Leveraging Recommender Systems to Enhance Student-Industry Alignment in Industrial Placements

This study explored the use of recommender systems to streamline student placement for industrial attachments. Traditional methods often restrict students to geographically limited options and a lack of industry knowledge. The research employed a mixed methodology, combining surveys and software development. A multi-criteria classifier named Placement-Ke was built, utilizing a hybrid collaborative filtering approach to create user profiles and recommend suitable companies. User profiles considered both individual interests and weighted past student ratings. Leveraging a pre-existing dataset and a pre-trained base model, the system achieved a commendable 92% accuracy. Parameters derived from the base model informed the development of a custom placement recommender system, implemented as a prototype application. User feedback gathered through surveys indicated a high satisfaction rate (77.78%). This research not only validates the need for automation in industrial placement, as identified in the literature review, but also demonstrates the effectiveness of the designed system in achieving both quantitative (accuracy) and qualitative (user satisfaction) objectives. The study recommends further development and implementation of Placement-Ke within higher education institutions.