Location-Sensitive Personalized Query Auto-Completion

Query auto-completion (QAC) suggests possible queries to users from the moment they start entering a query. This popular feature of modern search engines is thought to reduce physical and cognitive effort when typing a query. Meanwhile, geography is becoming increasingly important in Web search. Search engines can often return better results to users by analyzing features such as user location or geographic terms in Web pages and user queries. Despite of this, user's geographic search intent with QAC is under-studied. This paper begins to address this gap. We propose a hybrid QAC model that considers: query's forecasted frequency, user's previous submitted queries and user's location preference. Our experimental results on a real-world search log show that our location-sensitive personalized QAC model significantly outperforms two competitive baselines.