Negative-Unlabeled Tensor Factorization for Location Category Inference from Inaccurate Mobility Data

Abstract

Identifying significant location categories visited by mobile phone users is the key to a variety of applications. This is an extremely challenging task due to the possible deviation between the estimated location coordinate and the actual location, which could be on the order of kilometers. Using the collected location coordinate as the center and its associated location error as the radius, we can draw a location uncertainty circle that may cover multiple location categories, especially in densely populated areas. To estimate the actual location category more precisely, we propose a novel tensor factorization framework, through several key observations including the intrinsic correlations between users, to infer the most likely location categories within the location uncertainty circle. In addition, the proposed algorithm can also predict where users are even when there is no location update. In order to efficiently solve the proposed framework, we propose a parameter-free and scalable optimization algorithm by effectively exploring the sparse and low-rank structure of the tensor. Our empirical studies show that the proposed algorithm is both efficient and effective: it can solve problems with millions of users and billions of location updates, and also provides superior prediction accuracies on real-world location updates and check-in data sets.