Issues in Extraction and Categorization for Question Answering

We describe our approach in constructing question answering systems, which involves natural language parsing enhanced with named entity extraction and part-of-speech tagging. Performance of our question categorizer is discussed, particularly with respect to misclassifications and how they can be challenging to correct. Finally, recent work in connecting question answering to image and video retrieval components is discussed.

• Type: Conference
• Year: 2003
• Where: NDQA
• Authors: David Eichmann
• Keywords: Education, NDQA 2003, Question Answering, Question Answering Systems, Question Categorizer