User-Adaptive Ranking System on Mobile Trading Applications

Taeyoon Kim, Minyoung Bae, Hoonjin Jung, Youngeun Nam, Yooju Shin, Hyangsuk Min

Abstract: It becomes increasingly difficult for the stock trading app users to perceive useful information as the list of favorites grows and theinformation contained becomes complex. In this paper, we present a ranking system that rearranges the favorite items according to the user’s actionsusing a multi-armed bandit algorithm. Our solution ranks items according to the user’s interests and determines which features of each item arerelevant in making investment decisions. Users get an extra method for arranging the stocks, and this will be applicable to other features of a tradingapplication wherever stocks are presented in a list format.