no code implementations • 5 Feb 2024 • Ashley Shin, Qiao Jin, James Anibal, Zhiyong Lu
Our study suggests that repurposing user query logs of academic search engines can be a promising way to train state-of-the-art models for explaining literature recommendation.