Date of Award
Spring 2017
Document Type
Thesis
Degree Name
Master of Science (MS)
Department
Computing
First Advisor
Madiraju, Praveen
Second Advisor
Ahamed, Sheikh I.
Third Advisor
Kaczmarek, Thomas
Abstract
The interruption system is an application that prevents the user from noticing phone calls when he/she is busy, by turning off the ringtone. In a previous project, the user can enter his/her class and work schedule on Google Calendar. The intelligent interruption system can detect if the current time matches the range of one of the events in the user's Google Calendar. Other contexts considered were: driving, relationship of the callers, and proximity of Bluetooth devices. This project is a continuation of the interruption system. We consider additional context, social media such as Twitter. Research is done on when is the best time to turn off the ringer when the user is using Twitter. If the user is using social media, the user isn't as busy compared to, if the user is in class or at work. We further granularize social media activity such as reading messages, writing messages, and use these to help predict interruptions. We use the feedback provided by the user and employ machine learning approach which takes as input the different contexts and predicts if the user should be interrupted. We implemented a prototype application on Android operating system.