NEWS

Companies can now gain unprecedented insights and help raise customer interest in engaging with a company and its products. Shion Ishikawa and Young-joo Chung presented their latest paper at this year’s AdKDD workshop, which explores novel ways to understand and enhance the customer journey as well as inform actions in the digital sphere.

In “Dynamic collaborative filtering Thompson Sampling for cross-domain advertisements recommendation”, the authors propose a dynamic collaborative filtering Thomson Sampling (DCTS) to improve the performance of reinforcement learning-based recommender systems. Conducting an empirical analysis, the authors show that DCTS can outperform other state-of the-art models for cross-domain and dynamic recommendation, and quickly adapt to changing customer preferences.

“Dynamic collaborative filtering Thompson Sampling for cross-domain advertisements recommendation” was co-authored by Yu Hirate, respectively.  The paper has been made public so you can read it here and gain a deeper insight into their model, experiments, and results.

About AdKDD:

AdKDD workshops highlight state-of-the-art advances in computational advertising. AdKDD 2022 was held in conjunction with the 28^th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD 2022) in Washington DC on August 15.