Presentation 2012-03-13
Evaluating Documents on Historical Events by Comparing Tuple Sets Generated from Predicate-Argument Structure
Takayuki KITANO, Akihiro YAMAMOTO,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this paper, we propose an evaluation method for documents in Wikipedia. Explanation documents in Wikipedia can be edited by various users, and this causes doubt whether or not it contains a minimal standard of content. To solve this problem, we propose an evaluation method for explanation documents in Wikipedia by rating how much content of explanation documents in Wikipedia covers that in representative dictionaries on the same topic. We use predicate-argument structure to conduct an evaluation treating the meaning of sentences. We transform the structure generated from each text in an explanation document into several tuples and express the explanation document as a set of tuples. We compare these sets. Most attributes of the tuple correspond to case relations. We also introduce some methods, e.g. transforming tuples, detecting synonyms. We have succeeded in evaluating explanation documents on Japanese historical events in Wikipedia in approximately 49.1% of recall and approximately 93.4% of accuracy comparing with evaluation by humans.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Predicate-Argument Structure / Ellipsis Resolution / Evaluating Documents
Paper # AI2011-47
Date of Issue

Conference Information
Committee AI
Conference Date 2012/3/6(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Artificial Intelligence and Knowledge-Based Processing (AI)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Evaluating Documents on Historical Events by Comparing Tuple Sets Generated from Predicate-Argument Structure
Sub Title (in English)
Keyword(1) Predicate-Argument Structure
Keyword(2) Ellipsis Resolution
Keyword(3) Evaluating Documents
1st Author's Name Takayuki KITANO
1st Author's Affiliation Department of Intelligence Science and Technology, Graduate School of Informatics, Kyoto University()
2nd Author's Name Akihiro YAMAMOTO
2nd Author's Affiliation Department of Intelligence Science and Technology, Graduate School of Informatics, Kyoto University
Date 2012-03-13
Paper # AI2011-47
Volume (vol) vol.111
Number (no) 474
Page pp.pp.-
#Pages 6
Date of Issue