+ Home
+ ECRAN project
+ Project summary
+ Annual report
+ Project site
+ Related projects
+ RTD projects
+ LE1 projects
+ About this site
+ Search
+ Feedback


Extraction of Content: Research At Near-market


ECRAN will develop a new generation of Information Extraction (IE) applications, to be included in telematic services having a large textual content. ECRAN will analyse free texts (initially, financial information from specialised newswire services, and market information on the internet) extracting information content. The information can be compared against a model of user requirements so that the system can precisely identify text of interest to a customer.

Using state-of-the-art IE and lexical acquisition techniques, the ECRAN project will also develop a 'lexicon tuning' tool which enables the linguistic knowledge of the system to be extended and adapted to new text domains automatically. The IE technology developed has a wide range of applications, supporting for example, videotext and online services (online resources indexing, web browsers, active Email systems, intelligent TV guides).

Home - Gate - Back - Top - Ecran - Relevant