operational management of the data warehouse at 2nd level and
                               is associated to firm CSC to obtain the necessary knowledge in the fields of
                               datawarehousing and datamining.

                               The ONSS (National Office of Social Security) is the main supplier of data and
                               provides also for knowledge in the field via its inspectors. They take part in the
                               analysis of the needs and the activities resulting from this.

                               The ONEM (National Office of employment) siege at the steering committee and
                               provides for the basic data.

                               The BCSS (Crossroads Bank of the Social Security) is involved in the control of the
                               data flows between the administrations and on the safety aspects of the system
                               (including the application of the law on protection of the private life).

                               SMALS-MvM, with the authorization of the ONSS, provides some data which are
                               loaded in the data warehouse. It takes part in the improvement of the data quality. It
                               manages the production system, at the 1st level.

                               For the sake of completeness, it still should be mentioned that the SPF Employment,
                               Labour and Social Dialogue recently joined us and the SPF Finances is a data
                               supplier.

Technical ressources

                               Technical environment
                               It is made of the following elements:

                                        The development server, a SUN E250 running Solaris using an Oracle 8i
                                         database system. It is installed in the computer room of the SPF Social
                                         Security.

                                        The production server is a SUN E3500 running Solaris and using Oracle 8i. It
                                         is installed in one of the two computer sites of SMALS-MvM. It has a separate
                                         storage unit of A3500 type - 360 GB and a L1000 type archiving unit (or off-
                                         line storage of the data). It is inter-connected via buffers with the IBM
                                         mainframe of the Crossroads Bank of the Social Security with which it
                                         exchanges data under format In-House (EDI) and XML.

                                        The personal data forwarded via the BCSS and are sent in the form of
                                         encrypted structured messages (cfr. Chapter about implementation for the
                                         detail). The sources, from which we receive the mutations (term used for



OASIS eEurope Awards (Submission form)                                                                          5
                                          modifications), come from a Siemens BS2000 mainframe at SMALS-MvM and
                                          a Unisys mainframe running OS2200 at ONEM.

                                         “Enterprise-level” Sources are transferred via ftp (File Transfer) or tape
                                          which are read on the Octopus system of SMALS-MvM.

                                         The users (inspectors and controllers) work on four different sites and query
                                          the data warehouse via private lines through their Internet browser (3-tier
                                          architecture).


                                                                 ONEM
                                                                 Mainframe Unisys
                                       Staging ONSS
                                       Sun




                                                             E                                 E
                                         Mutations
                                         LATG
                                                             N            BCSS                 N      Staging
                                                             C                                 C
                                                             R
                                                             Y            Mainframe
                                                                                               R      OASIS
                                                                                               Y      Sun
                                                             P            IBM                  P
                                                             T                                 T
                                          DIMONA
                                          Stratus



                                                                                    Repertory of
                                                                                    employers Tapes     Taxes Tapes




                                         Connections are made safe by routers and firewalls.


                               Development environment
                               The procedures for loading, transformation, duplication, aggregation as well as the
                               integration of the metadata are developed with a ETL product : DataStage XE from
                               Ascential Software. Its graphic interfaces allow a fast development and an easy
                               maintenance. A layer known as "Presentation" offers to the users the following
                               functionalities:

                                         Managed query

                                         Semantic layer (or work-oriented) isolating the user from the physical data
                                          tables

                                         Transparent integration with the centralized metadata.

                               Business Objects is used as a reporting and ad hoc query tool to support the day-by-
                               day activities of the social inspectors. These reporting functionalities are available in
                               client-server mode (on LANs) or by means of a web browser reaching the
                               webintelligence engine of Objects Business.

