Recommendation systems were applied in various fields, such as e-tailing, movies, books,…, and so on. Among them, tourism recommendation systems are also one of the widely research topics. Many tourism recommendation system studies use Collaborative Filtering (CF) method and try to add personality traits to the recommendation system methods to improve the precision. Zhou et al. (2007) suggested that Mass Diffusion (MD) method has more precision than CF method, but this method is mostly applied to recommending movie genres or books, but less often in tourism. Compared to other recommendation systems, fewer studies have considered personality traits such as Big Five Factor (BFF) and Myers-Briggs Type Indicator (MBTI) 16 (personality type). In this study, we used the MD method to establish a model of tourism attraction recommendation, and combined the personality traits commonly used in other recommendation system studies, such as BFF and MBTI 16, to achieve personalized recommendation of tourism attractions. According to the experimental results of this study, compared with the CF method combined with personality traits, the MD method combined with personality traits can recommend attractions to users more accurately.