The goal(s) of this project is to:
1) investigate how clearly and consistently graph machine learning practitioners conceptualize graph neural network expressive power;
2) uncover how practitioners perceive the validity of common measurement models for expressive power;
3) inquire into the extent to which practitioners’ conceptualizations of expressive power and measurements thereof are driven by and impact real-world applications.
Survey results will be presented in aggregate in the form of a research paper, which will be submitted to a conference in 2023.
This survey will take approximately 5-10 minutes. We thank you for your participation.
Researchers involved:
1) Arjun Subramonian, UCLA
2) Levent Sagun, Meta AI
3) Adina Williams, Meta AI
4) Maximilian Nickel, Meta AI
If you have any questions or concerns, please contact arjunsub@cs.ucla.edu.