Matches in SemOpenAlex for { <https://semopenalex.org/work/W4310631728> ?p ?o ?g. }
Showing items 1 to 61 of
61
with 100 items per page.
- W4310631728 abstract "The effectiveness of contrastive learning technology in natural language processing tasks is yet to be explored and analyzed. How to construct positive and negative samples correctly and reasonably is the core challenge of contrastive learning. It is even harder to discover contrastive objects in multi-label text classification tasks. There are very few contrastive losses proposed previously. In this paper, we investigate the problem from a different angle by proposing five novel contrastive losses for multi-label text classification tasks. These are Strict Contrastive Loss (SCL), Intra-label Contrastive Loss (ICL), Jaccard Similarity Contrastive Loss (JSCL), Jaccard Similarity Probability Contrastive Loss (JSPCL), and Stepwise Label Contrastive Loss (SLCL). We explore the effectiveness of contrastive learning for multi-label text classification tasks by the employment of these novel losses and provide a set of baseline models for deploying contrastive learning techniques on specific tasks. We further perform an interpretable analysis of our approach to show how different components of contrastive learning losses play their roles. The experimental results show that our proposed contrastive losses can bring improvement to multi-label text classification tasks. Our work also explores how contrastive learning should be adapted for multi-label text classification tasks." @default.
- W4310631728 created "2022-12-13" @default.
- W4310631728 creator A5024135429 @default.
- W4310631728 creator A5030446424 @default.
- W4310631728 creator A5048852821 @default.
- W4310631728 creator A5070480766 @default.
- W4310631728 creator A5085281557 @default.
- W4310631728 date "2022-12-01" @default.
- W4310631728 modified "2023-09-27" @default.
- W4310631728 title "An Effective Deployment of Contrastive Learning in Multi-label Text Classification" @default.
- W4310631728 doi "https://doi.org/10.48550/arxiv.2212.00552" @default.
- W4310631728 hasPublicationYear "2022" @default.
- W4310631728 type Work @default.
- W4310631728 citedByCount "0" @default.
- W4310631728 crossrefType "posted-content" @default.
- W4310631728 hasAuthorship W4310631728A5024135429 @default.
- W4310631728 hasAuthorship W4310631728A5030446424 @default.
- W4310631728 hasAuthorship W4310631728A5048852821 @default.
- W4310631728 hasAuthorship W4310631728A5070480766 @default.
- W4310631728 hasAuthorship W4310631728A5085281557 @default.
- W4310631728 hasBestOaLocation W43106317281 @default.
- W4310631728 hasConcept C103278499 @default.
- W4310631728 hasConcept C115961682 @default.
- W4310631728 hasConcept C138885662 @default.
- W4310631728 hasConcept C153180895 @default.
- W4310631728 hasConcept C154945302 @default.
- W4310631728 hasConcept C177264268 @default.
- W4310631728 hasConcept C199360897 @default.
- W4310631728 hasConcept C203519979 @default.
- W4310631728 hasConcept C204321447 @default.
- W4310631728 hasConcept C2777629044 @default.
- W4310631728 hasConcept C41008148 @default.
- W4310631728 hasConcept C41895202 @default.
- W4310631728 hasConceptScore W4310631728C103278499 @default.
- W4310631728 hasConceptScore W4310631728C115961682 @default.
- W4310631728 hasConceptScore W4310631728C138885662 @default.
- W4310631728 hasConceptScore W4310631728C153180895 @default.
- W4310631728 hasConceptScore W4310631728C154945302 @default.
- W4310631728 hasConceptScore W4310631728C177264268 @default.
- W4310631728 hasConceptScore W4310631728C199360897 @default.
- W4310631728 hasConceptScore W4310631728C203519979 @default.
- W4310631728 hasConceptScore W4310631728C204321447 @default.
- W4310631728 hasConceptScore W4310631728C2777629044 @default.
- W4310631728 hasConceptScore W4310631728C41008148 @default.
- W4310631728 hasConceptScore W4310631728C41895202 @default.
- W4310631728 hasLocation W43106317281 @default.
- W4310631728 hasOpenAccess W4310631728 @default.
- W4310631728 hasPrimaryLocation W43106317281 @default.
- W4310631728 hasRelatedWork W1506512272 @default.
- W4310631728 hasRelatedWork W1968351314 @default.
- W4310631728 hasRelatedWork W2091133150 @default.
- W4310631728 hasRelatedWork W2157470625 @default.
- W4310631728 hasRelatedWork W2186284405 @default.
- W4310631728 hasRelatedWork W2797826682 @default.
- W4310631728 hasRelatedWork W3199095181 @default.
- W4310631728 hasRelatedWork W4281690070 @default.
- W4310631728 hasRelatedWork W4296300627 @default.
- W4310631728 hasRelatedWork W73805934 @default.
- W4310631728 isParatext "false" @default.
- W4310631728 isRetracted "false" @default.
- W4310631728 workType "article" @default.