-
Tytuł:
-
Non-negative matrix factorization for implicit aspect identification.
-
Autorzy:
-
Xu, Qiannan
Zhu, Li
Dai, Tao
Guo, Lei
Cao, Sisi
-
Źródło:
-
Journal of Ambient Intelligence & Humanized Computing; Jul2020, Vol. 11 Issue 7, p2683-2699, 17p
-
Sentiment analysis, also named opinion mining, is an important task in e-commerce. Recent years, many researchers have been focused on fine-grained sentiment analysis. Aspect level opinion mining detects the detailed sentiments about features of products. However, current aspect identification methods mainly focus on extracting explicit appeared aspects. The task of implicit aspect identification is still a big challenge in sentiment analysis. In this paper, we propose a novel implicit aspect identification approach based on non-negative matrix factorization. The approach first clusters product aspects by combining the co-occurrence information with intra-relations of aspect and opinion words, which can enhance the performance of aspect clustering substantially. In the next step, the approach collects context information of aspects, and represents review sentences by word vectors. Finally, a classifier is constructed to identify and predict the target implicit aspects. We also prove the convergence of our approach. Experimental results demonstrate that our approach outperforms baseline methods in most cases. [ABSTRACT FROM AUTHOR]
Copyright of Journal of Ambient Intelligence & Humanized Computing is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)