Intelligent Response Assistant to Support End User Queries in Enterprises
Abstract - In this paper, we have proposed a new approach to the Question-answering framework for effective handling of user queries. Our approach includes Universal Sentence Encoder (USE) for semantic similarities with an SVM-based classifier model for detecting the cluster IDs from a large FAQ knowledge base. This approach makes the responses faster and the framework scalable. Based on our test results, the model provides approximately 90-95% accuracy for query responses.
Keywords - FAQ, Semantic Search, clustering, Question- Answering