Abstract: Knowledge practices mirroring is one of the central elements of the Trialogical Learning Approach as it is supposed to be a driving force in processes of practice transformation and knowledge creation. The exploitation of historical logging data holds promise to provide a great deal of information about group activities without requiring additional efforts for recording of events by the users. Building on the Trialogical Learning Approach as well as related work in the fields of Data Mining and Knowledge Discovery, Computer-Supported Collaborative Learning, and Information Visualization, this paper suggests high-level requirements for mirroring tools in support of practice transformation and introduces a software tool called Timeline-Based Analyzer (TLBA) that was designed and developed in response to these requirements. One of the main TLBA features is the possibility to define patterns as sequences of relevant actions that resulted into critical moments in investigated practices. Such kind of patterns might also represent conceptually interesting practices that emerged within a particular context - either being positive (a sort of good practices), or negative (bad practices). The usability of the whole analytical solution has been tested through the first iteration of several practical experiments and case studies. One of them is described in this paper and illustrates how TLBA can be used to support collaborative analysis and mirroring. The results of these evaluations have been used for continued improvement of the TLBA in order to provide a stable and intuitive tool not only for researchers, but for daily use of teachers or other users.