earticle

영어

This paper proposes a recommender system based on periodicity analysis from purchase history data using mining sequential pattern with time-series and FRAT (Frequency, Regency, Amount and Type of merchandise or service) analysis. In this paper, not used user's profile for rating, it is necessary for us to make the task of mining sequential pattern with time-series and clustering of category with weight using FRAT analysis in order to recommend service by frequently changing trends of purchase pattern. For doing that, it is considered the importance of type of merchandise or service and then, we suggest recommending method based on periodicity analysis to reflect frequently changing trends of seasonable purchase pattern for the four seasons as time goes by and as often as customers need different merchandises on e-commerce being extremely diverse. We evaluated the proposing system on the data set collected in the same dataset collected on a cosmetic internet shop to measure its performance. We reported some of the experimental result, in the same condition, which is better performance than both the previous system and existing system.