Financial resources

                               The budgets (investment) since the starting of the project at the beginning of year
                               2000 are distributed as follows:




OASIS eEurope Awards (Submission form)                                                                                6
 Phase                                   Description                                            Budget

 Phase 1 .1                              Installation and configuration of the data warehouse       880.182 Euros
 « Copy management »                     platform                                                 (35.506.437 BEF)
 Phase 1.2                               Establishment of the bridges with the suppliers of         129.897 Euros
 « Connections »                         data                                                      (5.240.010 BEF)
 Phase 1.3                               Adaptation of the production systems of SMALS-              10.227 Euros
 « Adaptation on the                     MvM                                                        (412.528 BEF)
 operational servers »
 Phase 2                                 Development of all the structured messages                669.581 Euros
 « Labour broker (construction           Development of all the jobs which handle the 8           (27.010.830 BEF)
 sector) »                               sources
                                         Development of 24 alarms at the “presentation” layer
 Phase 3                                 Update of all the software used                           257.017 Euros
 « Integration DMFA and                  Preparation of the disappearance of one of the data      (10.368.040 BEF)
 technical upgrades »                    sources (LATG), to be replaced by a new declaration
                                         (DMFA) without regression
 Total:                                                                                            1.946.904 Euros
                                                                                                  ( 78.537.912BEF)



                               Annual operating costs are distributed as follows:

 Phase                                   Description                                            Budget

 Human resources                         Technical experts                                          133.880 Euros
                                                                                                   (5.400.705 BEF)
 Exploitation                            Day-by-day operating costs and the associated              144.152 Euros
                                         services                                                  (5.815.077 BEF)
 Support                                 Costs for maintenance and support on the servers            67.097 Euros
                                         and software used                                         (2.706.686 BEF)
 Total :                                                                                            345.129 Euros
                                                                                                  (13.922.470 BEF)




OASIS eEurope Awards (Submission form)                                                                          7
Implementation (B.4)
Models of fraud-process

                               The fraud detection is broken up into three levels:

                                           By sector, in order to emphasize the specificity of each sector - within the
                                            framework of the first project, this is restricted to the house-building sector;

                                           By scenario. A scenario describes the way in which a fraud is set up. It
                                            describes which are the actors and which are the possible attributes that can
                                            indicate their existence. The scenarios allow the development team to
                                            capture a structured knowledge from the inspectors.

                                           By alarm. An alarm is always associated with a scenario and each scenario "is
                                            tracked" by several alarms. An alarm provides for an indication to the
                                            detection of a scenario. Each alarm partially contributes to emphasize one of
                                            the aspects of a scenario. The combination of several alarms makes it possible
                                            to have an idea concerning the probability of a fraud.

                               The following figure illustrates the approach:




                                         Employers                  Workers

                                         attributes                 attributes                                   Inspecteurs

                                                                                                                                interviews

                                         Time           Alarms      Works done

                                         attributes                 attributes
                                                                                                                                                  Scénario
                                                                                                           Alarmes
                                                                                                                                                 Scenario
                                                                                                                                                     Les ouvriers engagés


                                                      Calc Alarms
                                                                                 Development
                                                                                                          Alarms                                     un jour par les
                                                                                                                                                     négriers de la
                                                                                                                                                 Les ouvriers engagés
                                                                                                                                                     construction ont
                                                                                                    Changement de raison sociale d’une société   un jour par les rester dans
                                                                                                                                                     tendance à
                                                                                                    existante (activité, commission paritaire)   négriers de la ...
                                                                                                                                                     ce circuit,
                                         Employers                  Workers                         vers le secteur de la construction (n° TVA   construction ont
                                                                                               Changement de raison sociale d’une société
                                                                                                    reste inchangé)                              tendance à rester dans
                                         attributes                 attributes                 existante (activité, commission paritaire)        ce circuit, ...
                                                                                               vers le secteur de la construction (n° TVA
                                                                                               reste inchangé)



                                         Time           Mesures     Works done

                                         attributes                 attributes




                                Presentation Server




                               For each scenario, a set of alarms is described. The scenarios and alarms are the result
                               of the interviews of the users and must be validated by them.

