FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4385488503> ?p ?o ?g. }

• W4385488503 abstract "Recently, contrastive learning (CL) has emerged as the fundamental framework for learning better sentence representations. In the unsupervised sentence representation task, due to the lack of labeled data, current CL-based approaches generally use various methods to generate or select positive and negative samples for the given sentence. Despite their success, existing CL-based unsupervised sentence representation methods underestimate hard positive samples and hard negative samples, which do not fully exploit the power of contrastive learning. In this paper, we argue that we need to focus more on hard positive and hard negative samples. To this end, we propose a novel contrastive learning model, HSimCSE, that extends SimCSE by considering both the hard positive and hard negative samples. Specifically, we first propose a novel adversarial positive sample generation module to generate an adversarial hard positive sample, then we propose a dual negative sample selection module to select hard negative samples from the in-batch samples and the entire training corpus. Finally, we propose a quadruplet loss to minimize the distance between the anchor sample and the adversarial hard positive sample and maximize the distance between the anchor sample and the two hard negative samples. Experiments conducted on seven semantic text similarity tasks demonstrate the effectiveness of our method. The source code can be found at https://github.com/xubodhu/HSimCSE." @default.
• W4385488503 created "2023-08-03" @default.
• W4385488503 creator A5020681500 @default.
• W4385488503 creator A5032547376 @default.
• W4385488503 creator A5035378893 @default.
• W4385488503 creator A5054796181 @default.
• W4385488503 creator A5056936324 @default.
• W4385488503 creator A5081861593 @default.
• W4385488503 creator A5087643562 @default.
• W4385488503 date "2023-06-18" @default.
• W4385488503 modified "2023-10-06" @default.
• W4385488503 title "HSimCSE: Improving Contrastive Learning of Unsupervised Sentence Representation with Adversarial Hard Positives and Dual Hard Negatives" @default.
• W4385488503 cites W2126400076 @default.
• W4385488503 cites W2133458109 @default.
• W4385488503 cites W2138621090 @default.
• W4385488503 cites W2462305634 @default.
• W4385488503 cites W2970641574 @default.
• W4385488503 cites W2988217457 @default.
• W4385488503 cites W3099700870 @default.
• W4385488503 cites W3100652389 @default.
• W4385488503 cites W3104033643 @default.
• W4385488503 cites W3105816068 @default.
• W4385488503 cites W3156636935 @default.
• W4385488503 cites W3173190788 @default.
• W4385488503 cites W3175362188 @default.
• W4385488503 cites W3176047188 @default.
• W4385488503 cites W3200253633 @default.
• W4385488503 cites W3206240757 @default.
• W4385488503 cites W3213139842 @default.
• W4385488503 cites W4225385501 @default.
• W4385488503 doi "https://doi.org/10.1109/ijcnn54540.2023.10191335" @default.
• W4385488503 hasPublicationYear "2023" @default.
• W4385488503 type Work @default.
• W4385488503 citedByCount "0" @default.
• W4385488503 crossrefType "proceedings-article" @default.
• W4385488503 hasAuthorship W4385488503A5020681500 @default.
• W4385488503 hasAuthorship W4385488503A5032547376 @default.
• W4385488503 hasAuthorship W4385488503A5035378893 @default.
• W4385488503 hasAuthorship W4385488503A5054796181 @default.
• W4385488503 hasAuthorship W4385488503A5056936324 @default.
• W4385488503 hasAuthorship W4385488503A5081861593 @default.
• W4385488503 hasAuthorship W4385488503A5087643562 @default.
• W4385488503 hasConcept C119857082 @default.
• W4385488503 hasConcept C120665830 @default.
• W4385488503 hasConcept C121332964 @default.
• W4385488503 hasConcept C154945302 @default.
• W4385488503 hasConcept C162324750 @default.
• W4385488503 hasConcept C17744445 @default.
• W4385488503 hasConcept C185592680 @default.
• W4385488503 hasConcept C187736073 @default.
• W4385488503 hasConcept C192209626 @default.
• W4385488503 hasConcept C198531522 @default.
• W4385488503 hasConcept C199539241 @default.
• W4385488503 hasConcept C204321447 @default.
• W4385488503 hasConcept C2776359362 @default.
• W4385488503 hasConcept C2777530160 @default.
• W4385488503 hasConcept C2780451532 @default.
• W4385488503 hasConcept C37736160 @default.
• W4385488503 hasConcept C41008148 @default.
• W4385488503 hasConcept C43617362 @default.
• W4385488503 hasConcept C59404180 @default.
• W4385488503 hasConcept C64869954 @default.
• W4385488503 hasConcept C8038995 @default.
• W4385488503 hasConcept C94625758 @default.
• W4385488503 hasConceptScore W4385488503C119857082 @default.
• W4385488503 hasConceptScore W4385488503C120665830 @default.
• W4385488503 hasConceptScore W4385488503C121332964 @default.
• W4385488503 hasConceptScore W4385488503C154945302 @default.
• W4385488503 hasConceptScore W4385488503C162324750 @default.
• W4385488503 hasConceptScore W4385488503C17744445 @default.
• W4385488503 hasConceptScore W4385488503C185592680 @default.
• W4385488503 hasConceptScore W4385488503C187736073 @default.
• W4385488503 hasConceptScore W4385488503C192209626 @default.
• W4385488503 hasConceptScore W4385488503C198531522 @default.
• W4385488503 hasConceptScore W4385488503C199539241 @default.
• W4385488503 hasConceptScore W4385488503C204321447 @default.
• W4385488503 hasConceptScore W4385488503C2776359362 @default.
• W4385488503 hasConceptScore W4385488503C2777530160 @default.
• W4385488503 hasConceptScore W4385488503C2780451532 @default.
• W4385488503 hasConceptScore W4385488503C37736160 @default.
• W4385488503 hasConceptScore W4385488503C41008148 @default.
• W4385488503 hasConceptScore W4385488503C43617362 @default.
• W4385488503 hasConceptScore W4385488503C59404180 @default.
• W4385488503 hasConceptScore W4385488503C64869954 @default.
• W4385488503 hasConceptScore W4385488503C8038995 @default.
• W4385488503 hasConceptScore W4385488503C94625758 @default.
• W4385488503 hasLocation W43854885031 @default.
• W4385488503 hasOpenAccess W4385488503 @default.
• W4385488503 hasPrimaryLocation W43854885031 @default.
• W4385488503 hasRelatedWork W1492132281 @default.
• W4385488503 hasRelatedWork W159132833 @default.
• W4385488503 hasRelatedWork W2889818188 @default.
• W4385488503 hasRelatedWork W3046843850 @default.
• W4385488503 hasRelatedWork W3105747907 @default.
• W4385488503 hasRelatedWork W3196155444 @default.
• W4385488503 hasRelatedWork W3209574120 @default.
• W4385488503 hasRelatedWork W4285260836 @default.
• W4385488503 hasRelatedWork W4386462264 @default.
• W4385488503 hasRelatedWork W4386716251 @default.
• W4385488503 isParatext "false" @default.

• next