Investigating academic researchers' perceptions of a recommender system for mentor-to-mentee matching

Abstract

The importance of mentoring among researchers cannot be overemphasized, and to this effect, several recommender systems have been presented in the literature to find experts for collaborations. However, matching researchers as mentors to their respective mentees has not been looked into. Earlier works suggest that expert-finding recommender systems primarily use metadata from experts' publications for making recommendations. Meanwhile, when matching for mentorship, more emphasis should be placed on the areas and major areas of both the mentors and mentees as intended in our proposed system. It is observed that many authors co-author in some areas in which they are not major, which may not certify them as experts in such areas. This research attempts to investigate the views of researchers about mentoring and also to determine the perceptions of the researchers on the need to develop a recommender system that can match researchers for mentorship. The study was conducted in two universities in south-western Nigeria, namely; Federal University of Technology, Akure, and Federal University, Oye-Ekiti. Data were collected with the aid of structured questionnaires using both online and paper surveys. The collected data were analyzed using frequency tables and charts. The results show that 99% strongly agreed and 1% agreed that mentorship is essential among researchers in academia. 98% strongly agreed and 2% agreed that there is need for tools that can help users get personalized items in all areas of life. 56% strongly agreed, 42% agreed, and 2% were indifferent that building a recommender system for matching researchers for mentorship is important. As can be seen from the results, the majority of the respondents acknowledged the importance of mentorship among academic researchers and the need to develop a recommender system to match these researchers for mentorship.