VIGHUB: a Technology Forecasting Tool based on Mining Software Repositories

Abstract

Introduction: Academics, developers, and companies focused on technological development seek to know what exists and what is still missing in this field. One of the ways they use is the review of bibliographic sources (state-of-the art). In this sense, a tool was developed that allows the current state to be identified semi-automatically. Objective: This article proposes a tool that extracts information from repositories hosted on GitHub. It analyzes the data using computational techniques and presents the results through visualizations that identify the field's technological evolution studied through the most used programming languages, central repositories, and organizations. Method: A model based on Mining Software Repositories (MSR) is used, which integrates an architecture based on microservices, using different programming languages, which allowed the construction of the VigHub tool. The model focuses on four aspects: Selection of a topic, extraction of the data source, analysis of information using computational techniques, and finally, the results are communicated through visualizations. Results: The VigHub tool was available online to carry out 3 case studies. The first in the academy, where technologies, programming languages, users, and companies interested in developing VLE's (Virtual Learning Environment) were identified from 2011 to 2021. The second and third were carried out by companies (industrial environment), which stated that using the VigHub tool supports data analysis and valuable results identification. Conclusions: A tool that allows identifying a part of the current state of technology could be a helpful tool for academics, developers, and companies, saving human resources, time, and possible repeated developments---code reuse. The VigHub tool aims to support the construction of state-of-the-art. Its results are complementary to the traditional method.