Back to: Gabor Cselle Homepage

Organizing Email

Gabor Cselle, Master's Thesis, ETH Zurich


Email clients were not designed to handle the volume and variety of messages users are dealing with today. While this problem, termed “email overload,” is widely acknowledged, there exist only few products which attempt to solve it. This thesis discusses an email client extension aimed at sensibly structuring the user’s incoming email.

We developed our ideas after a thorough review of the problem and existing approaches of organizing email in both research and practice. In particular, we explored automatic foldering, in which machine learning algorithms are used to automatically move emails to folders. However, this method has several drawbacks which we discuss in detail. Instead, we devised a scheme for topic detection and tracking in email: Related emails are grouped and displayed together. The underlying inbox data is not changed. We evaluated our scheme using existing metrics for news articles, and found good results for clustering quality. We created an extension called "BuzzTrack" for Mozilla Thunderbird, a popular email client, which makes this technology accessible to the end user.



	author   = {Gabor Cselle},
	title    = {Organizing Email},
	year     = {2006},
	school   = {ETH Zurich}

November 19, 2006