-
LearnDCG: End-to-End Joint Optimization of Ranker and Loss in Neural Ranking This paper, accepted at SIGIR 2026 (Melbourne, Australia), introduces LearnDCG, a differentiable and learnable approximation of NDCG that eliminates the rigid, hand-designed components of existing surrogate losses and trains them jointly with the ranker in a single end-to-end pipeline. The work was carried out…
-
This paper, published in Optics Express (Vol. 28, 2020), develops a comprehensive numerical model for broadband single-cycle terahertz pulse propagation that simultaneously accounts for non-paraxial diffraction, group velocity dispersion, and time-varying nonlinear refraction including both instantaneous and delayed Kerr contributions. The work was carried out at the University of Ottawa in collaboration with the Boyd…
-
Figure 1. Assessing human agreement with LLMs Figure 2. Using LLMs for user click-through behavior prediction and analysis Our paper, accepted at CIKM 2025, addresses a foundational bottleneck in entity retrieval research: the scarcity of high-quality relevance annotations. We investigate whether large language models can serve as reliable relevance assessors in this setting, evaluating their…