Better Extraction from Text Towards Enhanced Retrieval

Designing a compact cross-lingual and multi-lingual information retrieval and information extraction system to help an analyst with the user need across large-scale textual corpus.

BETTER is a 4 years project funded by IARPA to expand the informaion retreival (IR) and information extraction (IE) interest community beyond monolingual to cross-lingual and multi-lingual setting. Eugene Agichtein (Emory) lead the IR team, Benjamin Van Durme from John Hopkins leads the IE team, and Jordan Boyd-Graber leads the Human-in-the-loop (HITL) team.

• The BETTER interface was developed to make an easy to interact search interface with support for cross-lingual and query generation assistance & feedback. The details of the search Interaface, BetterSearch can be found at here.