As organizations and platforms increasingly rely on expert knowledge workers to answer questions, the problem of efficiently assigning requests to the right experts becomes critical. A new preprint on arXiv titled "Optimal Scheduling in a Question-Answering Forum of Knowledge Workers" tackles this challenge by modelling the request-answer process as a queuing system and developing schedulers that maximize capacity while maintaining stability.
The Research Problem
Question-answering (QA) forums have long been volunteer-based, but the researchers consider a future version where knowledge workers—paid experts in specific topics—field requests. In such a system, requests arrive in various topics, and each expert has a certain expertise level in each topic. The central problem is how to schedule the assignment of requests to experts to maximize the number of requests the system can handle without becoming unstable (i.e., without requests queuing indefinitely).
System Model and Capacity
The researchers model the QA forum as a queuing system where requests arrive randomly and are processed by experts. Each expert can handle multiple topics but with different proficiency levels. Using this model, they calculate the capacity of the system—the maximum rate of incoming requests that can be served while keeping the system stable. This capacity depends on the distribution of expertise across experts and the mix of topics in the request stream. The authors then design schedulers that can achieve this theoretical capacity, providing a blueprint for how to assign work in real-world platforms.
The Role of Collaboration
Beyond individual assignment, the paper investigates how collaboration between experts can further increase capacity. By allowing experts to jointly answer a request (e.g., combining knowledge from two specialities), the system may handle more requests or serve them faster. The researchers explore how such collaboration can be integrated into the scheduling framework to boost overall throughput.
Implications for Knowledge Workforce Management
While the study is theoretical, its findings have direct implications for enterprise knowledge management platforms, technical support systems, and any organization that deploys a team of specialists to handle varied requests. The model provides a quantitative way to evaluate the trade-offs between hiring generalists versus specialists, and between individual work versus team collaboration. For CTOs and operations leaders, these insights could inform the design of workforce scheduling algorithms that optimize both response time and resource utilization.
The research was authored by Negi, Rohit, Yilmaz, and Mustafa, and is available on arXiv as a preprint. It contributes to the growing field of AI-driven workforce optimization and offers a foundation for building more efficient knowledge-worker systems.
