vision

FF POIROT is about two things essentially: ontology and fraud It analyses and mines the knowledge of fraud examination and models it as ontology of fraud forensics and other subject matters as domain context. The ontology is the explicit computational knowledge representation of fraud examination expertise. It is deployed in and referenced by intelligence analysis and decision support systems. The use of ontology contributes the knowledge management among fraud examiners from disciplines and training. It knowledge-enables existing information systems. It facilitates the detection and prevention of financial frauds.

Project appraisal The final evaluation by the EC awards the project an overall score of 5/5. It was evaluated on quality, effectiveness, impact and IST objectives. " The management of the project, managerial and technical, is excellent... The project has produced very promising technical results and has identified and developed applications in close interaction with end users." The "innovation of the project" is described as "breakthrough". Exhibition events: Communicating European Research 2005, November 14, 15, Brussels. KM Europe 2004, Amsterdam, Nov. 2004.					Ontology is an explicit representation of knowledge concerning given subjects and domains. It is used with consensus of its user communities for knowledge management and communication. Topical ontology pertains to one or more themes, such as value added tax, intellectual property rights, financial fraud. It represents the knowledge structure of an expert on the theme. It typically deals with multiple subject matters and thus involves multiple domain ontologies. Though it reflects particular conceptual perspectives and structures associated with the themes, it is not intended to be application specific, rather it is usable by a family of knowledge applications.

	Fraud cases analysed are Illegal solicitation of financial instruments on the web Fraudulent insurance claims VAT carousel frauds in inter-community trade Unauthorised e-commerce activities Frauds by emails	Analyse frauds, model Ontology. FF POIROT studies the fraud cases and the legal, organizational and software instruments. In order to capture and represent the expert knowledge, A practical knowledge engineering methodology is followed by the consortium. A systematic approach is needed for effective collaboration due to the nature of the task: Multi-disciplinary subject domains Multi-applications under consideration Multi-disciplinary team of development Geographically and administratively distributed team		Tools and knowledge engineering facilities are created and adapted to maximise the effectiveness of knowledge engineering in the multidisciplinary knowledge domain. The facilities support knowledge requirements specification, knowledge analysis, automatic knowledge discovery, ontology modeller and integration, knowledge engineering portal.
Knowledge models. The outputs are the description of knowledge scope, knowledge breakdown structure, logics of each structural constituents, the ontology of their underlying concepts and relations and their use in particular applications and systems. Topical ontology of frauds is intended for knowledge management and detection of frauds. Topical ontology of VAT is intended for regulation enforcement to close the loopholes of VAT frauds. The verbalisation of these ontologies are in the form of multilingual terminology of English, French, Dutch and Italian.			Deployment of knowledge in applications. Ontologies are the knowledge framework for specifying business rules and constraints. Concepts and relations are grounded in specific tasks, business rationale and system contexts. They are more constrained, concrete and exclusive conceptually. They are connected to form a complex logical network. They are either converted into standard representation or compiled into specific required formats to load into the existing information systems.
Surveillance The surveillance on the web against fraudulent or illegal practices has become difficult with solely human resources. The ontology-based web search has been developed with the Knowledge Stones' technologies for CONSOB, which regulates the Italian stock exchanges. The results are evaluated in comparison with the existing process of web surveillance. Among the identified web sites are the known fraudulent sites as well as suspicious ones. The technology proves to be complementary to the human process. While the web search is ontology-driven key words search, a deeper analysis of web site contents is prototyped at VUB, by the use of natural language understanding and information extraction technologies. The natural language texts on the site are parsed and interpreted based on semantic model in terms of ontology. It is an information harvester according to a specified conceptual model developed by experts or analysts. Detection While web information search and extraction deals with unstructured data, the ontology of fraud is modeled to perform the model-driven link analysis of structured data in SQL databases. Forensic Pathways Ltd uses the ontological model to manage its investigative expertise and deploys it on its investigative system for insurance claim analysis to identify suspicious claims. Crucial to fraud examination and prosecution is the index, search and retrieve relevant evidences. The topical ontology of fraud suggests more systematic standard ways of evidence management and possible added value of automatic evidence interpretation. Prevention An effective way of proactive fraud prevention is to enable the potential victim to stay out of traps. VAT Application NV provides information systems for invoicing. The idea is to automate and guide users, SME, in validating data, at the stage of invoice preparation, to keep them out of traps of VAT carousel frauds. It builds an ontological layer and uses the topical ontology of VAT to scale its rule engine to 25 national legislation and 20 languages. Exploitation of project results The project consortium looks forward to the future exploitation of the project results by their intellectual property rights. The exploitation may be conducted with individual partners or group or the whole consortium. The consortium wishes to build further on the technologies, knowledge, methodology and expertise developed in the project in collaboration with other professional organizations of industrial, academic or public nature. The format of collaboration and exploitation can be as follows. industrial pilot applications commercialisation of the technologies and ontologies joint development of industrial applications joint R&D projects in the national or the European Commission's programmes consultation on knowledge management and methodologies Welcome to our exhibit booth 1.107 International Conference of Communicating European Research 2005, Brussels, 14 - 15, November 2005