Content-based Stock Recommendation Using Smartphone Data

The number of investors holding risky assets in Japan is much lower than that western countries even though it an effective way for building investor assets. Although Japanese investment companies offer a service to invest points through coalition loyalty programs instead actual currency address this situation, the problem still persists. One reason novice do not know which stocks invest. possible solution recommending stocks; however, we face cold-start because there no transaction history investors. In study, propose novel content-based recommendation approach utilizes touchpoint information, e.g., payment and app usage data, on smartphones daily life. This employs user-weighted recency, frequency, monetary, called UW-RFM complementary module comply with guidelines prohibit presenting only small establishing minimum be presented. We conduct online evaluation validate effectiveness proposed service. results show motivates users more, i.e., 0.352 more clicks area 3, 016 (yen), baseline method does consider information.