Premium
Joint model for subsentence‐level sentiment analysis with M arkov logic
Author(s) -
Chen Ziyan,
Huang Yu,
Tian Jing,
Liu Xiaoyan,
Fu Kun,
Huang Tinglei
Publication year - 2015
Publication title -
journal of the association for information science and technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.903
H-Index - 145
eISSN - 2330-1643
pISSN - 2330-1635
DOI - 10.1002/asi.23301
Subject(s) - computer science , sentiment analysis , sentence , latent dirichlet allocation , artificial intelligence , joint (building) , inference , natural language processing , subjectivity , topic model , principle of maximum entropy , focus (optics) , polarity (international relations) , machine learning , philosophy , architectural engineering , physics , optics , epistemology , engineering , genetics , biology , cell
Sentiment analysis mainly focuses on the study of one's opinions that express positive or negative sentiments. With the explosive growth of web documents, sentiment analysis is becoming a hot topic in both academic research and system design. Fine‐grained sentiment analysis is traditionally solved as a 2‐step strategy, which results in cascade errors. Although joint models, such as joint sentiment/topic and maximum entropy ( M ax E nt)/latent Dirichlet allocation, are proposed to tackle this problem of sentiment analysis, they focus on the joint learning of both aspects and sentiments. Thus, they are not appropriate to solve the cascade errors for sentiment analysis at the sentence or subsentence level. In this article, we present a novel jointly fine‐grained sentiment analysis framework at the subsentence level with M arkov logic. First, we divide the task into 2 separate stages (subjectivity classification and polarity classification). Then, the 2 separate stages are processed, respectively, with different feature sets, which are implemented by local formulas in M arkov logic. Finally, global formulas in M arkov logic are adopted to realize the interactions of the 2 separate stages. The joint inference of subjectivity and polarity helps prevent cascade errors. Experiments on a C hinese sentiment data set manifest that our joint model brings significant improvements.