With the development and penetration of mobile devices, the opportunity for M-Commerce has opened. Compared with case-based reasoning and collaborative filtering, few take the cognitive style into account when designing personalized recommender system. Adomain and feedback-based FD/I calculator algo-rithmwas proposed to address the using condition issues and FD/I evaluation problems. Experimental study shows that the users' FD/I degree can be evaluated transparently and the quality of personal-izedrecommendationhas been promoted in terms of both contents and form of presentations.