A group activity recommendation scheme based on situation distance and user similarity clustering

Abstract

Activity recommendation is meant to help user to choose appropriate activity in daily tasks. In order to provide a suitable recommended activity while increasing users’ satisfaction, users’ preferences should be considered. However, given mobility of users, they can reach an unknown environment, where there may be no information regarding users’ preferences. Therefore, the recommend mechanisms which use predefined users’ preference may not be able to work when user goes to a completely new environment. To overcome this issue, we present a group activity recommendation mechanism that recommends activities to users based on users’ preferences which are inferred from their behavior in the past. Moreover, our mechanism also recommends the people who are suitable to join the activity with him.