See, Sense, Spark: Profiling Granularity in Proactive AI for Creative Research Ideation

@inproceedings{10.1145/3757374.3771458,
author = {Kim, Gahui and Choi, Yebom and Lee, Changjun},
title = {See, Sense, Spark: Profiling Granularity in Proactive AI for Creative Research Ideation},
year = {2025},
isbn = {9798400721342},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
url = {https://doi.org/10.1145/3757374.3771458},
doi = {10.1145/3757374.3771458},
abstract = {Proactive AI systems are a new paradigm for supporting cognitively demanding tasks by anticipating user needs and providing timely interventions. While prior creativity research often suggests that broad, less constrained information cues can enhance associative flexibility, it remains unclear how deeply these systems should understand users to improve their thinking. This paper explores the psychological effects of profiling granularity on creative research ideation. Using a proactive AI system called See, Sense, Spark, we conducted a between-subjects study comparing coarse and fine-grained user profiling. Contrary to our initial hypothesis, participants with fine-grained profiles showed significantly higher I-type curiosity. This result establishes profiling granularity as a core design parameter for proactive AI, highlighting that topical relevance can outweigh breadth in stimulating exploratory reasoning. We conclude by suggesting the need for adaptive systems that can balance both breadth and relevance to foster meaningful human-AI collaboration in complex cognitive tasks.},
booktitle = {Proceedings of the SIGGRAPH Asia 2025 Posters},
articleno = {12},
numpages = {3},
location = {
},
series = {SA Posters '25}
}