Kang, Junmo ; Hong, Giwon ; Puerto San Roman, Haritz ; Myaeng, Sung-Hyon (2020)
Regularization of Distinct Strategies for Unsupervised Question Generation.
2020 Conference on Empirical Methods in Natural Language Processing (EMNLP 2020). virtual Conference (16.11.2020-20.11.2020)
doi: 10.18653/v1/2020.findings-emnlp.293
Konferenzveröffentlichung, Bibliographie
Kurzbeschreibung (Abstract)
Unsupervised question answering (UQA) has been proposed to avoid the high cost of creating high-quality datasets for QA. One approach to UQA is to train a QA model with questions generated automatically. However, the generated questions are either too similar to a word sequence in the context or too drifted from the semantics of the context, thereby making it difficult to train a robust QA model. We propose a novel regularization method based on teacher-student architecture to avoid bias toward a particular question generation strategy and modulate the process of generating individual words when a question is generated. Our experiments demonstrate that we have achieved the goal of generating higher-quality questions for UQA across diverse QA datasets and tasks. We also show that this method can be useful for creating a QA model with few-shot learning.
| Typ des Eintrags: | Konferenzveröffentlichung |
|---|---|
| Erschienen: | 2020 |
| Autor(en): | Kang, Junmo ; Hong, Giwon ; Puerto San Roman, Haritz ; Myaeng, Sung-Hyon |
| Art des Eintrags: | Bibliographie |
| Titel: | Regularization of Distinct Strategies for Unsupervised Question Generation |
| Sprache: | Englisch |
| Publikationsjahr: | 21 November 2020 |
| Verlag: | ACL |
| Buchtitel: | Findings of the Association for Computational Linguistics: EMNLP 2020 |
| Veranstaltungstitel: | 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP 2020) |
| Veranstaltungsort: | virtual Conference |
| Veranstaltungsdatum: | 16.11.2020-20.11.2020 |
| DOI: | 10.18653/v1/2020.findings-emnlp.293 |
| Kurzbeschreibung (Abstract): | Unsupervised question answering (UQA) has been proposed to avoid the high cost of creating high-quality datasets for QA. One approach to UQA is to train a QA model with questions generated automatically. However, the generated questions are either too similar to a word sequence in the context or too drifted from the semantics of the context, thereby making it difficult to train a robust QA model. We propose a novel regularization method based on teacher-student architecture to avoid bias toward a particular question generation strategy and modulate the process of generating individual words when a question is generated. Our experiments demonstrate that we have achieved the goal of generating higher-quality questions for UQA across diverse QA datasets and tasks. We also show that this method can be useful for creating a QA model with few-shot learning. |
| Fachbereich(e)/-gebiet(e): | 20 Fachbereich Informatik 20 Fachbereich Informatik > Ubiquitäre Wissensverarbeitung |
| Hinterlegungsdatum: | 06 Jul 2023 08:18 |
| Letzte Änderung: | 07 Jul 2023 08:59 |
| PPN: | 509437834 |
| Export: | |
| Suche nach Titel in: | TUfind oder in Google |
 |
Frage zum Eintrag |
Optionen (nur für Redakteure)
 |
Redaktionelle Details anzeigen |
Drucken
Impressum