A dynamic skill-weighted matching algorithm achieves higher percentage of skills fulfilled than manual and random assignment across both evaluated semesters¶

Claim · All claims
Evidence · 1 study · design r2 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? The algorithm outperformed both manual and random assignment on Percentage of Skills Fulfilled in Fall 2023 and Spring 2024 capstone cohorts. → Brandon Pardi 2026

Evidence¶

Brandon Pardi 2026¶

Brandon Pardi, Garret Castro, Michael Pisman, Avash Adhikari, and Santosh Chandrasekhar. (2026). Improving Capstone Team Outcomes through Dynamic Skill Matching and Preference Alignment. https://arxiv.org/abs/2606.15572 doi:10.48550/arxiv.2606.15572

q2 · i? · design · r2

Comparison of three assignment scenarios (algorithm, instructor manual, random) on real UC Merced capstone data from Fall 2023 (16 projects, 68 students) and Spring 2024 (27 projects, 122 students). Table 2A reports the algorithm "outperforms both the manual and random assignments across both semesters" on skill coverage.

"Table 2A presents the averagePercentage of Skills Fulfilledfor the Fall 2023 and Spring 2024 semesters. Here, our algorithm outperforms both the manual and random assignments across both semesters."

Discussion¶

Related Claims¶

• Manual team formation required over 20 hours of instructor effort, while the algorithm produced assignments in milliseconds — related
• Manual assignment achieved the highest average student preference toward assigned projects, with the algorithm competitive and both substantially above random assignment — related
• Former capstone students' team-formation priorities centered on interesting topics, preferred teammates, and skill or career alignment — related
• LearnOpt's skill-weighted objective produces a modest but real reordering of recommended topics over a mastery-conditioned frequency baseline — related