Fabio Ciravegna, Luca Gilardoni, Alberto Lavelli, Silvia Mazza, William J. Black, Massimo Ferraro, Nadia Mana, Johannes Matiasek, Fabio Rinaldi
This paper describes an advanced system for multilingual text classification adaptable to different user needs. The system has been initially developed as an applied research project involving both research centres, industrial bodies and end-user organisations. The project was a considerable success story in the financial field. Three different successful applications were released at the users' site involved in the project. Moreover the system was adopted by the main Italian financial news agency where it is used to provide classified news for an external pay-to-view service. The system has been running continuously since January 1998. Its architecture integrates modules based on both some innovative artificial intelligence methodologies and commercial tools. This shows that state-of-the-art AI techniques are mature enough to provide real world applications.
Keywords: Natural Language Processing, Text Classification
Citation: Fabio Ciravegna, Luca Gilardoni, Alberto Lavelli, Silvia Mazza, William J. Black, Massimo Ferraro, Nadia Mana, Johannes Matiasek, Fabio Rinaldi: Flexible Text Classification for Financial Applications: The Facile System. In W.Horn (ed.): ECAI2000, Proceedings of the 14th European Conference on Artificial Intelligence, IOS Press, Amsterdam, 2000, pp.696-700.