Presentation 2007-01-30
Extraction of Computer Virus Information using Dependency Structure
Yusaku SUZUKI, Tsuyoshi YAMAMURA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) This paper proposes the method of information extraction of noun phrases and sentences to create an encyclopedia from internet news articles. We extract the computer virus information (a virus name, infection routes and symptoms) using SVM (Support Vector Machine). Our previous method which only used features of surrounding morphemes or words in the sentence didn't have good performance in extraction of clauses or sentences. So, we applied dependency structures to improve the performance. As a result, our method greatly improved accuracy compared to chunking method. The information route's F-value increased by 11.04 and the symptom's F-value increased by 19.83.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Information Extraction / Support Vector Machine / Dependency Structure
Paper # NLC2006-79
Date of Issue

Conference Information
Committee NLC
Conference Date 2007/1/23(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 Natural Language Understanding and Models of Communication (NLC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Extraction of Computer Virus Information using Dependency Structure
Sub Title (in English)
Keyword(1) Information Extraction
Keyword(2) Support Vector Machine
Keyword(3) Dependency Structure
1st Author's Name Yusaku SUZUKI
1st Author's Affiliation Graduate School of Information Science and Technology, Aichi Prefectural University()
2nd Author's Name Tsuyoshi YAMAMURA
2nd Author's Affiliation Faculty of Information Science and Technology, Aichi Prefectural University
Date 2007-01-30
Paper # NLC2006-79
Volume (vol) vol.106
Number (no) 517
Page pp.pp.-
#Pages 6
Date of Issue