Attentive User-Engaged Adversarial Neural Network for Community Question Answering

Xie, Yuexiang; Shen, Ying; Li, Yaliang; Yang, Min; Lei, Kai

doi:10.1609/aaai.v34i05.6472

“…The current CQA task aims to automate the process of finding appropriate answers to questions in a community-created discussion forum [34]. Recent researches [50,53] often regard the CQA task as a text-matching problem and have proposed various deep learning networks to learn the semantic representation of question-answer pairs.…”

Section: Related Workmentioning

confidence: 99%

HeteroQA

Gao

¹

,

Zhang²,

Wang³

et al. 2022

Proceedings of the Fifteenth ACM International Conference on Web Search and Data Mining

4

0

View full text Add to dashboard Cite

Community Question Answering (CQA) is a well-defined task that can be used in many scenarios, such as E-Commerce and online user community for special interests. In these communities, users can post articles, give comment, raise a question and answer it. These data form the heterogeneous information sources where each information source have their own special structure and context (comments attached to an article or related question with answers). Most of the CQA methods only incorporate articles or Wikipedia to extract knowledge and answer the user's question. However, various types of information sources in the community are not fully explored by these CQA methods and these multiple information sources (MIS) can provide more related knowledge to user's questions. Thus, we propose a question-aware heterogeneous graph transformer to incorporate the MIS in the user community to automatically generate the answer. To evaluate our proposed method, we conduct the experiments on two datasets: MSM plus the modified version of benchmark dataset MS-MARCO and the AntQA dataset which is the first large-scale CQA dataset with four types of MIS 1 . Extensive experiments on two datasets show that our model outperforms all the baselines in terms of all the metrics.

show abstract

“…The enhancement part of the study (Lyu et al 2019) over the work (Zhao et al 2017) is attributed to the answer representation learning with latent user expertise and hierarchical attention mechanisms, as compared to the previous work that models answer embeddings independent on their authors. Moreover, AUANN (Xie et al 2020) employs generative adversarial networks (GANs) (Goodfellow et al 2014) to help to learn from user past answers for acquiring relevant user representations, which are then directly fed into score computation.…”

Section: Related Workmentioning

confidence: 99%

“…Thus recent efforts (Shen et al 2015;Qiu and Huang 2015;Fang et al 2016;Zhang et al 2017;Zhao et al 2017;Hu et al 2018;Lyu et al 2019) are devoted to deep representation learning approaches for measuring the matching degree between target questions and answers. In particular, a few recent studies (Zhao et al 2017;Hu et al 2018;Lyu et al 2019;Xie et al 2020) additionally build respondent representations based on their IDs or historical answers, hoping to reveal user expertise on answer quality. It is worth noting that the respondent role embodies the unique characteristics of CQA, as compared to general question answering.…”

Section: Introductionmentioning

confidence: 99%

Graph-Based Tri-Attention Network for Answer Ranking in CQA

Zhang¹,

Chen²,

Dong³

et al. 2021

Preprint

0

View full text Add to dashboard Cite

In community-based question answering (CQA) platforms, automatic answer ranking for a given question is critical for finding potentially popular answers in early times. The mainstream approaches learn to generate answer ranking scores based on the matching degree between question and answer representations as well as the influence of respondents. However, they encounter two main limitations: (1) Correlations between answers in the same question are often overlooked. (2) Question and respondent representations are built independently of specific answers before affecting answer representations. To address the limitations, we devise a novel graph-based tri-attention network, namely GTAN, which has two innovations. First, GTAN proposes to construct a graph for each question and learn answer correlations from each graph through graph neural networks (GNNs). Second, based on the representations learned from GNNs, an alternating tri-attention method is developed to alternatively build target-aware respondent representations, answerspecific question representations, and context-aware answer representations by attention computation. GTAN finally integrates the above representations to generate answer ranking scores. Experiments on three real-world CQA datasets demonstrate GTAN significantly outperforms state-of-the-art answer ranking methods, validating the rationality of the network architecture.

show abstract

“…The current CQA task aims to automate the process of finding appropriate answers to questions in a community-created discussion forum [33]. Recent researches [49,52] often regard the CQA task as a text-matching problem and have proposed various deep learning networks to learn the semantic representation of question-answer pairs.…”

Section: Related Workmentioning

confidence: 99%

HeteroQA: Learning towards Question-and-Answering through Multiple Information Sources via Heterogeneous Graph Modeling

Gao¹,

Zhang²,

Wang³

et al. 2021

Preprint

0

View full text Add to dashboard Cite

Community Question Answering (CQA) is a well-defined task that can be used in many scenarios, such as E-Commerce and online user community for special interests. In these communities, users can post articles, give comment, raise a question and answer it. These data form the heterogeneous information sources where each information source have their own special structure and context (comments attached to an article or related question with answers). Most of the CQA methods only incorporate articles or Wikipedia to extract knowledge and answer the user's question. However, various types of information sources in the community are not fully explored by these CQA methods and these multiple information sources (MIS) can provide more related knowledge to user's questions. Thus, we propose a question-aware heterogeneous graph transformer to incorporate the MIS in the user community to automatically generate the answer. To evaluate our proposed method, we conduct the experiments on two datasets: MSM plus the modified version of benchmark dataset MS-MARCO and the AntQA dataset which is the first large-scale CQA dataset with four types of MIS 1 . Extensive experiments on two datasets show that our model outperforms all the baselines in terms of all the metrics.

show abstract

Attentive User-Engaged Adversarial Neural Network for Community Question Answering

Cited by 12 publications

References 11 publications

HeteroQA

HeteroQA

Graph-Based Tri-Attention Network for Answer Ranking in CQA

HeteroQA: Learning towards Question-and-Answering through Multiple Information Sources via Heterogeneous Graph Modeling

Contact Info

Product

Resources

About