Purdue-Chen-MC3



Ava Zhao, Zhanqi Su, William C Fei, Na Zhuo, Hao Wang, Tianzhou Yu, Zuotian Li, Zhenyu Cheryl Qian, Yingjie Victor Chen

 Room: 104

2023-10-22T22:00:00ZGMT-0600Change your timezone on the schedule page
2023-10-22T22:00:00Z

Exemplar figure, described by caption below — Self-structured mind map with data visualizations that generated via ChatGPT-assisted learning.

Fast forward

Full Video

Abstract

To solve the VAST Challenge 2023 MC3, our team employed a large language model, ChatGPT, to explore the potential of AI-guided visual analytics for the detection of anomalies within a knowledge graph in the context of illegal fishing and marine trade. We employed a systematic and iterative approach, guided by GPT augmentation, that enabled problem understanding, data processing, solution explo- ration, code writing, and results analysis. By generating and analyz- ing various graphs, we identified anomalies related to revenue and product services. Further analyses unveiled potential illegal fishing activities and identified instances warranting additional investigation. Overall, our work highlights both the strengths and limitations of ChatGPT in aiding the visual analytics process and emphasizes the importance of human judgment in refining AI-generated outputs.