Department of Computer Science & Engineering

One way to improve inferences on sensor data is to tune the algorithms through an offline procedure. A potentially less expensive and more accurate method is to use an online procedure based on feedback from users, who often know best what the data means to them. We present a method for user-assisted location inference based on 802.11b wireless signal strengths. A user `corrects' system geolocations by clicking on a map, recording a `virtual access point' (VAP) at the selected point for future inferences. A best VAP is selected using simple criteria, including the VAP's creator. This permits using other's VAPs while getting their own if one exists, capturing user-specific behavior. Indoor experiments in our environment show over 10 times improvement in accuracy.

Pre-2018 CSE ID: CS2003-0765