Weather is a big factor in tourist decisions, andcertain places or events aren’t even recommendedduring dangerously bad weather. It is difficult to providea better recommendation to a group of tourists in thesecircumstances. We propose gTravel, a weather assistantframework that predicts weather in specified pointsof interest for a group of tourists. We demonstratehow prior knowledge of climatic patterns at a POI,as well as prior insights into how visitors rank theirdestinations in a variety of weather conditions, can helpimprove choice reliability. According to our findings, therecommendations are significantly more valid, and therecommended remedy is more comfortable.

POI, tourist, weather, recommendation, interest