PSCA: Position-Syntax and Cross-Aspect for Aspect-based Sentiment Analysis in One Model

Aspect-based sentiment analysis (ABSA) is a fine-grained sentiment analysis task that aims to identify the senti-ment polarity of specific aspects within sentence.methods based on graph neural networks that heavily rely on Syntax tree and semantic composition tree have been used to mine syntactic information In recent years, However, these methods overlook the feature of location influence factor of words and may provide irrelevant or even interfering noisy signals for ABSA due to the word association relationship mined by the Syntax tree and semantic composition tree. To alleviate the impact of noise information and fully explore the multiple factors of ABSA, In this paper, we propose a new framework PSCA which Integrating Position,Syntax and Cross-Aspect module and mining infor-mation on position importance, syntax and semantic structure dependencies and cross-aspect correlation information. Moreover, Experimental results on four real-world datasets demonstrate the effectiveness of PSCA compared to previous state-of-the-art methods.