Filtering Graduate Courses based on LinkedIn Profiles

• Leandro Figueira Lessa PUC-Minas
• Wladmir Cardoso Brandão PUC-Minas

Resumo

People are overloaded with the everyday massive amount of information they consume. There are several options available for choice, from TV shows, books, traffic routes to graduate courses. In this scenario of multiple choices, the manual search and evaluation of all possibilities to make decisions is unfeasible. In the academic context, the HEIs (Higher Education Institutions) offer several graduate courses and, with so many options, students need mechanisms to choose relevant courses to their interest in order to reduce the dropout and financial loss risks. In this article, we propose a recommendation approach that filters graduate courses for students using their LinkedIn professional profiles. Experiments show that features based on competences and activity area are more effective than professional summary and experiences to recommend graduate courses within a content-based approach. In addition, our proposed approach performs recommendations with precision of up to 68.50% in the top-1 recommendation lists, achieving 100% coverage.

Palavras-chave: Content-based recommendation, course recommendation, LinkedIn