Previous Thesis Descriptions

This page will list previous descriptions of my thesis topic. As the list becomes larger the format will probably have to change but for now they all live on this page to allow side-by-side comparison. The current version of my thesis description is on my main page.

9/10/97:

My thesis will study individual differences: how they can be determined using knowledge elicitation techniques, and how they can be used in developing a system (probably an intelligent interface). I am also interested in how understanding these differences can be used to benefit a team of users.

10/?/97:

My thesis will study individual differences: how differences in knowledge, ability, and cognitive style can be determined using knowledge elicitation techniques, and how they can be used in developing a system (either an expert system or an "intelligent" interface). Since most systems are used by more than a single user, I am interested in how a system can be designed to accomodate these differences yet still be useful for an entire team.

1/15/98:

The optimal thesis area would be one where I can find an overlap between my interests and those of my advisors. One area that may fit this description would be knowledge elicitation for design. I've been doing some background reading in this area but need to get a better idea of what is being done currently. I also need to find out what areas fit in best with my interests (and what those interests are...)

Some possibilities would be to extend an existing system already, use an existing system for a new application, use an existing system for a different problem solving method (?), or to build (prototype) something new. I'm not sure if any of these ideas make sense or are realistic. Plus I'm sure there ideas that I have not come up with yet.

3/11/98:

I have been investigating different KE techniques and how they might be used for design. In the process of doing this I have generated a matrix of techiques that groups them by category of interaction with the domain expert (interview, protocol analysis, construct elicitation etc.) and by type of data collected (procedures, strategy, classification of domain entities, etc.). There are several directions I am considering to turn this into a thesis topic:

1. Investigate how techniques or a combination of techniques could be used to extract sequencing/ordering for a design plan.
2. Take a type of knowledge or an elicitation method and investigate it for design.
3. Investigate coacting KE methods - how a combination of methods could be used, either sequentially or interleaved, to elicit a particular type of design knowledge.

I am currently leaning toward investigating how a combination of methods could be used to extract sequencing/ordering knowledge. There are direct methods that can be used to get this information by just asking for it but sometimes the designer may not be aware that (or why) they approach pieces of a problem in a specific order. I would like to investigate using a combination of indirect and direct methods where the indirect method is used to find when ordering is important and direct methods are used to determine why and how.