Hybrid Visualisation Recommender System of Patient-Generated Complex Diabetes Data

This research introduces MTViz, a novel hybrid visualisation recommender system leveraging machine learning models to predict data analysis tasks and rule-based heuristics to generate tailored visualisations from diverse multidimensional diabetes datasets. MTViz allows users to visualise data without requiring specific visualisation knowledge. MTViz supports the insertion and simultaneous display of multiple patient-generated datasets, benefiting practitioners in extracting essential insights from various sources. Incorporating data analysis tasks in MTViz ensures the recommended visualisations are aligned with user-defined tasks, enhancing task-solving efficiency. Furthermore, MTViz organises visualisations using juxtaposed and superimposed views, assisting practitioners in comparing multiple health datasets within a single view.