Graph mining approach towards recommendation system using feedback

Success of a system is ensured with effective implementation of Feedback System. In academia the purpose of Feedback system varies with management intention as a controlling tool or evaluation of teaching-learning process. Feedback of instructor or trainer by students is a legitimate source to know about the performance of a faculty and consequently of college/ institute. The prime perspective of many institutes to conduct feedback from students and analyze it for improving performance of individual and organization´s quality education goal. This work presents an alternative schema using clustering based data analysis technique to analyze feedback data and as a basis for recommendation system. In the result, Two resultant clusters of faculties are formed, one includes all those who performed well and second includes those who didn´t. The resultant clusters of faculties with similar attributes are formed by generating Minimum Spanning Trees from Euclidian distance matrix, through faculty rates(rated by students) adjacency matrix. This method not only provide a new insight to management for decision making process but also suggests re-exploration of existing techniques. The result of analysis and implementation resembles with the actual samples.