Supervised attention for answer selection in community question answering

Thanh Thi Ha (Hanoi University of Science and Technology)
Atsuhiro Takasu (National Institute of Informatics)
Thanh Chinh Nguyen (ThaiNguyen University of Information and Communication Technology)
Kiem Hieu Nguyen (ThaiNguyen University of Information and Communication Technology)
Van Nha Nguyen (ThaiNguyen University of Information and Communication Technology)
Kim Anh Nguyen (ThaiNguyen University of Information and Communication Technology)
Son Giang Tran (University of Science and Technology of Hanoi)

IAES International Journal of Artificial Intelligence (IJ-AI)
Vol 9, No 2: June 2020

Answer selection is an important task in Community Question Answering (CQA). In recent years, attention-based neural networks have been extensively studied in various natural language processing problems, including question answering. This paper explores matchLSTM for answer selection in CQA. A lexical gap in CQA is more challenging as questions and answers typical contain multiple sentences, irrelevant information, and noisy expressions. In our investigation, word-by-word attention in the original model does not work well on social question-answer pairs. We propose integrating supervised attention into matchLSTM. Specifically, we leverage lexical-semantic from external to guide the learning of attention weights for question-answer pairs. The proposed model learns more meaningful attention that allows performing better than the basic model. Our performance is among the top on SemEval datasets.