thoughtcrumbs

"Active learning" web browser at Penn State

Information Science professor Guangfeng Song at Penn State McKeesport is developing a browser that learns from its user's behavior and makes adjustments to improve its user's experience. The goal is not just to customize browsers to individuals, but to collect aggregate activity from multiple users to better inform future web design. Privacy issues aside, centralizing methods for assisting the user (breadcrumbs, visited link colors, saved search queries) within the browser removes some of the onus from individual site designers, each of whom has her own method for "helping" visitors. Browsers already offer a lot of this functionality: bookmarks, saved passwords, finishing typed URLs, all of which are fairly standardized across browsers, and people can decide which ones to use. Adding a level of machine learning has a lot of potential.