Multi-View Domain Adapted Sentence Embeddings for Low-Resource Unsupervised Duplicate Question Detection

Nina Poerner, Hinrich Schütze


Abstract
We address the problem of Duplicate Question Detection (DQD) in low-resource domain-specific Community Question Answering forums. Our multi-view framework MV-DASE combines an ensemble of sentence encoders via Generalized Canonical Correlation Analysis, using unlabeled data only. In our experiments, the ensemble includes generic and domain-specific averaged word embeddings, domain-finetuned BERT and the Universal Sentence Encoder. We evaluate MV-DASE on the CQADupStack corpus and on additional low-resource Stack Exchange forums. Combining the strengths of different encoders, we significantly outperform BM25, all single-view systems as well as a recent supervised domain-adversarial DQD method.
Anthology ID:
D19-1173
Volume:
Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)
Month:
November
Year:
2019
Address:
Hong Kong, China
Venues:
EMNLP | IJCNLP
SIG:
SIGDAT
Publisher:
Association for Computational Linguistics
Note:
Pages:
1630–1641
Language:
URL:
https://aclanthology.org/D19-1173
DOI:
10.18653/v1/D19-1173
Bibkey:
Cite (ACL):
Nina Poerner and Hinrich Schütze. 2019. Multi-View Domain Adapted Sentence Embeddings for Low-Resource Unsupervised Duplicate Question Detection. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), pages 1630–1641, Hong Kong, China. Association for Computational Linguistics.
Cite (Informal):
Multi-View Domain Adapted Sentence Embeddings for Low-Resource Unsupervised Duplicate Question Detection (Poerner & Schütze, EMNLP-IJCNLP 2019)
Copy Citation:
PDF:
https://preview.aclanthology.org/emnlp-22-attachments/D19-1173.pdf
Attachment:
 D19-1173.Attachment.zip