Interpretability in Sentiment Analysis: A Self-Supervised Approach to Sentiment Cue Extraction
In this paper, we present a novel self-supervised framework for Sentiment Cue Extraction (SCE) aimed at enhancing the interpretability of text sentiment analysis models. Our approach leverages self-supervised learning to identify and highlight key textual elements that significantly influence sentim...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-03-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/14/7/2737 |
| _version_ | 1797212905055191040 |
|---|---|
| author | Yawei Sun Saike He Xu Han Yan Luo |
| author_facet | Yawei Sun Saike He Xu Han Yan Luo |
| author_sort | Yawei Sun |
| collection | DOAJ |
| description | In this paper, we present a novel self-supervised framework for Sentiment Cue Extraction (SCE) aimed at enhancing the interpretability of text sentiment analysis models. Our approach leverages self-supervised learning to identify and highlight key textual elements that significantly influence sentiment classification decisions. Central to our framework is the development of an innovative Mask Sequence Interpretation Score (MSIS), a bespoke metric designed to assess the relevance and coherence of identified sentiment cues within binary text classification tasks. By employing Monte Carlo Sampling techniques optimized for computational efficiency, our framework demonstrates exceptional effectiveness in processing large-scale text data across diverse datasets, including English and Chinese, thus proving its versatility and scalability. The effectiveness of our approach is validated through extensive experiments on several benchmark datasets, including SST-2, IMDb, Yelp, and ChnSentiCorp. The results indicate a substantial improvement in the interpretability of the sentiment analysis models without compromising their predictive accuracy. Furthermore, our method stands out for its global interpretability, offering an efficient solution for analyzing new data compared to traditional techniques focused on local explanations. |
| first_indexed | 2024-04-24T10:49:48Z |
| format | Article |
| id | doaj.art-4e8a4379e6dc4b8298c15d1a1775d9fd |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-04-24T10:49:48Z |
| publishDate | 2024-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-4e8a4379e6dc4b8298c15d1a1775d9fd2024-04-12T13:14:40ZengMDPI AGApplied Sciences2076-34172024-03-01147273710.3390/app14072737Interpretability in Sentiment Analysis: A Self-Supervised Approach to Sentiment Cue ExtractionYawei Sun0Saike He1Xu Han2Yan Luo3Key Laboratory of Trustworthy Distributed Computing and Service (BUPT), Ministry of Education, Beijing University of Posts and Telecommunications, Beijing 100876, ChinaState Key Laboratory of Multimodal Artificial Intelligence Systems, Institute of Automation, Chinese Academy of Sciences, Beijing 100190, ChinaInstitute of Scientific and Technical Information of China, Beijing 100038, ChinaInstitute of Medical Information, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing 100020, ChinaIn this paper, we present a novel self-supervised framework for Sentiment Cue Extraction (SCE) aimed at enhancing the interpretability of text sentiment analysis models. Our approach leverages self-supervised learning to identify and highlight key textual elements that significantly influence sentiment classification decisions. Central to our framework is the development of an innovative Mask Sequence Interpretation Score (MSIS), a bespoke metric designed to assess the relevance and coherence of identified sentiment cues within binary text classification tasks. By employing Monte Carlo Sampling techniques optimized for computational efficiency, our framework demonstrates exceptional effectiveness in processing large-scale text data across diverse datasets, including English and Chinese, thus proving its versatility and scalability. The effectiveness of our approach is validated through extensive experiments on several benchmark datasets, including SST-2, IMDb, Yelp, and ChnSentiCorp. The results indicate a substantial improvement in the interpretability of the sentiment analysis models without compromising their predictive accuracy. Furthermore, our method stands out for its global interpretability, offering an efficient solution for analyzing new data compared to traditional techniques focused on local explanations.https://www.mdpi.com/2076-3417/14/7/2737sentiment cue extractionself-supervised learninginterpretable machine learning |
| spellingShingle | Yawei Sun Saike He Xu Han Yan Luo Interpretability in Sentiment Analysis: A Self-Supervised Approach to Sentiment Cue Extraction Applied Sciences sentiment cue extraction self-supervised learning interpretable machine learning |
| title | Interpretability in Sentiment Analysis: A Self-Supervised Approach to Sentiment Cue Extraction |
| title_full | Interpretability in Sentiment Analysis: A Self-Supervised Approach to Sentiment Cue Extraction |
| title_fullStr | Interpretability in Sentiment Analysis: A Self-Supervised Approach to Sentiment Cue Extraction |
| title_full_unstemmed | Interpretability in Sentiment Analysis: A Self-Supervised Approach to Sentiment Cue Extraction |
| title_short | Interpretability in Sentiment Analysis: A Self-Supervised Approach to Sentiment Cue Extraction |
| title_sort | interpretability in sentiment analysis a self supervised approach to sentiment cue extraction |
| topic | sentiment cue extraction self-supervised learning interpretable machine learning |
| url | https://www.mdpi.com/2076-3417/14/7/2737 |
| work_keys_str_mv | AT yaweisun interpretabilityinsentimentanalysisaselfsupervisedapproachtosentimentcueextraction AT saikehe interpretabilityinsentimentanalysisaselfsupervisedapproachtosentimentcueextraction AT xuhan interpretabilityinsentimentanalysisaselfsupervisedapproachtosentimentcueextraction AT yanluo interpretabilityinsentimentanalysisaselfsupervisedapproachtosentimentcueextraction |