Michigan State University at the 2007 TREC ciQA Evaluation

rst participation in the ciQA task. Instead of exploring conversation strategies in question answering (3, 4), we decided to focus on simple interaction strategies using relevance feedback. In our view, the ciQA task is not designed to evaluate user initiative interaction strategies. Since NIST assessors act as users, the motivation to take an initiative is lacking. It is not clear how to encourage the assessors to take the initiative (e.g., by asking additional questions) during the interaction process. We feel in such a setting, relevance feedback or any kind of system initiative interaction strategies seem more appropriate. Therefore, we have focused on variations of relevance feedback in this year's evaluation. For the initial runs, our two submissions were based on two distinct approaches, a heuristic approach and a machine learning approach. Since only two interactive runs can be submitted for evaluation, we decided to focus on one aspect of variation. The only dierence between the two interactive and