Global-Local Dynamic Adversarial Learning for Cross-Domain Sentiment Analysis
As one of the most widely used applications in domain adaption (DA), Cross-domain sentiment analysis (CDSA) aims to tackle the barrier of lacking in sentiment labeled data. Applying an adversarial network to DA to reduce the distribution discrepancy between source and target domains is a significant...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-07-01
|
| Series: | Mathematics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-7390/11/14/3130 |
| _version_ | 1797588406615670784 |
|---|---|
| author | Juntao Lyu Zheyuan Zhang Shufeng Chen Xiying Fan |
| author_facet | Juntao Lyu Zheyuan Zhang Shufeng Chen Xiying Fan |
| author_sort | Juntao Lyu |
| collection | DOAJ |
| description | As one of the most widely used applications in domain adaption (DA), Cross-domain sentiment analysis (CDSA) aims to tackle the barrier of lacking in sentiment labeled data. Applying an adversarial network to DA to reduce the distribution discrepancy between source and target domains is a significant advance in CDSA. This adversarial DA paradigm utilizes a single global domain discriminator or a series of local domain discriminators to reduce marginal or conditional probability distribution discrepancies. In general, each discrepancy has a different effect on domain adaption. However, the existing CDSA algorithms ignore this point. Therefore, in this paper, we propose an effective, novel and unsupervised adversarial DA paradigm, Global-Local Dynamic Adversarial Learning (<span style="font-variant: small-caps;">GLDAL</span>). This paradigm is able to quantitively evaluate the weights of global distribution and every local distribution. We also study how to apply <span style="font-variant: small-caps;">GLDAL</span> to CDSA. As GLDAL can effectively reduce the distribution discrepancy between domains, it performs well in a series of CDSA experiments and achieves improvements in classification accuracy compared to similar methods. The effectiveness of each component is demonstrated through ablation experiments on different parts and a quantitative analysis of the dynamic factor. Overall, this approach achieves the desired DA effect with domain shifts. |
| first_indexed | 2024-03-11T00:51:32Z |
| format | Article |
| id | doaj.art-2a249eacce1245bab17a6b9c210e6fae |
| institution | Directory Open Access Journal |
| issn | 2227-7390 |
| language | English |
| last_indexed | 2024-03-11T00:51:32Z |
| publishDate | 2023-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Mathematics |
| spelling | doaj.art-2a249eacce1245bab17a6b9c210e6fae2023-11-18T20:21:00ZengMDPI AGMathematics2227-73902023-07-011114313010.3390/math11143130Global-Local Dynamic Adversarial Learning for Cross-Domain Sentiment AnalysisJuntao Lyu0Zheyuan Zhang1Shufeng Chen2Xiying Fan3School of Computer and Communication Engineering, University of Science and Technology Beijing, Beijing 100083, ChinaSchool of Computer and Communication Engineering, University of Science and Technology Beijing, Beijing 100083, ChinaSchool of Computer and Communication Engineering, University of Science and Technology Beijing, Beijing 100083, ChinaSchool of Computer and Communication Engineering, University of Science and Technology Beijing, Beijing 100083, ChinaAs one of the most widely used applications in domain adaption (DA), Cross-domain sentiment analysis (CDSA) aims to tackle the barrier of lacking in sentiment labeled data. Applying an adversarial network to DA to reduce the distribution discrepancy between source and target domains is a significant advance in CDSA. This adversarial DA paradigm utilizes a single global domain discriminator or a series of local domain discriminators to reduce marginal or conditional probability distribution discrepancies. In general, each discrepancy has a different effect on domain adaption. However, the existing CDSA algorithms ignore this point. Therefore, in this paper, we propose an effective, novel and unsupervised adversarial DA paradigm, Global-Local Dynamic Adversarial Learning (<span style="font-variant: small-caps;">GLDAL</span>). This paradigm is able to quantitively evaluate the weights of global distribution and every local distribution. We also study how to apply <span style="font-variant: small-caps;">GLDAL</span> to CDSA. As GLDAL can effectively reduce the distribution discrepancy between domains, it performs well in a series of CDSA experiments and achieves improvements in classification accuracy compared to similar methods. The effectiveness of each component is demonstrated through ablation experiments on different parts and a quantitative analysis of the dynamic factor. Overall, this approach achieves the desired DA effect with domain shifts.https://www.mdpi.com/2227-7390/11/14/3130adversarial domain adaptioncross-domain sentiment analysisglobal-local dynamic adversarial learning |
| spellingShingle | Juntao Lyu Zheyuan Zhang Shufeng Chen Xiying Fan Global-Local Dynamic Adversarial Learning for Cross-Domain Sentiment Analysis Mathematics adversarial domain adaption cross-domain sentiment analysis global-local dynamic adversarial learning |
| title | Global-Local Dynamic Adversarial Learning for Cross-Domain Sentiment Analysis |
| title_full | Global-Local Dynamic Adversarial Learning for Cross-Domain Sentiment Analysis |
| title_fullStr | Global-Local Dynamic Adversarial Learning for Cross-Domain Sentiment Analysis |
| title_full_unstemmed | Global-Local Dynamic Adversarial Learning for Cross-Domain Sentiment Analysis |
| title_short | Global-Local Dynamic Adversarial Learning for Cross-Domain Sentiment Analysis |
| title_sort | global local dynamic adversarial learning for cross domain sentiment analysis |
| topic | adversarial domain adaption cross-domain sentiment analysis global-local dynamic adversarial learning |
| url | https://www.mdpi.com/2227-7390/11/14/3130 |
| work_keys_str_mv | AT juntaolyu globallocaldynamicadversariallearningforcrossdomainsentimentanalysis AT zheyuanzhang globallocaldynamicadversariallearningforcrossdomainsentimentanalysis AT shufengchen globallocaldynamicadversariallearningforcrossdomainsentimentanalysis AT xiyingfan globallocaldynamicadversariallearningforcrossdomainsentimentanalysis |