Chenglong Ma

Chenglong Ma

Ph.D. | Research Fellow at ADM+S

ADM+S Centre

RMIT University

Biography

Platforms like Amazon and YouTube train their recommender systems by understanding user interest from their interaction history. In the normal days, no problem. But during extraordinary events like the COVID-19 pandemic, people tend to exhibit unusual behavior, such as panic buying. My work is to differentiate between genuine user interests and irrational behavior, in order to improve the quality of recommendations.

Interests
  • Recommender Systems
  • Information Retrieval
  • Responsible AI
  • User Modeling
  • Social Influence
  • Simulation
Education
  • PhD in Recommender Systems, 2020

    RMIT University

  • Master of Information Technology, 2018

    RMIT University

  • Bachelor of Mechatronic Engineering, 2010

    Beijing Information Science & Technology University

Recent Publications

Quickly discover relevant content by filtering publications.
(2024). Temporal Conformity-aware Hawkes Graph Network for Recommendations. Proceedings of the ACM Web Conference 2024.

Cite Poster DOI URL

(2022). NEST: Simulating Pandemic-like Events for Collaborative Filtering by Modeling User Needs Evolution. Proceedings of the 31st ACM International Conference on Information & Knowledge Management.

Cite Video DOI URL

(2022). Evaluation of Herd Behavior Caused by Population-scale Concept Drift in Collaborative Filtering. Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval.

Cite DOI URL

(2021). From Changing User Needs to Changing Behavior: The Impact of User Macro-Trends on Recommender Systems. SimuRec: Workshop on Synthetic Data and Simulation Methods for Recommender Systems Research.

Cite DOI

Projects

*
Zoplicate
The plugin for Zotero specializes in detecting and managing duplicate items within the library. It automatically identifies if a newly imported item is a duplicate of an existing one, prompting the user to handle and process any duplicate items accordingly.
Zoplicate
Skin Tone Classifier
The library is user-friendly and aims to classify skin tones in images, detecting faces and categorizing skin tones into specified colors. It generates a report highlighting detected faces, dominant skin tones, and color categories.
Skin Tone Classifier

Experience

 
 
 
 
 
Blobfish International
Software Engineer
December 2019 – Present Melbourne, Australia

Responsibilities include:

  • Web app development
  • Mobile Android, iOS apps development
  • DevOps
 
 
 
 
 
RMIT School of Computing Technologies
Tutor for Practical Data Science
February 2020 – July 2023 Melbourne, Australia

Responsibilities include:

  • Teaching tutorial and lab classes
  • Marking assignments and exams
  • Mentoring students
 
 
 
 
 
CAXA Technology CO., LTD.
Software Engineer
July 2014 – April 2017 Beijing, China

Responsibilities include:

  • Adapted and extended existing MES/ERP solutions
  • Developed Automatic Production Scheduling Modules
  • Developed Server Modules for Data Synchronism between MES, ERP and CAPP systems.

Contact

Hello there! 👋 I’m thrilled that you’ve reached out.

Whether you’re here for a research discussion, to explore potential projects, propose a collaboration, or just to say hi, your message is eagerly awaited!