Breaking the information cocoon, causal inference empowers the recommendation system

301a721ba2fe864cba51992475739cf6.gif

Invited lecturer : Wang Wenjie , a researcher at the National University of Singapore, whose main research directions are recommendation systems, causal reasoning and multimodal computing. In 2023, he will receive a Ph.D. from the National University of Singapore under the tutelage of Prof. Chua Tat-Seng. He has published several first-authored papers in high-level conferences and journals such as SIGIR, KDD, WWW, WSDM, ACMMM and TOIS, and served as a reviewer for conferences and journals such as SIGIR, WWW, TPAMI, TOIS, etc. He has obtained 4 national invention patents , whose work has won the best paper candidate of ACM MM 2019. He has won honors such as Google PhD Scholarship, NUS Dean's Scholarship, and NUS Research Achievement Award.

23eb1b395532b9b161e64ed8aab3f4df.gif

0feefac424c5eb4e889125f414e2cf4c.jpeg

​​

Content preview


In today's information age, the recommendation system has become a key infrastructure for users to obtain information, affecting people's understanding and cognition of society. Although the recommendation system has achieved extensive development in recent years, the existing recommendation services have shown a trend of disordered growth, resulting in problems such as recommendation deviation, information cocoons, and poor generalization capabilities, which affect the normal order of information dissemination. In response to these problems, we analyzed the underlying mechanism from the perspective of causality, and proposed a causal recommendation framework based on causal knowledge, using causal reasoning to empower the recommendation system. This report focuses on our series of work in the field of causal recommendation in recent years, and discusses the future development trend of this field.

Syllabus:
1. Background overview of recommender systems and causal reasoning
2. Causal recommendation framework
3. Recent technological developments in causal recommender systems
4. Future research directions in causal recommender systems

9bf4975bc6f7ca4a9a48da94fb2f1b67.png

There are benefits for booking a live broadcast!

Scan the QR code to register and make an appointment, and there will be a text message reminder before the broadcast; watch the live broadcast to participate in the interactive lottery , and you can also receive 5,000 core hours/500 yuan card hour computing resources for free.

Interactive lottery 

gift

Beijing supercomputer customized multifunctional desktop metal bracket

cf4e42df9f11edaa4e9a54ed86eec107.jpeg

793583b88c644ea2c0f63820e74509ef.jpeg

569906bb3f3ff2b6820d98266f730c75.jpeg

a9e122cb5cdd10f1a7f65388186990b0.png

Guidance unit: China Computer Federation High Performance Computing Professional Committee

Sponsor: Beijing Super Cloud Computing Center

"Super Cloud Lecture Hall" is an online science popularization and public welfare live lecture hall integrating scientific research, academics, and education. Invite technical experts and well-known scholars in various fields to interpret the latest and hottest scientific research and popular science topics, and share application experience and successful practices based on the scientific research environment on the cloud. Committed to creating a scientific research, science popularization, and academic exchange platform that is easy to participate, happy to share, diligent in thinking, and good at practice, and builds a bond and bridge between scientific research, teaching and research workers and the public, not only focusing on the dissemination of scientific ideas, but also paying attention to the integration of ideas in different fields Intersection and collision aim to create a diverse and speculative scientific and cultural exchange atmosphere.

At present, the cloud lecture hall has launched 71 episodes in a row, with 50,000+ users' attention, and 500,000+ views in total. It also has 31 live exchange groups with 6,000+ active fans.

Links to watch past courses:

https://space.bilibili.com/505497082

Live QQ communication group number: 794834983

Live broadcast platform: bilibili, video account, Douyin

Popular science, public welfare, sharing, happiness

4996d634fd9bc202bfd547cda725534a.gif

ae3682c45aa9b4b928520915d61de6fd.png

Guess you like

Origin blog.csdn.net/qq_33431368/article/details/130256596