Presentation	2013-11-29 On Retrieval Performance of a Sequence-based Retrieval Model Yoshihisa Udagawa,
PDF Download Page	PDF download Page Link
Abstract(in Japanese)	(See Japanese page)
Abstract(in English)	Retrieving similar source code is required in order to support the efficient program development and maintenance work. Source code is essentially a sequence of statements; therefore, we define a similarity model that is tailored to sequentially structured texts for retrieving similar source code fragments. Our similarity model is evaluated by comparing two well-known similarity models using Struts 2, which is a medium-size open source Java program. The results show that our similarity model outperforms the vector space retrieval model by 66.8% on average and 98.7% at the maximum.
Keyword(in Japanese)	(See Japanese page)
Keyword(in English)	Sequence structure retrieval model / Control statement / Method identifier / Vector space model
Paper #	SWIM2013-12
Date of Issue

Conference Information
Committee	SWIM
Conference Date	2013/11/22(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	Software Interprise Modeling (SWIM)
Language	JPN
Title (in Japanese)	(See Japanese page)
Sub Title (in Japanese)	(See Japanese page)
Title (in English)	On Retrieval Performance of a Sequence-based Retrieval Model
Sub Title (in English)
Keyword(1)	Sequence structure retrieval model
Keyword(2)	Control statement
Keyword(3)	Method identifier
Keyword(4)	Vector space model
1st Author's Name	Yoshihisa Udagawa
1st Author's Affiliation	Faculty of Engineering, Tokyo Polytechnic University()
Date	2013-11-29
Paper #	SWIM2013-12
Volume (vol)	vol.113
Number (no)	334
Page	pp.pp.-
#Pages	8
Date of Issue