Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This study investigates the design of recommender systems in the context of Timee, a spot-work platform, focusing on how recommendations influence access to limited job opportunities. By implementing a thresholded eligibility control (TEC) mechanism, the authors effectively reallocate exposure to job templates based on real-time posting activity and labor demand, addressing the issue of over-concentration on popular templates. The results demonstrate a significant increase in job-finding rates, from 57.6% to 70.0%, alongside improved exposure and matching outcomes in a randomized field experiment.
A novel exposure-control mechanism boosts job-finding rates by reallocating recommendations from oversaturated to underexposed job templates.
How should recommender systems be designed when recommendations shape access to scarce, short-lived opportunities? We study this question in a production setting: Timee, Japan's largest platform for spot work, where workers favorite job templates and receive notifications when firms post shifts from those templates. Maximizing predicted favoriting can generate misdirected concentration: recommendations accumulate on popular templates that create few viable job openings, while templates with unmet labor demand receive too little exposure. We design exposure-control mechanisms for favorite-list management, reallocating template exposure based on posting activity and unfilled capacity. The proposed recommender, thresholded eligibility control (TEC), is fully parallelizable and suitable for large-scale digital platforms. In simulations calibrated to Timee data, TEC raises the per-round job-finding rate from 57.6% to 70.0%. A prefecture-level randomized field experiment increases realized matches and exposure per active template, reduces the share of low-exposure templates, and improves impression-level favoriting and downstream matching.