The project aims at developing the technologies for analyzing text in the Italian language in order to build a Question Answering system based on semantic relations.
An initial prototype is available called DeepSearch.
As part of this activity, we have developed a full pipeline of NLP tools, called Tanl (Text Analytics and Natural Language) and pronounced 'tunnel', that we are applying to the Italian Wikipedia, creating a large body of text annotated with semantic tags.
Parsing Web Services are available for:
- Related Work
- Semantically Annotated English Wikipedia (was: http://www.yr-bcn.es/dokuwiki/doku.php?id=semantically_annotated_snapshot_of_wikipedia)
- Yahoo! Correlator (was: http://sandbox.yahoo.com/Correlator)
- Mining Meaning from Wikipedia