                               The development team programs these alarms at the "Presentation server" level : they
                               use the data prepared for the analysis and provide for complementary tables on the
                               results of calculations.

                               The data resulting from the analysis as well as the alarms are made available for the
                               inspectors via Business Objects.




OASIS eEurope Awards (Submission form)                                                                                                                              8
Knowledge management

                               On the data
                               The data warehouse OASIS contains data coming from the following databases:

                                        The Employers’ Directory (ONSS),
                                        The database LATG related to the career and salary of employees (ONSS),
                                        The database DIMONA for the immediate declaration at the time of
                                         recruitment (ONSS),
                                        "Building sites and subcontractors" (ONSS),
                                        "Status of the enterprises, regularity of payments" (ONSS),
                                        "Accounts" and balance of contributions (ONSS),
                                        "Recovering" and legal procedures (ONSS),
                                        Temporary/economic unemployment (ONEm),
                                        Tax declarations, debts and customer-suppliers listings (Finances).

                               Each source mentioned above brings a part of the information necessary to calculate
                               the alarms. It should be mentioned that a zone can be selected for three different
                               reasons:

                               The data offers technical information: logical keys, dates, etc, allowing to sort the
                               records and to manage the history;

                               The data is part of the calculation algorithm of an alarm;

                               The data provides contextual information: An alarm must be interpreted in a broader
                               context which makes it possible for the inspectors to appreciate its value as a clue that
                               there may be a fraud.


                               On the metadata
                               The users have access to following information:

                                        Loading date and volume of the various sources. Reports give information making
                                         it possible to the inspectors to know "at which stage of loading we are" for the
                                         various sources. In the case of the LATG for example, we give the number of
                                         records loaded as well as the number of employers loaded during the n last
                                         quarters.

                                        Definition of the analysis dimensions. In the analysis universe, a description of
                                         the significance of the various objects and their attributes is available.

                                        Identification of the sources for the analysis dimensions. A Business Objects report
                                         gives for each attribute, the sources this information comes from.

                               This very important when it is known that OASIS gives access to data managed by
                               other institutions than the one to which the inspector who launched the interrogation
                               belongs.

Quality control and correction system

                               The role of the correction and quality control system is to ensure an optimum quality
                               of data at the analysis-level.


OASIS eEurope Awards (Submission form)                                                                                 9
                               Our data warehouse is separated into two components:

                                        The "Staging area", which receives the bulk data and carries out a quality
                                         control, calculates a certain amount of new variables, allows corrections, etc.
                                         before the data are transferred to the analysis component : the "Presentation
                                         server".

                                        The "Presentation server" where the data are supposed to be of a satisfactory
                                         quality and where no more change is brought. This is the level where the
                                         analysis and reporting tools are connected and put in the hands of the users.

                               We distinguish two types of controls:

                                        Technical controls: field type, correct domain definition, key uniqueness, etc.

                                        Contents controls: field contents validation: for example, is code NACE good,
                                         does the amount of the salary seem correct, etc.

                               The figure below shows the architecture implemented for control and corrections.




                                                                                        Quality reports


                                                                           Inspectors



                                                                         Presentation

                                                            Immediate transfer

                                          Controled
                                          data
                                                                                          Coherence tests



                                          Controls
                                                                                           Technical tests
                                                                                                                Application Manager
                                                                            Staging                results


                                                                                   Sources          Feed-back to Data Owners




Respect of the law on privacy

                               The data are provided to the data warehouse through flows forwarded by the
                               Crossroads Bank.

                               The prime objective of the data warehouse is to detect and analyze possible fraud
                               scenarios . This does not require individualized personal data and this is why the law
                               on privacy requests that the personal data be provided to the datawarehouse in a
                               coded form.

