Abstract
In order to obtain evaluation information about the various aspects of products or services, the Fine-grained Topic Sentiment Unification (FG-TSU) model is proposed based on the improvement of LDA (Latent Dirichlet Allocation) model. Firstly, the topics are divided into local and global topics. And the sliding window is introduced to lower co-occurrence information from document to sentence level, to implement fine-grained extraction of local topics. Secondly, the indicator variables are used to distinguish aspects and opinions. Finally, we incorporate the sentiment layer into LDA model to obtain the sentiment polarity of the whole review and specific aspects. The datasets of hotel and mobile phone are selected to verify the domain adaptability of this model. The experimental results verified the feasibility of FG-TSU model in the realization of opinion mining.
Highlights
With the rapid development of online shopping, E-commerce reviews have shown an explosive growth
We propose the Fine-Grained Topic Sentiment Unification (FG-TSU) model based on the extended LDA model
The FG-TSU model is worse than MG-LDA on the hotel data, probably because of the incorporation of the sentiment layer in sentiment analysis
Summary
With the rapid development of online shopping, E-commerce reviews have shown an explosive growth. Fine-grained opinion mining based on aspects of comment is proposed to extract aspects, the corresponding evaluations and sentiment analysis from online reviews. We propose the Fine-Grained Topic Sentiment Unification (FG-TSU) model based on the extended LDA model. It can be seen that fine-grained opinion mining achieves the deeper task, which can obtain the emotional opinion information of an evaluated entity and its related aspects, and can better meet the needs of users. The goal of fine-grained viewpoint mining is to extract the evaluated entities and the corresponding emotions from comments, and generate evaluation summaries (not involving the elements of time and viewpoint holders). The above model can’t distinguish aspects and opinions, and few studies combined topic modeling and sentiment analysis to provide a topic-based review summary of the product or service [16].
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.