Of Online Community: Clustering Group of Compatible Mentor and Mentee
Pratya Nuankaew*, and Punnarumol Temdee
Abstract: Clustering group of compatible mentor and mentee in online community is challenging because compaibility can be measured in many different aspects. This paper proposes clustering method of compatible mentors and mentees in online community based on the assumption that the mentors and the mentees have both common and compatible different attributes. This paper consists of three phases including data gathering, clustering, and member indentifying. The k-Means clustering method is used in this paper. The effective k value is determined by adapting the Elbow method with the proposed idea of this paper which is the most suitable k value is the one decreasing less than half of average within centroid distances. The results from 32 mentors and 205 mentees from Mae Fah Luang University, University of Phayao, Chiang Rai Rajabhat University, and Rajabhat Maha Sarakham University show that k=4 is the most suitable numbers of clusters. Additionally, k-Means clustering method is suitable for clustering groups of compatible mentors and mentees in this paper.
Keywords: Mentoring; Online Comminity; Mentor; Mentee; Clustering; k-Means Clustering.