Cluster and discover services in the Smart Campus platform for online programming contests

Abstract

Nowadays we can see and feel the technological change around us. Most of the tools are extended with online functionality, new online collaboration environments are appearing and bring along wide range of data for collection. Following this shift with the appearance of cognitive infocommunication we established a framework and created a usage scenario as an online system for supporting programmers in our Smart Campus platform. When we created the platform, extendibility and adaptivity were the primary goals to stand for efficient data collection through crowdsourced collectors. This paper is the continuation of our work and describes our online programming contest environment where services are applied to help the users solving assessments and grading them. The collected data help us to discover hidden correspondences and fine tune our problem suggestion service. While we continuously collect data, the applied clustering algorithms could help us to understand the problems arriving through the students' learning path and along these we could achieve a much more fine-grained solution for our services.