• Dr. Kamal Sutaria
• kruti khalpada

Answer selection is one of the popular tools and has gained an increased research interest in extracting the quality answers in Community Question Answers (CQA) community. The aim of the QA is to retrieve the best answers for the user’s query from the large collection of documents from the database. The CQA services allow the users to post questions about their interest and other users share their opinions with their knowledge. Traditional approaches on deep learning allow the network to capture a part of the semantic relationship but ignore the rich feature learning which limits the quality answer prediction. To overcome the certain limitation in those models, we propose a new approach named multi-scale stacked dilated convolution (Ms-SDC) with the attention-based deep bidirectional long short term memory (Att-deep BiDLSTM) to estimate the quality answers for the input query. This model tackles the problems of: i) high-level or rich feature representation learning, ii) quality answer prediction from the set of candidate answers, iii) modeling the complicate relationship among question-answer and answer-answer. The process is classified into four stages: embedding, feature learning, retrieval and classification. In this, a dilated convolution is performed that recognizes the high-quality answers for the user query. Also, the attention-based networks allow the query to process on a certain part of the candidate answers. This allows the architecture to exploits important semantic and the syntactic features among the QA pair to determine the high-quality answers. Experimental analysis is conducted over semEval dataset and evaluated in terms of several classification metrics such as recall, precision, f1-score and accuracy measures. The performance of the proposed network outperforms and achieves a better improvement than the state-of-art techniques.

Answer selection, deep neural network, community question answering, answer ranking

"A Comprehensive Analysis for Quality Answer Selection in Community Question Answering", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.9, Issue 4, page no.f478-f483, April-2022, Available :http://www.jetir.org/papers/JETIR2204562.pdf

"A Comprehensive Analysis for Quality Answer Selection in Community Question Answering", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.9, Issue 4, page no. ppf478-f483, April-2022, Available at : http://www.jetir.org/papers/JETIR2204562.pdf