SBIR Phase I: Data Mining Tool for Assessing Computer Science Curricula

This Small Business Innovation Research Phase I project provides a tool to allow for computer science curricula modifications to better prepare computer science graduates for workforce demands. 69% of working software developers are self-taught and 43% utilized on-the-job training as their primary learning source. The United States is becoming increasingly dependent on technology to remain competitive on the global stage and our education system must evolve to more effectively train our next generation of technology workers. There were only 59,581 computer science graduates in 2015 compared to 527,169 open computing jobs. This shortfall is magnified considering the majority of graduates must teach themselves critical skills to become productive workers in private industry. This tool not only impacts higher education institutions and their graduates, but also coding academies, high schools, workforce development agencies, private companies, recruiters, and government bodies. There are 2,650 institutions offering computer science degrees in the United States and all are potential users of this tool. The market size for this tool is estimated to be over $400 million. The intellectual merit of this project is the development of a tool that can 1) assess what higher educational institutions should be teaching within computer science as demanded by private industry, 2) determine how effectively institutions are currently teaching these skills, and 3) benchmark these institutions compared to their peers. The innovation inherent in this project is a real-time measurement of skills demanded in the private sector, rather than lagging years behind, and a real-time score of how well institutions are currently performing. The data provided by this tool will allow institutions to modify curricula at a significantly faster pace and dramatically increase the productivity of graduates. The tool proposed will be based upon an aggregation of millions of data points from disparate sources such as university catalogs, job postings, and resumes. Raw data is then cleansed and analyzed to produce actionable insights for end users in a visual presentation layer, tailored to individual institutions.