Intelligent Question Answering System Based on Cognitive Graph: A Case Study of Textile Fabrics

With the continuous development of artificial intelligence technology, more and more natural language processing technologies are applied to the textile industry. How to extract information from massive textile data and give more targeted answers to users' questions is a vital issue in promoting textile intelligence. Therefore, taking textile fabrics as an example, this paper proposes an intelligent question-answering system based on cognitive graphs. Based on the knowledge of textile fabrics as the knowledge source, a cognitive graph question-answering system is constructed. Firstly, the Bidirectional Encoder Representation from Transformers-Bi-directional Long Short-Term Memory-Conditional Random Fields (BERT+BiLSTM+CRF) model is applied to perform entity recognition on the question sentence. Subsequently, node traversal is performed utilizing Cypher statements. Following that, answer reasoning is conducted employing the Graph Convolutional Network (GCN) algorithm, and the obtained answer is implemented as the output result. Finally, the system undergoes functional testing to verify its capability to accurately return the question's answer. The empirical results indicate the feasibility of the proposed cognitive graph-based question-answering system and its specific generalization and scalability.