University of Maryland

Project Motivation – DICE

DICE (Detector of Innovation Communities and Ecosystems) is a system that allows users to identity, monitor, and model the communities and ecosystems associated with complex innovations. These communities consist of collections of organizations, people, and concepts that interact to form complex, dynamic innovation ecosystems.

DICE is a simple extendable platform that uses topic modeling, machine learning, and crowdsourcing techniques to enable users to specify reference terms of innovation, draw materials from diverse sources (including media repositories, academic paper repositories, and other public document collections), identify the organizations and individuals who comprise the innovation community. We hope to extend it further, allowing DICE to construct network diagrams and other visualizations of the innovation ecosystem.

DICE pursues problems that often require the aid of data; lots of data. The process of moving from a set of terms that we want to know about to producing networks that represent the innovation ecosystem has been time-consuming and fraught with problems. This tool aims to streamline the process, while, at the same time, making the process by which datasets are derived well-documented and easy to reproduce. Adding terms is as easy as writing those terms out to a file and setting the tool off running.