                               Whenever there is a strong presumption of fraud for an employer, the social
                               inspectors from the various inspection services (SPF Social Security/ONSS/SPF
                               Employment/ONEm) can obtain access to the full data concerning the workers of this


OASIS eEurope Awards (Submission form)                                                                                         10
                               company according to a targeted examination in this company. For this specific
                               mission, we decipher the coded NISS (national identification number) to find the
                               identity of the persons.

Collecting data from various institutions

                               The gathering of information is carried out by means of standardized data flows: via
                               the Crossroads Bank, the OASIS data warehouse sends, for example, a request for
                               information about a determined employer (using the ONSS immatriculation
                               number).


                               The ONSS seeks in the databases mentioned above the data available about this
                               employer and his workers. Then, the ONSS sends to the Crossroads Bank the data for
                               which an authorization was granted.


                               The Crossroads Bank codes the NISS of the workers and transmits the information to
                               the data warehouse .


                               From this information the data warehouse is able to address a complementary
                               request, for example, to ONEm with regard to the coded NISS.


                               The Crossroads Bank decodes the coded NISS from the request and transmits the
                               request to ONEm, which communicates to the Crossroads Bank all the information
                               over the unemployed persons, the nature and the duration of his/her unemployment,
                               as well as the contract type of (part-time of full-time).


                               The Crossroads Bank then codes again the NISS of the involved workers and
                               transmits these data to the data warehouse OASIS.


                               For the personal data protected by the laws on private life, five messages have been
                               developed and are specific to OASIS, these are:


 Message       Description                                                         Submitter            Target

 L830          Data flow allowing the consultation every 14 days of the              OASIS             ONSS
               DIMONA database at the ONSS by the data warehouse                                     (DIMONA)
               OASIS with regard to the sectors defined in the
               authorization.
 L831          Data flow allowing the consultation every 14 days of the              OASIS              ONSS
               LATG database at the ONSS by the OASIS data warehouse                                   (LATG)
               with regard to the sectors defined in the authorization.
 L840          Data flow allowing the monthly consultation of the database           OASIS              ONEM
               of ONEm by the datawarehouse OASIS with regard to the
               workers in the sectors defined in the authorization.
 L835          Deciphering allowing the social inspectors to decipher coded          OASIS              BCSS
               NISS. This makes it possible for the social inspectors to
               examine in detail if a scenario of potential fraud were
               detected.
 M700          Flow for the communication of changes in a NISS number.                BCSS              OASIS




OASIS eEurope Awards (Submission form)                                                                       11
Results (B.5)
Authorization of the “Comité de Surveillance”

                               OASIS owes mainly its existence to an authorization of the “Comité de Surveillance”
                               supervising the Crossroads Bank its deliberation n° 01/06 of March 6, 2001 the
                               “Comité” authorizes the transfer of data in personal matter towards a data
                               warehouse in an anonymized form.

                               The exact heading of this authorization is Deliberation n° 01/06 of March 6, 2001 relative
                               to a request of the Federal Ministry of the Social Affairs, Public health and Environment for
                               the creation and the management of a bank of data OASIS for the fight against the social fraud
                               in the sectors of construction, the steel construction, electricity and the parks and gardens"

                               The system of anonymization implemented for the OASIS project was since then used
                               on other projects and uses the encoding and hashing processors of the BCSS. It opens
                               the way for a whole series of new applications.

                               Within the framework of OASIS, it should be mentioned that the encoding process is
                               reversible to allow the inspectors and controllers to find the data about the workers in
                               the case of suspicion of frauds.




Implementation planning

                               From April 2000 to June 2001: Realization of the three sub-projects of phase 1, namely
                               the installation and the configuration of the datawarehouse platform (project "Copy
                               management"), definition of all the structured messages and establishment of the
                               bridges with the data suppliers including the encoding systems (project
                               "Connections") and then adaptation of the production systems of SMALS-MvM
                               (project "Adaptation on the operational servers").

                               From July 2001 to July 2002: Development of the first application project for the
                               house-building sector and delivery of a first series of alarms in August 2002.

