Learn2Mine: An Open-Source Cloud-Based Informatics Platform For Integrated Teaching and Data Exploration

Informatics platforms exist for a multitude of purposes and specialties (e.g., Taverna, Galaxy, RapidMiner). The development of such platforms is vital when considering the expanding role of Big Data. In concordance with the trend of programs reaching toward a open-source, cloud-based environment, we are developing an informatics and data mining platform that integrates teaching and data exploration through the development of gameful experiences. Learn2Mine is built upon the widely used open-source project Weka, unlike other informatics platforms, Learn2Mine utilizes a decompositionbased flow rather than a connected GUI workflow like that of RapidMiner, Galaxy, and Taverna. With this interface, focus is shifted towards discovery by offering levels of abstraction. The system is designed to build a user’s confidence in data science techniques intuitively and effectively by completing the feedback loop between application and domain expert. This develops a positive feedback mechanism through the integrated teaching environment, as students will be learning complex data mining techniques while earning achievements and being directed toward optimal solutions. Scientists are able to disable this feedback when a pure data analysis and exploration system is desired. Learn2Mine’s core data storage environment, built upon Google Drive, provides significant utility to multidisciplinary teams. Datasets, analyses, and customized algorithms will be able to be shared between collaborating scientists and with the data science community. The customization of templates allows for scientists who have a strong specialization in a field, such as bioinformatics, to develop a statistical tool that is much more meaningful to them rather than a geoinformatician. The system’s initial implementation and associated feedback integration is tested on its ability to analyze standard pedagogical datasets. The utility of this software will also be demonstrated through ongoing collaborations with the Air Force Research Laboratory, where datasets of human fatigue and toxicology will be analyzed.