LEARNED CONTEXT CORRELATION THROUGH NETWORK COMMUNICATION OBSERVATIONS

摘要

Establishing a correlation between a user's context and wireless signal observations from that context enables wireless computing devices to provide users with targeted information. Rather than performing an expensive calibration, mobile sensor data indicative of a user's context is bundled with wireless signal observations and transmitted to server computing devices that can obtain a context from the mobile sensor data, correlate that to the wireless signal observations, and store the same in a context/wireless correlations database. When other wireless computing devices request a context based on wireless signal observations made by those wireless computing devices, reference can be made to the database and a context correlated to those wireless signal observations is obtained and provided to them, enabling targeted information to be delivered to their users. To accommodate heterogeneity among wireless computing devices the database maintains correlations according to the wireless computing devices from which the data originated.