Sentilyzer: Aspect-Oriented Sentiment Analysis of Product Reviews: Short Paper submitted to CSCI18: Symposium on Computational Science (CSCI-ISCS)

摘要

Aspect-oriented sentiment analysis is used in particular on textual product reviews to identify positive or negative product characteristics. This information is beneficial not only for customers related to purchasing decisions, but also used by manufacturers to improve their products. In this short paper, we introduce Sentilyzer, a system which performs state-of-the-art aspect-oriented sentiment analysis based on automatically generated category-specific sentiment and aspect dictionaries from Amazon product reviews.