MiSAS: a multi-domain feature-level sentiment analysis system on micro-blog

Big data from micro-blog has been an important access to social groups' psychology, market feedback and so on. Unlike the review corpus which is usually related to the specific object (e.g. a product), the micro-blog content covers the opinion of many domains. It is less useful to extract the fine-grained feature-level opinion target without detect the domain. This paper proposed a systematic feature-level sentiment analysis approach on Micro-blog that recognize data related to the interesting topic automatically. Working with the big micro-blog data we figure out valuable text features to train the opinion targets extraction and sentimental polarity detection models that achieve better multi-domain adaption. We implement the MiSAS system, which crawls micro-blog raw data, outputs opinion targets and orientation summarization on the giving domains, offering valuable analytical tool for practical applications.

[1]  Preslav Nakov,et al.  SemEval-2015 Task 10: Sentiment Analysis in Twitter , 2015, *SEMEVAL.

[2]  Claire Cardie,et al.  Joint Inference for Fine-grained Opinion Extraction , 2013, ACL.

[3]  Ding Shengchun,et al.  Opinion Targets Extraction from Chinese Microblogs Based on Conditional Random Fields and Domain Ontology , 2016 .

[4]  Guodong Zhou,et al.  Opinion Target Extraction Using a Shallow Semantic Parsing Framework , 2012, AAAI.

[5]  Lin Hong-fei.  Micro-Blog Hot Events Detection Based on Emotion Distribution , 2012 .

[6]  Björn W. Schuller,et al.  New Avenues in Opinion Mining and Sentiment Analysis , 2013, IEEE Intelligent Systems.

[7]  Xuanjing Huang,et al.  FudanNLP: A Toolkit for Chinese Natural Language Processing , 2013, ACL.

[8]  Li Chen,et al.  Comparison of Model-Based Learning Methods for Feature-Level Opinion Mining , 2011, 2011 IEEE/WIC/ACM International Conferences on Web Intelligence and Intelligent Agent Technology.

[9]  Ming Zhou,et al.  Learning Sentiment-Specific Word Embedding for Twitter Sentiment Classification , 2014, ACL.

[10]  Omer Levy,et al.  word2vec Explained: deriving Mikolov et al.'s negative-sampling word-embedding method , 2014, ArXiv.

[11]  Xiao Shi-bin.  Sentiment analysis of Chinese micro-blog based on multi-feature and combined classification , 2013 .

[12]  Jian Su,et al.  NLANGP: Supervised Machine Learning System for Aspect Category Classification and Opinion Target Extraction , 2015, *SEMEVAL.

[13]  Frédéric Béchet,et al.  Lsislif: CRF and Logistic Regression for Opinion Target Extraction and Sentiment Polarity Analysis , 2015, SemEval@NAACL-HLT.

[14]  Diego Reforgiato Recupero,et al.  Frame-Based Detection of Opinion Holders and Topics: A Model and a Tool , 2014, IEEE Computational Intelligence Magazine.