ADM+S Student presents at the 2024 World Wide Web Conference in Singapore
Author Natalie Campbell
Date 31 May 2024
ADM+S PhD Student Chenglong Ma recently presented at the 2024 World Wide Web Conference (WWW’24) in Singapore.
Chenglong presented a poster and oral presentation on Temporal Conformity-aware Hawkes Process on Recommendations, which challenges the assumption that user behaviour in recommender systems is solely driven by personal interests, and highlights the influence of peer effects and conformity behaviour.
His work criticizes existing solutions that overlook this influence and introduces the TCHN model which employs attentional Hawkes processes to separate user self-interest from conformity, and temporal graph attention networks to capture users’ changing dynamics.
“I’m thrilled that my work on the Temporal Conformity-aware Hawkes Proess on Recommendations received significant attention and valuable feedback,” said Chenglong.
“It was a wonderful experience, and I had the opportunity to meet many outstanding researchers.”
The WWW’24 is an annual academic conference on the topic of the future direction of the World Wide Web. It remains the premier venue to present and discuss progress in research, development, standards, and applications of the topics related to the Web.
In addition to making important connections during the conference, Chenglong also connected with leading researchers in his field while visiting Nanyang Technological University, including Prof Aixin Sun.
“We share common views on research issues in recommender systems, and I greatly admire his rigorous and critical thinking.
He criticised the simplification of research task definitions for overemphasising modelling decision outcomes rather than the decision-making process. This approach hinders the ability to predict users’ decisions in subsequent interactions within a dynamic, evolving, and application-specific context.”
Prof Sun praised Chenglong’s ability, as a PhD student, to discover and identify valuable research questions in addition to merely solving the problems, noting that some studies overly focus on improving recommendation accuracy, neglecting research questions that are more valuable and worthy of exploration.
Chenglong’s WWW’24 experience was supported by ADM+S.
