W2VLDA: Almost unsupervised system for Aspect Based Sentiment Analysis

作者：

Highlights：

• Almost unsupervised Aspect Based Sentiment Analysis (ABSA).

• Based on guided topic modelling and continuous word embeddings.

• Easy to apply to different languages and domains changing a few seed words.

• Evaluated for four languages and several domains.

摘要

•Almost unsupervised Aspect Based Sentiment Analysis (ABSA).•Based on guided topic modelling and continuous word embeddings.•Easy to apply to different languages and domains changing a few seed words.•Evaluated for four languages and several domains.

论文关键词：Opinion mining,Aspect Based Sentiment Analysis,Almost unsupervised,Multilingual,Multidomain

论文评审过程：Received 19 May 2017, Revised 29 August 2017, Accepted 30 August 2017, Available online 1 September 2017, Version of Record 5 September 2017.

论文官网地址：https://doi.org/10.1016/j.eswa.2017.08.049