Journals
  Publication Years
  Keywords
Search within results Open Search
Please wait a minute...
For Selected: Toggle Thumbnails
Sentiment-aspect analysis method based on seed words
CHEN Yongheng, ZUO Wanli, LIN Yaojing
Journal of Computer Applications    2015, 35 (9): 2560-2564.   DOI: 10.11772/j.issn.1001-9081.2015.09.2560
Abstract516)      PDF (884KB)(353)       Save
The analysis of sentiment-aspect for product or service is useful for finding the information of sentiment-aspect from the mess of comment set. This paper proposed a new method of sentiment-aspect based on seed words of aspect. Firstly, seed words of aspect and documents of aspect automatically could be achieved by this method. Secondly, Sentiment-Aspect Analysis model Supervised by Seed Words (SAA_SSW) was employed by this method to find aspect and related sentiment. The experimental results show that, compared with traditional Joint Sentiment/Topic Model (JST) and Aspect and Sentiment Unification Model (ASUM), SAA_SSW can find the sentiment labels for same word under different topics and achieve higher relevance between sentiment word and topic. In addition, SAA_SSW model, compared with traditional JST and ASUM model, can improve the classification accuracy by at least 7.5%. So, SAA_SSW model can achieve the extraction of sentiment-aspect well and improve the classification accuracy.
Reference | Related Articles | Metrics
Application of labeled author topic model in scientific literature
CHEN Yongheng, ZUO Wanli, LIN Yaojin
Journal of Computer Applications    2015, 35 (4): 1001-1005.   DOI: 10.11772/j.issn.1001-9081.2015.04.1001
Abstract542)      PDF (712KB)(626)       Save

Author Topic (AT) model is widely used to find the author's interests in scientific literature, but AT model cannot take advantage of the correlation between category labels and topics. Through integrating the inherent category labels of documents into AT model, Labeled Author Topic (LAT) model was proposed. LAT model realized the predicate of multi-labels by optimizing the mapping relation between labels and topics and improved the clustering results. The experimental results suggest that, compared with Latent Dirichlet Allocation (LDA) model and AT model, LAT model can improve the decision accuracy of multi-labels, and optimize the generalization ability and operating efficiency.

Reference | Related Articles | Metrics