JD Showcases Ten Papers and DEERS Recommendation System at KDD 2018

The International Conference on Knowledge Discovery and Data Mining (KDD) 2018, held in London, featured JD’s extensive participation, with ten peer‑reviewed papers demonstrating the company’s advances in big‑data and artificial‑intelligence research across multiple domains.

One highlighted paper, “Recommendations with Negative Feedback via Pairwise Deep Reinforcement Learning,” introduces DEERS, a novel reinforcement‑learning framework that models the recommendation process as a Markov Decision Process, enabling continuous optimization of recommendation strategies through real‑time user interaction.

JD also organized a themed “KDD in Retail Industry” technical night, gathering ACM SIGKDD Innovation Award winners, senior JD scientists, and global experts to discuss future directions, opportunities, and challenges in data mining for retail.

Beyond recommendation, JD’s research spans AI‑driven medical case discovery, smart‑city analytics, and intelligent supply‑chain solutions, reflecting a strong integration of academic research with practical business scenarios.

JD’s leadership emphasized the importance of coupling technology with rich business contexts, positioning the company as a platform for talent to develop, validate, and apply cutting‑edge AI and big‑data solutions.