EDU-Capsule: aspect-based sentiment analysis at clause level

• Published in: Knowledge and information systems Vol. 65; no. 2; pp. 517 - 541
• Main Authors: Lin, Ting, Sun, Aixin, Wang, Yequan
• Format: Journal Article
• Language: English
• Published: London Springer London 01.02.2023; Springer Nature B.V
• Subjects: Annotations; Artificial intelligence; Computer Science; Data mining; Data Mining and Knowledge Discovery; Database Management; Datasets; Food; Information Storage and Retrieval; Information Systems and Communication Service; Information Systems Applications (incl.Internet); IT in Business; Neural networks; Regular Paper; Semantics; Sentiment analysis; Source code; Text structure; Representation learning; Elementary discourse unit (EDU); Aspect-based sentiment analysis; Capsule network
• ISSN: 0219-1377; 0219-3116
• DOI: 10.1007/s10115-022-01797-z

Many studies on aspect-based sentiment analysis (ABSA) aim to directly predict aspects and polarities at sentence level. However, it is not rare that a long sentence expresses multiple aspects. In this paper, we propose to study ABSA at EDU-level. Elementary discourse unit (EDU) in rhetorical structure theory is an atomic semantic unit, similar to a clause in a sentence. Through manual annotation of 8,823 EDUs, obtained from the SemEval-2014 Task 4 Restaurant Review dataset, we show that more than 97% of EDUs express at most one aspect. Based on this observation, we propose an EDU-level Capsule network for ABSA. EDU-Capsule learns EDU representations within its sentential context for aspect detection and sentiment prediction. EDU-Capsule outperforms strong baselines in our experiments on two benchmark datasets. Both the EDU-level annotations and EDU-Capsule source code are released to support further studies in this area.