Extracting Aspects Hierarchies using Rhetorical Structure Theory

Abstract

We propose a novel approach to generate aspect hierarchies that proved to be consistently correct compared with human-generated hierarchies. We present an unsupervised technique using Rhetorical Structure Theory and graph analysis. We evaluated our approach based on 100,000 reviews from Amazon and achieved an astonishing 80% coverage compared with human-generated hierarchies coded in ConceptNet. The method could be easily extended with a sentiment analysis model and used to describe sentiment on different levels of aspect granularity. Hence, besides the flat aspect structure, we can differentiate between aspects and describe if the charging aspect is related to battery or price.

arXiv commentACAI 2018 MLNLP

References in corpus (4)

• Ups and Downs: Modeling the Visual Evolution of Fashion Trends with One-Class Collaborative Filtering
• Two-pass Discourse Segmentation with Pairing and Global Features
• All Maximal Independent Sets and Dynamic Dominance for Sparse Graphs
• Method for Aspect-Based Sentiment Annotation Using Rhetorical Analysis