Collaborative learning in professional development requires dynamic coordination between cognitive engagement and affective regulation. This talk presents emerging findings from an analysis of teacher discussion forums designed to support and extend professional learning community (PLC) meetings. Using a dataset comprising over one million tokens of teacher discussion, the preliminary results shared here integrate natural language processing (NLP) with measure of engagement, including coach-provided ratings and platform clickstream data, to examine affective shifts, discourse cohesion, and thematic patterns, we explore how teachers sustain participation, regulate emotions, and align cognitively in collaborative learning settings. These insights contribute to process-oriented models of teacher development and highlight new opportunities for using NLP to support AI-enhanced professional learning.
