Traditional computer-supported collaborative learning (CSCL) systems focus primarily on facilitating group work amongst students, providing several different modes for different types of communication, from chatting applications to shared whiteboards. The power of a good CSCL system, in addition to the interactions it supports, lies with the data the system can gather from those interactions (both relational and content-based). Thus, we have developed the Intelligent Multiagent Infrastructure for Distributed System in Education (I-MINDS), a CSCL system built on the Microsoft ConferenceXP platform that tracks real-time data to support student learning and classroom management. I-MINDS facilitates (1) student-student interactions through message dialogs, stored as trees, and (2) student-teacher interactions through question/answer pairs. I-MINDS logs these interactions in a MySQL database. Our VirtualClassroom analysis tool derives statistics from these interactions such as student responsiveness, degree of dialog involvement, amount of contribution, and quality of involvement and contribution. Then, an instructor can view overall trends for individual students and groups and detailed charts for all statistics for each student or all students for each statistic, which is especially helpful in large classrooms and in distance education. The instructor can also adjust a "highlight" thresh-old for each statistic such that any value failing to meet the threshold will be highlighted for easy observation. Our DatabaseSearch analysis tool allows students and teachers to search through recorded interactions and filter the results based on metadata such as time interval, class relevance, or keyword content.