An Architecture Model of Recommender System for Pedagogical Design Patterns

This work in progress research-to-practice paper proposes an architecture model for active learning design pattern retrieval based on the concept of conversational recommender systems (C-RS). In earlier work, we proposed an object-based pedagogical design pattern model to generate abstraction of multiple unique implementations of collaborative active learning practices. This model helped in formalizing the problem-solution pairs that the educators face in collaborative active learning such as students' engagement in teams and class activities, social skill improvement, and assessment issues. Based on this model multiple patterns in different categories were developed through a systematic pattern development cycle combined with deductive elicitation methods and workshops. In this study, we propose an architectural model for an interactive pattern recommendation system. We use NLP models to extract the context of the problem that each pattern addresses. The extracted aspects are combined with the dimensions of each pattern based on attributes of the developed pattern model. The proposed architectural model is based on the model, view controller (MVC) pattern. In this model, the user interacts with the system through the ‘view’ layer. The ‘model’ layer stores the pattern language in the form of a graph in which each node of the graph represents one pattern, and the edges denote the relationship between the patterns. The ‘controller’ handles text analysis by NLP algorithms based on user input. One of the main features of our object-based design pattern model is its emphasis on team attributes by which different dimensions of collaboration in engineering education are addressed. The implementation of the developed architectural model would enable the instructors to identify possible solutions to the pedagogical challenges they face in teaching engineering classes where teamwork is an integrated part of the learning process.