Abstract:
I will describe ongoing work on a context-aware knowledge management system called I2I. I2I works by looking at the content of the user's document in word processing and web browsing applications. By comparing the content of documents multiple networked users are manipulating, I2I builds a conceptual space around the document that is separate from the physical (or hyper linked) space it occupies. I2I users viewing or creating documents can then be situated within this conceptual space and given access to:

1. neighboring users in that space
2. related documents from the Web or internal information repositories
3. relevant ongoing on-line conversations
4. captured expertise in the form of questions and answers
5. notes that other people have left there

I2I represents the next step in a new opportunistic knowledge management and information access framework we are developing. The system's goal is to provide users opportunities to access information and have conversations with people based on their shared interests, instead of requiring users have a strong goal in mind, know what information they are looking for, and make a deliberate effort to access that information. The result is that I2I users on the same "conceptual page" have access to relevant electronic documents, and a dynamic community of common interest without requiring any explicit intervention on their part.

In this talk, I will ground I2I's development in the context of previous systems we (and others) have built and continue to develop, and close with a brief description of our current efforts in extending this framework to information objects in the physical world.

About the speaker:
Jay Budzik is a Ph. D. student in the Computer Science Department at Northwestern University, where he has been a member of the Intelligent Information Laboratory for over four years. Jay's research has addressed a wide array of information access problems. At its core is the idea that human-machine communication can be improved by making the machine more aware of the user and her current task context. This has led to the development of task-embedded information access systems that are tightly coupled with everyday applications so they can proactively retrieve relevant information from distributed repositories, as well as glean information about the context of an explicit request. The Watson system he developed has been featured in the national press, and is used by tens of thousands of people worldwide.

Bringing Innovative Web Technologies to Library Settings - May 17, 2000