Context-aware Answer Selection in Community Question Answering Exploiting Spatial Temporal Bidirectional Long Short-Term Memory

Community Question Answering (CQA) sites provide knowledge sharing facility as the users can post questions and other share their answers. The selection of top-quality answers from set in a thread is significant challenging task Natural Language Processing (NLP). To address this issue, we propose deep learning based spatial temporal Bidirectional Long Short-Term Memory (Bi-LSTM) algorithm. existing studies mainly focus only computing semantic similarity between using votes given by users. proposed hybrid approach, on both forward backward, consider question to answer similarity. LSTM captures impact estimate relevancy, whereas backward learns features with predict best quality answer. Moreover, Bi-LSTM past future dependencies for better understanding context improve effectiveness selection. For extracting meaningful information noisy text data, data preprocessed following standard steps such tokenization, parsing, lemmatization, stop words removal, part speech tagging entities extraction. Word embeddings-based Paragraph vector (par2vec) has additional input nodes represent paragraph understanding. empirical analysis carried out SemEval CQA dataset shows that model outperforms state-of-art approaches.