Multilingual Document Mining and Navigation Using Self-Organizing Maps
120 likes | 249 Vues
This paper presents a novel approach to multilingual document mining and navigation through the use of self-organizing maps (SOM). It addresses the challenge posed by monolingual interfaces that restrict access for users unfamiliar with a specific language. The proposed method automatically organizes multilingual web pages into a coherent web directory, facilitating easier navigation. Key methodologies include preprocessing, document encoding, and clustering. The approach is fully automated and effectively enhances multilingual web information retrieval, breaking down language barriers for users.
Multilingual Document Mining and Navigation Using Self-Organizing Maps
E N D
Presentation Transcript
Multilingual document mining and navigation using self-organizing maps Presenter : Keng-Yu Lin Author : Hsin-Chang Yang , Han-Wei Hsiao , Chung-Hong Lee IPM .2011
Outlines • Motivation • Objectives • Methodology • Experiments • Conclusions • Comments
Motivation Monolingual interface may limit the spread of users who unfamiliar with the language.
Objectives • To propose an approach that could automatically arrange multilingual Web pages into a multilingual Web directory to break the language barriers in Web navigation.
Methodology • Preprocessing • Word segmentation • Stopword elimination • Stemming • Keyword selection • Encoding • All keywords of all documents are collected to build a vocabulary VE. • A document is encoded into a binary vector according to those keywords that occurred in it. Ex: Xi=[0,1,1,0,1,0,1,1]
Methodology => document cluster map (DCM) => keyword cluster map (KCM) • SOM Algorithm
Methodology Determining dominating clusters algorithm
Methodology (C1,C3)=4 (C3,C5)=3 (C1,C5)=3 PK=(4+3+3)/3=3.33 Evaluation of quality of generated hierarchies
Methodology • Multilingual web directory generation • Semantic similarity • Structural similarity
Conclusions The approach is fully automated and requires no human intervention. The result of the alignment can be applied to tackle tasks such as multilingual information retrieval.
Comments • Advantage • The research result can help people to break language barrier. • Applications • Multilingual information retrieval.