Unsupervised Candidate Ranking for Lexical Substitution via Holistic Sentence Semantics
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866912586298032128 |
|---|---|
| author | Hu, Zhongyang Gu, Naijie Tao, Xiangzhi Gu, Tianhui Zhou, Yibing |
| author_facet | Hu, Zhongyang Gu, Naijie Tao, Xiangzhi Gu, Tianhui Zhou, Yibing |
| contents | A key subtask in lexical substitution is ranking the given candidate words. A common approach is to replace the target word with a candidate in the original sentence and feed the modified sentence into a model to capture semantic differences before and after substitution. However, effectively modeling the bidirectional influence of candidate substitution on both the target word and its context remains challenging. Existing methods often focus solely on semantic changes at the target position or rely on parameter tuning over multiple evaluation metrics, making it difficult to accurately characterize semantic variation. To address this, we investigate two approaches: one based on attention weights and another leveraging the more interpretable integrated gradients method, both designed to measure the influence of context tokens on the target token and to rank candidates by incorporating semantic similarity between the original and substituted sentences. Experiments on the LS07 and SWORDS datasets demonstrate that both approaches improve ranking performance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_11513 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Unsupervised Candidate Ranking for Lexical Substitution via Holistic Sentence Semantics Hu, Zhongyang Gu, Naijie Tao, Xiangzhi Gu, Tianhui Zhou, Yibing Computation and Language Artificial Intelligence A key subtask in lexical substitution is ranking the given candidate words. A common approach is to replace the target word with a candidate in the original sentence and feed the modified sentence into a model to capture semantic differences before and after substitution. However, effectively modeling the bidirectional influence of candidate substitution on both the target word and its context remains challenging. Existing methods often focus solely on semantic changes at the target position or rely on parameter tuning over multiple evaluation metrics, making it difficult to accurately characterize semantic variation. To address this, we investigate two approaches: one based on attention weights and another leveraging the more interpretable integrated gradients method, both designed to measure the influence of context tokens on the target token and to rank candidates by incorporating semantic similarity between the original and substituted sentences. Experiments on the LS07 and SWORDS datasets demonstrate that both approaches improve ranking performance. |
| title | Unsupervised Candidate Ranking for Lexical Substitution via Holistic Sentence Semantics |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2509.11513 |