Abstract
With the increasing frequency of documents appearing in scenarios including online comments, car reviews and movie evaluations, the text length in sentiment analysis tasks has become longer. Since documents consist of multiple sentences, aspect terms and opinion terms in aspect sentiment triplets may not be in the same sentence, which requires multi-hop reasoning across sentences for extraction, which exceeds the scope of intra-sentence aspect sentiment triplet extraction. Apparently, the existing models at the sentence level are not suitable for extracting sentiment triples directly from documents. Therefore, a well-performing document sentiment analysis model is needed to address the degraded performance of triplet extraction caused by cross-sentence sentiment elements in long texts. To promote the development of aspect sentiment triplet extraction from documents, a dataset based on movie review texts is proposed and the existing methods are evaluated based on this dataset. Through error analysis, this study explores the difficulties and possible future research directions of document-level aspect sentiment triplet extraction.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
