This paper mainly studies how to use gesture behavior generated by mobile terminal to mine user interest and build user interest profile. On the basis of this, the interest recommendation effect based on user interest profile is studied. By collecting the user′s gesture behavior data and the corresponding content of gesture, the user′s interest is mined and the user′s interest profile is constructed. Finally, recommendation experiments are conducted based on user interest profile. The results show that the interest profile construction based on user gesture behavior has achieved good results. The results of this study have important value in the field of personalized recommendation system and marketing. On the one hand, it can provide users with interesting content; on the other hand, it can improve user experience and increase user loyalty.