MIT EECS - MediaTek Inc Undergraduate Research and Innovation Scholar
Smart iDiary: Inferring behavior of people according to their ‘backgrounds’ based on GPS
Daniela L. Rus
iDiary is a system that through a mobile application, algorithmically turns data into textual descriptions, allowing users to search their previous destinations with queries. Currently, iDiary streams live GPS data and compresses the input into a more manageable dataset on which real-time queries can be performed. We wish to extend iDiary to be able to categorize the data based on a user’s occupation/background, so that insights can be gained on user’s behavioral patterns based on their work. This kind of insight could be useful for services catered to people based on specific work.
More than a year of experience with Java and web development will allow me to present the findings on websites and Android apps. In terms of machine learning, I have taken Artificial Intelligence (6.034) and am taking Machine Learning (6.867). This background will allow me to grasp project concepts and have an idea of how the whole project should head.