Personalized Behavior-Powered Systems

Опубликовано 2 марта 2017, 21:57

I will present work that leverages user behavioral data to build personalized applications, which I call "behavior-powered systems". Two applications use online user interactions: 1) WebGazer uses interaction data made on any website to continuously calibrate a webcam-based eye tracker, so that users can manipulate any web page solely by looking. 2) Drafty tracks interactions with a detailed table of computer science professors to ask the crowd of readers to help keep structured data up-to-date by inferring their interests. And two applications use mobile sensing data: 3) SleepCoacher uses smartphone sensors to capture noise and movement data while people sleep to automatically generate recommendations about how to sleep better through a continuous cycle of mini-experiments. 4) Rewind uses passive location tracking on smartphones to recreate a person's past memory through a fusion of geolocation, street side imagery, and weather data. Together, these systems show how subtle footprints of user behavior collected remotely can reimagine the way we gaze at websites, improve our sleep, experience the past, and maintain changing data.

See more on this video at microsoft.com/en-us/research/v...