Preliminary results of a curriculum adjuster based on professional network analysis

The topics related to the structure, evolution and usability of social and professional networks, drawing on knowledge from disciplines as diverse as sociology, mathematics, economics and physics, are relatively new in the area of computer science. Nevertheless, social/professional network data analysis (SNA/PNA) have already imposed a standard and in conjunction with a recommendation system can rank profiles based on skills, endorsements or recommendations from colleagues, and also analyse learning areas and provide potential corrections on learning curricula. We address in this paper the latter issue, of using data from professional networks for adjusting an engineering specialization curriculum. To this end, we built an adaptive recommender system that consists of several tools (modules) to create a data pipeline, perform data analysis tasks and provide recommendations for professors' council regarding the adjustment of the learning curriculum. The system includes an educational component with modules for course annotation, timeline generator based on learning curriculum and students' skills analysis, and a novel professional network analysis component, with modules to parse students' and graduates' profiles and extract their skills, creating a skills' timeline and identifying the most desired skills that should be (eventually more thoroughly) covered by the curricula. Through the automatic extraction of the set of skills covered by syllabi, compared with LinkedIn profile information for a significant percentage of our alumni (graduates) for the last 5 years, we were thus able to provide valuable feedback regarding the efficiency of our learning programmes.