A Text Classification Based Method for Context Extraction from Online Reviews

Recommender systems are systems that filter information depending on users' profiles and suggest items that might match their preferences. While the majority of existing researches compute recommendation by considering only users and items, Context Aware Recommendation Systems (CARS) consider, in addition to users and items, others features related to the context. A first issue in CARS studies is to identify the contextual features. In this paper, we investigate the use of Text Classification techniques to extract contextual features from users' reviews. We conduct experiments to identify the best classification algorithm for our dataset. We evaluate our approach on hotel reviews. We focus on extracting the trip type, as contextual information, from these reviews. Results show that the Multinomial Naive Bayes performs best in our dataset, with a F1 score of 60.1%. Since contextual information are not always provided in the reviews, we think that our results are promising. We conclude that this research area needs deeper studies.