Fairness in Algorithmic Multi-disciplinary Team Formation
Athalage, Shehani (2023)
Athalage, Shehani
2023
Master's Programme in Computing Sciences
Informaatioteknologian ja viestinnän tiedekunta - Faculty of Information Technology and Communication Sciences
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2023-05-02
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-202304254293
https://urn.fi/URN:NBN:fi:tuni-202304254293
Tiivistelmä
Fairness in team creation is becoming an increasingly important subject of study in computer science and artificial intelligence. As algorithms increasingly automate decision-making processes, ensuring these systems are fair and unbiased has become a key concern. Team formation is one area of study where algorithms are used to match individuals with complementary talents and expertise to establish productive teams. When it comes to the process of forming teams, fairness is an essential component. Based on the relevant research, the thesis proposes the Rule-Based Expert Extraction Method and the Group-project distance and Unfairness Optimization Method to improve fairness during the team-formation process. Additionally, To assess the unfairness, the two proposed approaches are compared with the Pair-round selection method, which was previously examined by Machado and Stefanidis (2019). The fairness improvement is evaluated and compared.
Several metrics were taken to assess the refined performance in the team formation process to create a balanced and fair team. The primary goal was to increase fairness when forming multidisciplinary teams. In terms of promoting fairness, the results reveal that the Group-project distance and Unfairness Optimization Method and the Rule-Based Expert Extraction Method perform slightly better than the Pair-round choosing method. The Rule-Based Expert Extraction Method has the most significant Group-project distance, followed by the Group-project distance and Unfairness Optimization Method, and the Pair-rounds choosing method. However, the new approaches have improved fairness and mitigated the increased Group-project distance. Overall, the experimental evaluation demonstrates the potential of the Group-project distance and Unfairness Optimization method and the Rule-Based Expert Extraction method to improve fairness in team formation, which has significant consequences for businesses and organizations that rely on team collaboration.
Several metrics were taken to assess the refined performance in the team formation process to create a balanced and fair team. The primary goal was to increase fairness when forming multidisciplinary teams. In terms of promoting fairness, the results reveal that the Group-project distance and Unfairness Optimization Method and the Rule-Based Expert Extraction Method perform slightly better than the Pair-round choosing method. The Rule-Based Expert Extraction Method has the most significant Group-project distance, followed by the Group-project distance and Unfairness Optimization Method, and the Pair-rounds choosing method. However, the new approaches have improved fairness and mitigated the increased Group-project distance. Overall, the experimental evaluation demonstrates the potential of the Group-project distance and Unfairness Optimization method and the Rule-Based Expert Extraction method to improve fairness in team formation, which has significant consequences for businesses and organizations that rely on team collaboration.