A knowledge extraction system from online reviews using fuzzy logic

Currently, the most common method that tourists make decisions on selecting their preferred hotels is a data analysis from online reviews of experienced tourists to support the decision. Unfortunately, tourists must read enormous reviews in order to select the preferred hotels. This paper presents the design and development of an automatic knowledge extraction system from online hotel reviews of tourists. The proposed system focuses on a semantic analysis from hotel reviews using the ontology as a knowledge base for hotel information extraction and storing, and using the parser to solve the semantic analysis problem. In addition, we propose a fuzzy-logic based method for calculating a tourist satisfaction level with hotel services and facilities. The experimental results indicate that the semantic analysis method achieves with 88.91% of accuracy, 100% of precision and 73.56% of recall. While the satisfaction level calculation method achieves with 0.366 of MAE and 0.464 of RMSE.