IPASJ INTERNATIONAL JOURNAL OF INFORMATION TECHNOLOGY (IIJIT)

Run by A Publisher for Research Motivation ........

ISSN 2321-5976

www.ipasj.org/IIJIT/IIJIT.htm

IMPACT FACTOR

Cross-lingual Sentiment Lexicon Learning With Bilingual Word Graph Label Propagation, Authors : Haig, Alexander,IPASJ INTERNATIONAL JOURNAL OF INFORMATION TECHNOLOGY (IIJIT) ,www.ipasj.org/IIJIT/IIJIT.htm

Volume & Issue no: Volume 5, Issue 11, November 2017


Title:
Cross-lingual Sentiment Lexicon Learning With Bilingual Word Graph Label Propagation
Author Name:
Haig, Alexander
Abstract:
ABSTRACT In this article we tend to address the task of cross-lingual sentiment lexicon learning, that aims to mechanically generate sentiment lexicons for the target languages with obtainable English sentiment lexicons. we tend to formalize the task as a learning drawback on a bilingual word graph, within which the intra-language relations among the words within the same language and therefore the lingua franca relations among the words between totally different languages area unit properly drawn. With the words within the English sentiment lexicon as seeds, we tend to propose a bilingual word graph label propagation approach to induce sentiment polarities of the untagged words within the target language. notably, we tend to show that each equivalent word and opposite word relations is wont to build the intra-language relation, which the word alignment info derived from bilingual parallel sentences is effectively leveraged to make the inter-language relation. The analysis of Chinese sentiment lexicon learning shows that the projected approach outperforms existing approaches in each exactness and recall. Experiments conducted on the NTCIR information set any demonstrate the effectiveness of the learned sentiment lexicon in sentence-level sentiment classification.
Cite this article:
Haig, Alexander , " Cross-lingual Sentiment Lexicon Learning With Bilingual Word Graph Label Propagation " , IPASJ INTERNATIONAL JOURNAL OF INFORMATION TECHNOLOGY (IIJIT), Volume 5, Issue 11, November 2017 , pp. 012-016 , ISSN 2321-5976.
Full Text [PDF]                          Home