Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Shih, Min-Han, Chung, Ho-Lam, Pai, Yu-Chi, Hsu, Ming-Hao, Lin, Guan-Ting, Li, Shang-Wen, Lee, Hung-yi
Format: Preprint
Veröffentlicht: 2023
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2312.09781
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866916446766891008
author Shih, Min-Han
Chung, Ho-Lam
Pai, Yu-Chi
Hsu, Ming-Hao
Lin, Guan-Ting
Li, Shang-Wen
Lee, Hung-yi
author_facet Shih, Min-Han
Chung, Ho-Lam
Pai, Yu-Chi
Hsu, Ming-Hao
Lin, Guan-Ting
Li, Shang-Wen
Lee, Hung-yi
contents In recent advancements in spoken question answering (QA), end-to-end models have made significant strides. However, previous research has primarily focused on extractive span selection. While this extractive-based approach is effective when answers are present directly within the input, it falls short in addressing abstractive questions, where answers are not directly extracted but inferred from the given information. To bridge this gap, we introduce the first end-to-end Generative Spoken Question Answering (GSQA) model that empowers the system to engage in abstractive reasoning. The challenge in training our GSQA model lies in the absence of a spoken abstractive QA dataset. We propose using text models for initialization and leveraging the extractive QA dataset to transfer knowledge from the text generative model to the spoken generative model. Experimental results indicate that our model surpasses the previous extractive model by 3% on extractive QA datasets. Furthermore, the GSQA model has only been fine-tuned on the spoken extractive QA dataset. Despite not having seen any spoken abstractive QA data, it can still closely match the performance of the cascade model. In conclusion, our GSQA model shows the potential to generalize to a broad spectrum of questions, thus further expanding the spoken question answering capabilities of abstractive QA. Our code is available at https://voidful.github.io/GSQA
format Preprint
id arxiv_https___arxiv_org_abs_2312_09781
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle GSQA: An End-to-End Model for Generative Spoken Question Answering
Shih, Min-Han
Chung, Ho-Lam
Pai, Yu-Chi
Hsu, Ming-Hao
Lin, Guan-Ting
Li, Shang-Wen
Lee, Hung-yi
Computation and Language
Artificial Intelligence
In recent advancements in spoken question answering (QA), end-to-end models have made significant strides. However, previous research has primarily focused on extractive span selection. While this extractive-based approach is effective when answers are present directly within the input, it falls short in addressing abstractive questions, where answers are not directly extracted but inferred from the given information. To bridge this gap, we introduce the first end-to-end Generative Spoken Question Answering (GSQA) model that empowers the system to engage in abstractive reasoning. The challenge in training our GSQA model lies in the absence of a spoken abstractive QA dataset. We propose using text models for initialization and leveraging the extractive QA dataset to transfer knowledge from the text generative model to the spoken generative model. Experimental results indicate that our model surpasses the previous extractive model by 3% on extractive QA datasets. Furthermore, the GSQA model has only been fine-tuned on the spoken extractive QA dataset. Despite not having seen any spoken abstractive QA data, it can still closely match the performance of the cascade model. In conclusion, our GSQA model shows the potential to generalize to a broad spectrum of questions, thus further expanding the spoken question answering capabilities of abstractive QA. Our code is available at https://voidful.github.io/GSQA
title GSQA: An End-to-End Model for Generative Spoken Question Answering
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2312.09781