This project develops a new computational and educational resource, the Distributome, for exploring, discovering and interacting with varieties of probability distributions. The Distributome project leverages the results of successful NSF projects that have been sustained over decades of work in developing interactive learning materials, forging technological advances, and building and sustaining digital libraries; all integrated with an effective dissemination and professional development infrastructure to ensure on-going use. There are several novel features of this project. This resource builds the infrastructure for community based development, expands and validates the distributions' meta-data that is stored, processed, searched, traversed and updated by experts, learners, and educators. The Distributome provides a graphical user interface for interactive exploration of diverse distribution resources, as well as a web-service for query, discovery and computational utilization of these distribution resources by other software programs and tools.
Specifically, this project provides an open (development and utilization), platform-agnostic, extensible and broad framework for navigation, discovery and usage of probability distributions in diverse applications. The entire framework is built using XML/JAVA/HTML/Wiki/MathML/LaTeX and is freely made available to the entire community via www.Distributome.org. The user-base of the Distributome infrastructure includes both educators (integrating these graphical tools and instructional materials in their course curricula and participating in a unique virtual community led by a cadre of activists) and most importantly learners (exploring, validating and understanding the use of probability distributions and models for practical problem solving). Probability modeling is at the root of solving driving biological, engineering, health, physical science, and social problems fundamental to the modern STEM curriculum. The Distributome infrastructure enables representation, demonstration, computation and visualization of a large number of probability distributions, their interrelations and their applications integrated with associated class and out-of-class activities to advance learning.
Summary of the outcomes The table below includes a cumulative summary of the Distributome Project Outcomes (2010-2014) Cumulative Summary of Distributome Developments (2010-2014) Resource Type Examples of Distributome Resources Infrastructure Deployed the Distributome server www.Distributome.org Introduced the Distributome Blog www.Distributome.org/blog Introduced the Distributome Game www.distributome.org/V3/DistributomeGame.html Developed open-source Java, HTML and JavaScript library for Navigating the Distributome and exploring distribution properties and inter-distribution relations Developed an HTML Distributome Carousel www.distributome.org/tools.html Introduced Distributome Preferences distributome.org/V3/data/Distributome.xml.Book.pref Computational Tools Developed a large collection of Java and JavaScript probability distribution calculators Designed Simulators and Experiments for different probability distributions Developed the Distributome Navigator www.distributome.org/V3 Instructional Modules Introduced 31 new instructional modules demonstrating the classroom use of the Distributome resources for a wide spectrum of applications Developed 50+ examples of problems and corresponding probability distribution models www.distributome.org/V3/data/DistributomeGame_ProblemExamples.csv Training & Dissemination Validated the Distributome infrastructure in 21 courses at UMich, UAH, OSU, UCLA Trained 17 undergraduate and 5 graduate students at UMich, UAH, OSU, UCLA Organized 8 Distributome Continuing Education Workshops and dissemination events Published 6 peer-reviewed scientific publications reporting on our theoretical, educational and applied R&D efforts Intellectual merit outcomes We have demonstrated how advanced scientific concepts, powerful computational libraries and driving motivational challenges can be integrated into a framework (Distributome) that enabled the open-scientific development and collaborative validation of resources. Our training activities go beyond exposing learners to mathematical models and software tools. We engage students in the design, implementation, validation and distribution of Distributome resources. The interactive Distributome calculators, simulators and experiments provide instructors with resources for engaging students into the process of formulating conjectures, testing tools and confirming properties or characteristics of different models (e.g., probability distributions). Distributome web-infrastructure utilizes advanced and modern information and communication technologies to enhance the learning experience of students and enable instructors to customize or enhance their course materials. Broader impact outcomes Some Distributome activities involve popular science, important current events or interesting biomedical challenges (e.g., Ohio Turnpike, Colorblindness and other Distributome Activities). We aim to deliver advanced scientific concepts and computational analytics to the general population, as well as translate scientific progress to broad audiences. These resources have been accessed by over 33,000 learners and instructors. Anonymous evaluations of the Distributome resources, including our hands-on dissemination events, indicate that the broad Probability community values these tools, materials and services.
