Opinion mining using ensemble text hidden Markov models for text classification
作者：
Highlights：
• Proposed a new sentiment analysis method, based on text-based hidden Markov models, that uses word orders without the need of sentiment lexicons.
• Proposed an ensemble of text-based hidden Markov models using boosting and clusters of words produced by latent semantic analysis.
• Showed the method has potential to classify implicit opinions by the proposed ensemble method.
• Showed better performance in comparison to several previous algorithms in several datasets.
• Applied it to a real-life dataset to classify paper titles.
摘要
•Proposed a new sentiment analysis method, based on text-based hidden Markov models, that uses word orders without the need of sentiment lexicons.•Proposed an ensemble of text-based hidden Markov models using boosting and clusters of words produced by latent semantic analysis.•Showed the method has potential to classify implicit opinions by the proposed ensemble method.•Showed better performance in comparison to several previous algorithms in several datasets.•Applied it to a real-life dataset to classify paper titles.
论文关键词：Opinion mining,Sentiment analysis,Hidden Markov models,Ensemble,Boosting,Clustering
论文评审过程：Received 21 March 2017, Revised 13 July 2017, Accepted 14 July 2017, Available online 19 July 2017, Version of Record 13 November 2017.
论文官网地址：https://doi.org/10.1016/j.eswa.2017.07.019
