What Matters in Hiring Professionals for Global Software Development? A SLR and NLP Criteria Clustering

• Published in: IEEE transactions on engineering management Vol. 71; pp. 6291 - 6318
• Main Authors: Santos, Erivelton Antonio dos, de Souza, Dalton Garcia Borges, da Silva, Carlos Eduardo Sanches
• Format: Journal Article
• Language: English
• Published: New York IEEE 2024; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Cluster; Clustering; Clustering algorithms; Companies; Context; Contracts; Criteria; criteria selection; Germanium; Gig economy; Globalization; hierarchical structure; Hiring; mind map; Natural language processing; NLP; Organizations; Professionals; sentence embeddings; Skilled workers; SLR; Software; Software algorithms; Software development; Software industry; Systematic review
• ISSN: 0018-9391; 1558-0040
• DOI: 10.1109/TEM.2023.3279769

Globalization stimulated a new era of Global Software Development (GSD), followed by the gig economy (GE) phenomenon, which jointly caused considerable transformations in software development markets, mainly after the recent supply chain disruptions. The cultural and geographic barriers have compelled numerous organizations to devise comprehensive digital technologies to overcome this situation. Likewise, the rising unemployment rates led the workforce into short-term contracts or to the on-demand market known as the gig economy. Together with the enhancement in global software development, the organizations found a direction to restore their activities. However, when organizations are immersed in fast-paced environments, selecting skilled professionals is difficult and risky, especially with a lack of qualified professionals. This article identifies the criteria for hiring professionals in the GSD or GE context and proposes a novel approach to clustering them. To do so, we collected the criteria from a broad subject through a systematic literature review, then applied natural language processing with the SBERT algorithm to get the sentence embeddings. Further, we cluster the criteria by applying the k -means algorithm. After that, we innovatively and responsively grouped the clusters formed by repeating the SBERT and k -means algorithms and created its mind map. Our findings disclosed 319 criteria and 6 cluster groups comprising a mind map hierarchical structure. Consequently, these outcomes have pedagogical implications to assist specialists from education institutions in designing new course domains. Such as, it can be helpful to practitioners to assist in hiring professional processes in the GSD or GE context.