                               From August 2002 to November 2002: Increase in the number of alarms available and
                               delivery of a second series of predefined reports.

                               Phase 3 finishes in June 2003 and mainly aims at integrating the new multifunctional
                               declaration or DMFA of the ONSS (which completely replaces the source LATG). This
                               project will ensure the continuity of operation of the data warehouse without
                               regression of service.



Functionalities accessible to the inspection services

                               OASIS receives information from 9 different data sources and produces 23 reports
                               called alarms. Presentation Universes (6 in total) are also available for the inspectors
                               wanting to produce their own reports.

                               Global review reports (5) are also available to give:

                                            On the basis of one quarter and one year, a view of the various
                                             employers having generated one or more alarms over this period (report
                                             "All Alarms per Employer")


OASIS eEurope Awards (Submission form)                                                                                 12
                                            Similar to the "detail" part of the preceding report, except for that it
                                             directly makes it possible to obtain the results for an employer knowing
                                             his immatriculation number (report "Alarms for 1 Employer")

                                            On the basis of one quarter and one year, a view of which alarms are
                                             more frequently fired for which employers (report "All Employers per
                                             Alarm")

                                            Between two dates, the total number of alarms (and their types) which
                                             appeared (report "Evolution All Alarms")

                                            Gives a total number of alarms which appeared for a particular company
                                             (report "Evolution Alarms for 1 Employer")

                                    The definition of alarms cannot be disclosed publicly.




Results (partial) reported by the inspectors and controllers

                               Tests related to the effectiveness of OASIS for the detection of the fraud were carried
                               out.

                               On a sample of known cases from anti-fraud cells and/or, if necessary, consecutive
                               controls at the building sites, the tests carried out appeared all "positive".

                               There was no company for which no alarm fired. Moreover, the results were
                               extremely interesting regarding mainly with the completeness of the various alarms
                               having been activated by OASIS.

                               All the "parameters", without any exception, defined during various workshops
                               appeared to be relevant.




Available statistics

                               The only available statistics are related to flows feeding OASIS. Here are the key
                               figures. In 2002 (half-year of exploitation), the number of question raised by OASIS to
                               the BCSS was distributed as follows:

                                        Flow L830 - Consult. database DIMONA: 420.111 interrogations

                                        Flow L831 - Consult. database LATG: 713.645 interrogations

                                        Flow L840 - Consult. database ONEm : 4.307.710 interrogations

                               Phase 4 of this project will have to extend the reporting system to have factual
                               information on the ROI from the cases detected via OASIS.

Future projects

                                The contents of phase 4 are in the definition stage at the time of the writing of this
                               document. Two BPR projects (or improvement programs ) related respectively to
                               control methodologies and case treatment are started.

                               One of the stakes is of course an optimal use of the results produced by OASIS in the
                               process of creation and of treatment of the cases (evolution of a reactive system


OASIS eEurope Awards (Submission form)                                                                          13
                               towards a pro-active system) and the improvement of the collaboration between the
                               inspection services which belong to departments on different sites.




OASIS eEurope Awards (Submission form)                                                                     14
Learning points and conclusions (B.6)
From the organization point of view

                               Adoption process
                               If one wants a project as OASIS to succeed, it matters that it is fully supported by the
                               top-management of the institutions taking part in the project. The leaders must
                               unambiguously give their downstream to it. Indeed, the datawarehousing is much
                               more than a technical project. It leads to upheavals in the culture.

                               To convince management poses particular problems. The concept of a data
                               warehouse is a particularly abstract construction, which moreover is very far away
                               from the known operational systems. This is why it is particularly important to take
                               the necessary steps to explain the advantages, the impact awaited on the processes
                               and the culture. It was extremely important to start from the request for a new
                               information system, not of an information processing system. Finally, technology is
                               only an instrument to help to satisfy the data-processing needs.


