User Emotion and Personality in Context-aware Travel Destination Recommendation

Abstract

At present, many recommendation systems focus on adapting psychological parameters, such as user emotion and personality, in recommendation algorithms in order to enhance the accuracy of recommendation lists according to user needs. Since users with different personalities tend to prefer items with different features, the personality of a user provides valuable information in exploiting personalized recommendations. Among personality models, the five-factor model appears to be suitable for applying recommendation systems, as it can be quantitatively measured. Moreover, emotion has also been widely adapted to a wide variety of recommendation domains, although few studies have examined tourist destination recommendation. In the present study, we use Plutchick’s emotion classification for emotion acquisition in the recommendation process. In addition, personality traits can be directly related to user emotions in decision making and exploring the relationship between emotion and personality in recommending items is also important. In order to incorporate user emotion in the recommendation process together with user personality, we propose a travel recommendation system that incorporated user personality and emotion. Thus, we compare and clarify the effectiveness of using emotion and personality in the recommendation process together with collaborative filtering techniques.