NATURAL LANGUAGE PROCESSING FOR QUESTION-ANSWERING SYSTEM

Chalermpol  Tapsai; Wilailuk  Rakbumrung

Chalermpol Tapsai College of Innovation and Management, Suan Sunandha Rajabhat University, Thailand
Wilailuk Rakbumrung College of Innovation and Management, Suan Sunandha Rajabhat University, Thailand

Keywords: Natural language processing, Question-Answering, Deep learning

Abstract

Information is essential to create understanding and knowledge among stakeholders for organization management. The Question-Answering system is an important service that provides accurate information for users and stakeholders that can drive the organization to achieve its objectives. This research study on Natural language processing to analyze question sentences and predict the appropriate answer for each question using the Deep learning technique. The experiment was conducted to compare the prediction results between the analysis with and without data pre-processing. The results show that the accuracy of predictions using pre-processed data is higher than raw data. In addition, some techniques, such as ontology and similarity, can be applied to infer the meaning of words and sentences, which improves the results and presents more precise predictions.