Alert button
Picture for Chen-Yu Hsu

Chen-Yu Hsu

Alert button

Unsupervised Multilingual Dense Retrieval via Generative Pseudo Labeling

Add code
Bookmark button
Alert button
Mar 06, 2024
Chao-Wei Huang, Chen-An Li, Tsu-Yuan Hsu, Chen-Yu Hsu, Yun-Nung Chen

Figure 1 for Unsupervised Multilingual Dense Retrieval via Generative Pseudo Labeling
Figure 2 for Unsupervised Multilingual Dense Retrieval via Generative Pseudo Labeling
Figure 3 for Unsupervised Multilingual Dense Retrieval via Generative Pseudo Labeling
Figure 4 for Unsupervised Multilingual Dense Retrieval via Generative Pseudo Labeling

Dense retrieval methods have demonstrated promising performance in multilingual information retrieval, where queries and documents can be in different languages. However, dense retrievers typically require a substantial amount of paired data, which poses even greater challenges in multilingual scenarios. This paper introduces UMR, an Unsupervised Multilingual dense Retriever trained without any paired data. Our approach leverages the sequence likelihood estimation capabilities of multilingual language models to acquire pseudo labels for training dense retrievers. We propose a two-stage framework which iteratively improves the performance of multilingual dense retrievers. Experimental results on two benchmark datasets show that UMR outperforms supervised baselines, showcasing the potential of training multilingual retrievers without paired data, thereby enhancing their practicality. Our source code, data, and models are publicly available at https://github.com/MiuLab/UMR

* Accepted to Findings of EACL 2024 
Viaarxiv icon
Github IconAccess Paper or Ask Questions
Twitter IconFacebook IconLinkedin IconWhatsapp IconMessenger Icon

CONVERSER: Few-Shot Conversational Dense Retrieval with Synthetic Data Generation

Add code
Bookmark button
Alert button
Sep 13, 2023
Chao-Wei Huang, Chen-Yu Hsu, Tsu-Yuan Hsu, Chen-An Li, Yun-Nung Chen

Figure 1 for CONVERSER: Few-Shot Conversational Dense Retrieval with Synthetic Data Generation
Figure 2 for CONVERSER: Few-Shot Conversational Dense Retrieval with Synthetic Data Generation
Figure 3 for CONVERSER: Few-Shot Conversational Dense Retrieval with Synthetic Data Generation
Figure 4 for CONVERSER: Few-Shot Conversational Dense Retrieval with Synthetic Data Generation

Conversational search provides a natural interface for information retrieval (IR). Recent approaches have demonstrated promising results in applying dense retrieval to conversational IR. However, training dense retrievers requires large amounts of in-domain paired data. This hinders the development of conversational dense retrievers, as abundant in-domain conversations are expensive to collect. In this paper, we propose CONVERSER, a framework for training conversational dense retrievers with at most 6 examples of in-domain dialogues. Specifically, we utilize the in-context learning capability of large language models to generate conversational queries given a passage in the retrieval corpus. Experimental results on conversational retrieval benchmarks OR-QuAC and TREC CAsT 19 show that the proposed CONVERSER achieves comparable performance to fully-supervised models, demonstrating the effectiveness of our proposed framework in few-shot conversational dense retrieval. All source code and generated datasets are available at https://github.com/MiuLab/CONVERSER

* Accepted to SIGDIAL 2023 
Viaarxiv icon
Github IconAccess Paper or Ask Questions
Twitter IconFacebook IconLinkedin IconWhatsapp IconMessenger Icon