Current methods of credentialing and assessment for the purposes of job and career placement are either too coarse or too granular. The bachelor's degree is one of the few universally recognized credentials but provides prospective employers little information regarding the fit of a particular candidate to a position the employer has available. Certification exams tell employers a candidate has a specific skill set, but most skills and abilities are not covered (and likely never will be covered) by such exams.
This project explores the use of machine learning and similar strategies to develop alternative credentialing and assessment methods to help improve the fit between employers and prospective employees. Our initial approach focuses on the analysis of eportfolios.