Enhancing Factoid Question Answering Using Frame Semantics: How FrameNet may Improve Answer Passage Retrieval and Answer Processing Effectiveness in Natural Language Question Answering Systems - Softcover

Ofoghi, Bahadorreza

Softcover

ISBN 10: 3639292707 ISBN 13: 9783639292701

Publisher: VDM Verlag Dr. Müller, 2011

Synopsis

As the rapid rise of information stored in document databases continues, there is a real possibility of using these textual databases in systems that automatically provide answers to questions issued by users in natural language. Identification of candidate answers, within these documents, using Question Answering systems, is a challenging task that has been tackled by many researchers. One of the main problems identified in this domain is to retrieve text passages that potentially contain answers to the questions. Another problem is the extraction of text excerpts that are highly likely to answer the questions. This book discusses the work on utilizing Frame Semantics encapsulated in FrameNet to enhance the performance of semantic QA systems. FrameNet provides scenario-based generalizations over lexical items that share similar semantic backgrounds. This work also addresses some other challenges that influence frame semantic-based QA effectiveness. This includes the level of frame semantic parsing of texts, lexical coverage in FrameNet, and methods to fuse a FrameNet-based answer processing model with a non-semantic model.

About this title

About the Author

Dr. Ofoghi has carried out research in Intelligent Information Retrieval, Natural Language Processing, and Data Mining since 2002. He successfully completed his PhD at University of Ballarat, Australia in 2009 and is currently a Research Fellow at Victoria University and the Victorian Institute of Sport, Melbourne, Asutralia.

Edition details

• Publisher: VDM Verlag Dr. Müller
• Publication date: 2011
• Language: English
• ISBN 10: 3639292707
• ISBN 13: 9783639292701
• Binding: Paperback
• Number of pages: 264