摘要 |
Provided is a process of profiling a user of a mobile computing device, the process including: obtaining a location history of a user, the location history being based on signals from a mobile computing device of the user; obtaining a location-attribute score of a location identified in, or inferred from, the location history; determining, with a computer, a user-attribute score based on the location-attribute score; and storing the user-attribute score in a user-profile datastore. |
主权项 |
1. A method of profiling a user of a mobile computing device, the method comprising:
obtaining a location history of a user, the location history being based on signals from a mobile computing device of the user; obtaining a location-attribute score of a location identified in, or inferred from, the location history; determining, with a computer, a user-attribute score based on the location-attribute score; and storing the user-attribute score in a user-profile datastore, wherein:
the location history comprises an list of geolocation records, each geolocation record including geographic coordinates expressed as a latitude and longitude and a time at which the mobile computing device was at the respective coordinates, each geolocation record being obtained by an end-user portable device having access to a location identifying service;obtaining a location-attribute score comprises:
inferring locations between locations identified in the location history;for each identified or inferred location, retrieving respective tile records from a GIS, the tile records corresponding to a tile in which the respective location is disposed and adjacent tiles, each tile record corresponding to a geographic area of between 100 square meters and 100,000 square meters and being associated with one or more location-attribute scores, each location-attribute score corresponding to an activity of interest to advertisers and ordinal values indicative of a likelihood that a user is engaged in the respective activity in the tile during each of a plurality of time-bins, the time-bins defining different subsets of a week;determining a user-attribute score comprises:
determining that location attribute scores for the tile records for the time-bin in which the user was at the location are consistent among the adjacent tiles; andin response, determining a plurality of user-attribute scores corresponding the location-attribute scores, the respective user-attribute score being an average of the corresponding location-attribute score for a time-bin including the time at which the user was a the location and previous scores for the attribute from other locations;storing the user-attribute score comprises:
storing the averaged user-attribute scores in a user profile in the user-profile datastore, user profile being stored on a tangible, non-transitory, machine-readable medium, and the user-profile datastore being operative to respond to queries from advertisers for data relevant to the selection of advertisements;receiving a query for data relevant to the selection of advertisements; andresponding to the query based on responsive data stored in the user-profile datastore. |