A description of a user is estimated based on the context of the user's travel pattern. A disclosed data-processing system continually receives data points originating from the user's cell phone, for example, wherein each data point represents a visit by the user to a particular geolocation, at a particular time. The system then partitions the received data points into groups, wherein the groups represent the different places that the user has visited. The system counts how many data points are in each group. Based on the number of data points in each group, as well as the number of groups, the system then determines the entropy of the received data points that correspond to the particular user, which is also regarded as the entropy of the user's travel pattern. The data-processing system then estimates a user description for the user, based at least in part on the determined user entropy.