                               Project phases
                               The project phase has a double finality. First of all, it is essential in a phase to develop
                               a subsystem. This implies the determination of "limited" needs and the definition of
                               the corresponding information architecture as well as the technical implementation of
                               the concept.

                               Then, the project phase aims at the training of experts in information who, at the end
                               of the project, must be capable to continue to feed and use the system. The training of
                               the potential end-users is also a part of the project phase.

                               In this phase, agents (or reference users ) selected for this purpose must be assigned
                               for a long period (counted in years) to a single project group.


                               Implementation phase
                               After the closure of the project phase, the system is put in production and produces
                               its first results. However, gaps can be noted or an additional tuning can be necessary
                               to improve the effectiveness. In the various administrations, "relays" must be created
                               for the feedback and the definition of later developments.

                               On the basis of acquired knowledge, the existing products are refined or new
                               products are developed. The decentralized information relays must be supported on
                               the organizational, conceptual and technical level by a central team.

From the control methodology point of view

                               Re-definition of the control methodology and process for the treatment of files
                               OASIS transforms data into information. The traditional process of treatment of a file
                               must now be preceded by a pre-investigation non-related to a specific case. It is the
                               preliminary phase of interrogation of OASIS in an automated way or in a directed
                               way. We thus have the new workflow:

                               1) Pre-investigation not-dependent on the cases: addition of a pro-active system
                                  automated for fraud research ; in which the data are stored and organized for
                                  research for clues


OASIS eEurope Awards (Submission form)                                                                               15
                                         A. Influence the preparation of the cases because the bonds enrich
                                            knowledge on the studied cases and makes it possible to better inform
                                            the inspectors (improvement of the reactivity)

                                         B. Have an effect on the discovery of the cases because the links established
                                            between the data make it possible to spot cases where the probability of
                                            fraud is more important (pro-active aspect)

                               2) Pre-investigation related to the files, e.a consultation of the last situation

                               3) Check and recording of the results in the files

                               4) Finalization of the administrative procedure.

                               It is important in step 1 to take care that information is distributed in the most
                               adequate way. Information, with the adequate security levels, must be strongly
                               democratised with an aim to accessibility and availability.

                               The total process (4 steps) described above must be described in a rigorous and
                               detailed way and to constitute the base of a workflow system accepted in day-to-day
                               practice by the inspectors and controllers.


                               Importance of the relays of information
                               The use of professional tools for the management of the data exposes in a pitiless way
                               the problems of the quality of the sources (missing values, bad formats, etc).

                               The bad reports created show the weaknesses in the operational systems and can
                               consequently be used in order to improve these systems systematically.

                               Moreover, it is necessary to have a team of specialists in information (in our case
                               multi-institutions), trained to the use of the most advanced technologies for the
                               production of information.

                               They also have a role to select the most interesting cases and to open requests for
                               investigations in the services concerned. In practice, we cannot work without this
                               “coordination cell”.

Conclusion

                               This practice shows that specialized tools coming from the private sector (like the
                               data warehouses for the marketing or the telecommunications sector ) are perfectly
                               appropriate to support public institutions of Social Security in the phases of
                               planification of the investigations.

                               During the implementation of such a project, the difficulties are not so much in the
                               field of the technical realization or that of the technical barriers . The big problem
                               remains fragmentation (or dispersion) of the information and the quantity of the data
                               sources.

                               Moreover, the culture is still strongly centered on the use of data in the context
                               related to an existing case. The "automated" search for frauds does not exist and is
                               thus the weak link in the chain, which demands energy and generates waste of time.

                               Lastly, the problem is not so much to build a data warehouse, but much more to fill
                               the shelves by modelizing the knowledge of the inspectors and controllers.




OASIS eEurope Awards (Submission form)                                                                             16
References and links (B.7)
                               http://www.bcss.fgov.be/fr/fluxdonnees/fluxdonnees%5F26.htm




OASIS eEurope Awards (Submission form)                                                       17

								
To top