Question Answering from Large Document Collections

March 2000
Eric J. Breck, The MITRE Corporation
John D. Burger, The MITRE Corporation
David House, The MITRE Corporation
Marc Light, The MITRE Corporation
Inderjeet Mani, The MITRE Corporation
Download PDF (58.46 KB)

We present a question answering system with a hybrid design, combining techniques from knowledge representation, information retrieval, and natural language processing. This combination enables domain independence and robustness in the face of text variability, both in the question and in the raw text documents used as knowledge sources. We describe the specific design of our current prototype, which has been entered in the 1999 Text Retrieval Conference's Question Answering track. We also comment on the evaluation of question answering systems.


Publication Search