Recommending Points-of-Interest (POIs) is surfacing in many location-based\napplications. The literature contains personalized and socialized POI\nrecommendation approaches which employ historical check-ins and social links to\nmake recommendations. However these systems still lack customizability\n(incorporating session-based user interactions with the system) and\ncontextuality (incorporating the situational context of the user), particularly\nin cold start situations, where nearly no user information is available. In\nthis paper, we propose LikeMind, a POI recommendation system which tackles the\nchallenges of cold start, customizability, contextuality, and explainability by\nexploiting look-alike groups mined in public POI datasets. LikeMind\nreformulates the problem of POI recommendation, as recommending explainable\nlook-alike groups (and their POIs) which are in line with user's interests.\nLikeMind frames the task of POI recommendation as an exploratory process where\nusers interact with the system by expressing their favorite POIs, and their\ninteractions impact the way look-alike groups are selected out. Moreover,\nLikeMind employs "mindsets", which capture actual situation and intent of the\nuser, and enforce the semantics of POI interestingness. In an extensive set of\nexperiments, we show the quality of our approach in recommending relevant\nlook-alike groups and their POIs, in terms of efficiency and effectiveness.\n