Personal Profile Mining Based on Mobile Phone Location Data

Objective Understanding personal profiles like preferences,income levels,and geographical areas isthe basis of providing aperson with personalized and accurate services.In order to acquire personalprofiles we propose a reasonable technical route that first extracts the geographic regions from person-al mobile phone location data based on a density-based clustering algorithm.Then,the geographic re-gions are tagged with semantic meaning and we analyze house descriptions by NLP(Natural LanguageProcessing).A division method for people’s daily time is given,based on the assumed the activitypatterns of people.At last,an individual’s taste for something or his income levels is analyzed usingthe statistics for POIs and house prices in the extracted places.An experiment with real data showsthat this method is an effective solution to mining personal profiles.