Maximizing diversity within and among teams in a large-scale project

Area of Science:

• Operations Research
• Computer Science
• Social Sciences

Background:

• Previous methodologies for team assignment focused solely on within-team diversity.
• This led to repetitive combinations of conflicting interests across different teams.
• A government agency required a method to generate more diverse team combinations.

Purpose of the Study:

• To enhance personnel assignment methodology for three-member teams.
• To maximize diversity not only within teams but also among different teams.
• To address the need for a greater variety of interest group combinations in teams.

Main Methods:

• Development of an integer linear programming (ILP) algorithm.
• Incorporation of penalty costs for deviations from project requirements.
• Implementation of upper bound constraints on the frequency of specific team assignments.

Main Results:

• The algorithm generates thousands of diverse three-member teams.
• It effectively represents the conflicting interests of various groups.
• Maximal diversity within and among teams is achieved.
• The algorithm demonstrates efficiency, scalability, and fast convergence.

Conclusions:

• The upgraded ILP algorithm successfully boosts diversity among teams by maximizing between-team differences.
• It maintains maximal within-team diversity while ensuring a wider range of team compositions.
• The method is efficient, scalable, and converges rapidly, providing a practical solution for complex team formation.