arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2510.09293cs.CL
收藏

一句话,两种嵌入:显式和隐式语义的对比学习

One Sentence, Two Embeddings: Contrastive Learning of Explicit and Implicit Semantic Representations

Japan Advanced Institute of Science and Technology(日本先进科学研究院);Toshiba Corporation(东芝公司)

Kohei Oda, Po-Min Chuang, Kiyoaki Shirai, Natthawut Kertkeidkachorn

AI总结:

DualCSE通过对比学习为句子生成显式和隐式语义嵌入,提升信息检索和文本分类任务性能。

摘要:

句子嵌入方法取得了显著进展,但它们仍然难以捕捉句子中的隐式语义。这可以归因于传统句子嵌入方法的固有局限,即每个句子只能分配一个向量。为克服这一限制,我们提出了DualCSE,一种句子嵌入方法,为每个句子分配两种嵌入:一种表示显式语义,另一种表示隐式语义。这些嵌入共存于共享空间中,使在特定用途如信息检索和文本分类中选择所需语义成为可能。实验结果表明,DualCSE能够有效编码显式和隐式含义,并提高下游任务的性能。

英文摘要:

Sentence embedding methods have made remarkable progress, yet they still struggle to capture the implicit semantics within sentences. This can be attributed to the inherent limitations of conventional sentence embedding methods that assign only a single vector per sentence. To overcome this limitation, we propose DualCSE, a sentence embedding method that assigns two embeddings to each sentence: one representing the explicit semantics and the other representing the implicit semantics. These embeddings coexist in the shared space, enabling the selection of the desired semantics for specific purposes such as information retrieval and text classification. Experimental results demonstrate that DualCSE can effectively encode both explicit and implicit meanings and improve the performance of the downstream task.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