A Conceptual Model and web platform for discovering digital heroes from stack overflow developer community forum using a Data Mining approach

Abstract

Heroes are crucial in our general life. They inspire us to do good deeds. Likewise, in the digital world people are helping others using online community forums even without knowing the people they are helping. They are referred to as digital heroes and their efforts enrich the community forums and digital world day by day. The available literature does not reveal an existing platform for analyzing people’s digital activities and identifying digital heroes in the digital world. This study aimed to establish digital hero selection criteria based on community forum activities using data mining approach and develop a web platform. The platform will aid in discovering and documenting digital heroes so that future generations can know about them and be inspired by their contributions to the digital community as well as be motivated to do good deeds. Using homogeneous purposive sampling technique, this study used a public dataset of 1,889,860 users between 2008 and 2017 from the Stack Overflow online developer community forum and analyzed using data mining approach in order to discover the digital heroes based on their digital activities and proposed selection criteria. The data mined was mainly user’s information including user name, registration date, about, user id, total votes (up and down) and reputation number. The collected information was analyzed based on four proposed hero selection criteria for identifying digital heroes: Experience, Accuracy, Activity and Trust, which were established by collecting experts’ opinion using questionnaires. In order to establish digital hero selection criteria, this study used convenience sampling technique to target 45 experts who know about the Stack Overflow community forum, understand how it works, and know about developer community members and their activities in the forum. Furthermore, this study developed a web platform referred to as Digital Hero Discovery (DHD) web platform that was guided by conceptual data analysis model to analyze, discover and document digital heroes. The study was able to discover a total number of 3,231 digital heroes who qualified in all the criteria using the Knowledge Discovery in Databases (KDD) process of data mining approach to analyze data as well as hero selection criteria. The digital heroes were classified into four proposed hero categories: A, B, C and D based on their contributions in the community. From the findings, the categories had the following number of heroes: A – 1; B – 45; C – 2,411; and D – 774. The proposed digital hero selection criteria and data mining approach would be the start of the journey of digital hero discovery to inspire digital world people to help others accurately and frequently. The developed DHD web platform could be used in other digital communities including academic, medical and social communities in order to discover digital heroes based on their contributions