Matches in Ghent University Academic Bibliography for { <https://biblio.ugent.be/publication/01HCYEXWFET0907E8EG2T46RAW> ?p ?o. }
Showing items 1 to 23 of
23
with 100 items per page.
- 01HCYEXWFET0907E8EG2T46RAW classification C1.
- 01HCYEXWFET0907E8EG2T46RAW date "2023".
- 01HCYEXWFET0907E8EG2T46RAW language "eng".
- 01HCYEXWFET0907E8EG2T46RAW type conference.
- 01HCYEXWFET0907E8EG2T46RAW hasPart 01HCYF087HQNDV1YK0WMJ3MRZF.pdf.
- 01HCYEXWFET0907E8EG2T46RAW subject "Technology and Engineering".
- 01HCYEXWFET0907E8EG2T46RAW doi "10.18653/v1/2023.nlp4convai-1.7".
- 01HCYEXWFET0907E8EG2T46RAW presentedAt urn:uuid:8b36a31a-9034-4af3-8bd9-41ca41219d6a.
- 01HCYEXWFET0907E8EG2T46RAW abstract "Intent discovery is the task of inferring latent intents from a set of unlabeled utterances, and is a useful step towards the efficient creation of new conversational agents. We show that recent competitive methods in intent discovery can be outperformed by clustering utterances based on abstractive summaries, i.e., ``labels'', that retain the core elements while removing non-essential information. We contribute the IDAS approach, which collects a set of descriptive utterance labels by prompting a Large Language Model, starting from a well-chosen seed set of prototypical utterances, to bootstrap an In-Context Learning procedure to generate labels for non-prototypical utterances. The utterances and their resulting noisy labels are then encoded by a frozen pre-trained encoder, and subsequently clustered to recover the latent intents. For the unsupervised task (without any intent labels) IDAS outperforms the state-of-the-art by up to +7.42\% in standard cluster metrics for the Banking, StackOverflow, and Transport datasets. For the semi-supervised task (with labels for a subset of intents) IDAS surpasses 2 recent methods on the CLINC benchmark without even using labeled data.".
- 01HCYEXWFET0907E8EG2T46RAW author 0628AA40-F0EF-11E1-A197-91C894A0A6B4.
- 01HCYEXWFET0907E8EG2T46RAW author 07BA1E3A-F0EE-11E1-A9DE-61C894A0A6B4.
- 01HCYEXWFET0907E8EG2T46RAW author F5DC76B8-F0ED-11E1-A9DE-61C894A0A6B4.
- 01HCYEXWFET0907E8EG2T46RAW author F7F44048-F0ED-11E1-A9DE-61C894A0A6B4.
- 01HCYEXWFET0907E8EG2T46RAW dateCreated "2023-10-17T09:19:34Z".
- 01HCYEXWFET0907E8EG2T46RAW dateModified "2024-12-12T20:12:03Z".
- 01HCYEXWFET0907E8EG2T46RAW editor urn:uuid:72643517-bfc8-4449-868f-b4d85c457a08.
- 01HCYEXWFET0907E8EG2T46RAW editor urn:uuid:c527d9a9-3da6-4b51-97a8-9e768ac4da9c.
- 01HCYEXWFET0907E8EG2T46RAW name "IDAS : intent discovery with abstractive summarization".
- 01HCYEXWFET0907E8EG2T46RAW pagination urn:uuid:ca51a450-db16-4d4b-a151-5937220e9728.
- 01HCYEXWFET0907E8EG2T46RAW publisher urn:uuid:aeb09889-8a7b-42b0-a899-99086da4ffb2.
- 01HCYEXWFET0907E8EG2T46RAW sameAs LU-01HCYEXWFET0907E8EG2T46RAW.
- 01HCYEXWFET0907E8EG2T46RAW sourceOrganization urn:uuid:e296702a-c73f-43eb-901e-19142be85e03.
- 01HCYEXWFET0907E8EG2T46RAW type C1.