Visiting Scholars: Maximizing Research's Impact

There’s synergy between CS research and our visiting scholars from Hitachi: Chi Heem Wong and Gaku Morio. A recent interview with our esteemed guests highlights the energy and passion that emerges from our visiting scholar program. Through the program, affiliate programs provide member companies the opportunity to send their scientists or engineers for an extended stay at Stanford to collaborate with faculty.

What is your title and role at Hitachi?
Chi Heem: “I am a Senior Research Scientist applying machine learning and data science techniques to solve problems in finance, supply chain, and heavy industry.”

Gaku: “I am a Researcher of NLP and AI, pursuing research in advanced technologies at the intersection of finance and climate change.”

Describe your research here at Stanford
Chi Heem: “I work with Percy Liang on various projects, such as evaluating vision-language models [1,2] or training large language models. Previously, I joined a project with Jure Leskovec on learning production functions for supply chains with graph neural networks [3]. These papers have been accepted in top conferences such as NeurIPS and AAAI.”

Gaku: “I work with Prof. Christopher D. Manning to contribute to the analysis of corporate climate stances and sustainability reporting. We have published in leading NLP and AI conferences such as NeurIPS [4], ACL [5], and IJCAI [6] and have collaborated with global experts in climate policy, sustainable finance, and AI.”

Briefly describe your educational background. How has your education and professional experience benefitted Stanford?
Chi Heem: “I obtained my B.Eng. (Hons) in Mechanical engineering and B.Soc.Sci (Hons) in Economics from the National University of Singapore and my S.M. and Ph.D. in Electrical Engineering and Computer Science from the Massachusetts Institute of Technology.

I previously interned in a central bank and Microsoft, among others, and worked full-time as an operations research engineer in an industrial giant and also as a quantitative researcher in a start-up hedge fund. As such, I bring fresh perspectives and different domain knowledge to the discussions at Stanford.”

Gaku: “I obtained a BEng with early graduation honor and MEng from the Tokyo University of Agriculture and Technology. Since I was in the computer science department at the university and had a professional career at Hitachi after graduation, I didn’t encounter any difficulties starting research at Stanford.”

What has impressed you about the caliber of research, the culture, and community of people you’ve encountered at Stanford?
Chi Heem: “The students and professors I work with are smart, hardworking, and committed to the highest standard of research.”

Gaku: “I think Stanford is a stimulating yet friendly and open environment. The professors are open to discussing various research ideas, and they provide generous help with everything from experimental design to writing papers.”

How does Hitachi benefit from your research at Stanford?
Chi Heem: “Hitachi can use the research internally in some of their product offerings. Furthermore, collecting new ideas from academia can help the company develop new technologies ahead of the competition.”

Gaku: “Collaboration with Stanford is a powerful way to improve the quality of our research and maximize its impact. It is beneficial for both Stanford and Hitachi to lead communities and initiatives in specific research fields.”

What is the value of the visiting scholar program? What can you do here in the US that you cannot do from Japan?
Gaku: “One of the great benefits of the Visiting Scholar program is that I have easy access to not only my advisor, but also to a variety of professionals within the university. As an industrial researcher who works in an interdisciplinary field, this is very attractive and something I would not be able to do if I did not come to the university.”

How can other divisions at Hitachi benefit if they were to send a visiting scholar to Stanford?
Gaku: “First of all, we are conducting research with Stanford that will benefit everyone. In principle, the research data and code we have developed are made public. This is completely in line with the policy of Stanford Data Applications, which supports open IP. We then verify the applicability of the research to our businesses.”

[1] Roberts, Josselin Somerville, et al. "Image2Struct: Benchmarking Structure Extraction for Vision-Language Models." Advances in Neural Information Processing Systems, 37, 115058-115097.
[2] Lee, Tony, et al. "VHELM: A holistic evaluation of vision language models." Advances in Neural Information Processing Systems, 37, 140632-140666.
[3] Chang, Serina, et al. "Learning production functions for supply chains with graph neural networks." arXiv preprint arXiv:2407.18772 (2024).
[4] An NLP Benchmark Dataset for Assessing Corporate Climate Policy Engagement. Gaku Morio, and Christopher D Manning. In Advances in Neural Information Processing Systems: NeurIPS 2023 Datasets and Benchmarks Track, Aug 2023.
[5] Predicting Narratives of Climate Obstruction in Social Media Advertising. Harri Rowlands, Gaku Morio, Dylan Tanner, and Christopher Manning. In Findings of the Association for Computational Linguistics: ACL 2024, Aug 2024.
[6] ReportParse: A Unified NLP Tool for Extracting Document Structure and Semantics of Corporate Sustainability Reporting. Gaku Morio, Soh Young In, Jungah Yoon, HarriRowlands, and Christopher Manning. In Proceedings of the Thirty-Third International Joint Conference on Artificial Intelligence, IJCAI-24, Aug 2024. Demo Track.