Chinese aspect-level sentiment analysis model based on mixture vector

• 1、Beijing University of Posts and Telecommunications, Institute of network technology, 100876

Abstract：The aspect-level sentiment analysis task aims to mine the sentimental polarity of a given aspect in a sentence. The sentiment analysis model using attention mechanism has achieved excellent results in sentiment analysis at the English level, but the performance in the Chinese scene is not ideal, mainly because Chinese is more complicated in language structure than English, and it is difficult to extract semantic information of Chinese words. In order to more accurately extract the semantic features of Chinese words, based on the existing research, this paper proposes a word vector joint learning model, and builds an emotion analysis model based on the model. The experimental results show that the model based on word vector joint learning proposed in this paper is comparable to the best current results in Chinese aspect-level sentiment analysis tasks.

Keywords： aspect-level sentiment analysis character embedding word embedding