Skip to main content

Claire Wang & Michael Huang

Claire Wang & Michael Huang

2023 Davidson Fellows
$25,000 Team Scholarship

Age: Claire, 18 & Michael, 18
Hometown: Andover, MA

Technology: “TEAM Efficient Algorithms for Parallel Bi-core Decomposition”

About Claire

Hey! I’m Claire, a soon-to-be freshman at MIT hoping to study Computer Science & Neuroscience and maybe dabble in some Maths and Philosophy. I’m fascinated by how little we know about the brain and how we can create beautiful art/stories with today's technological advancements. Beyond just these subjects, I’m very interested in how we design today’s systems—computer systems, cities, tech policy, etc.

Beyond academics, I really enjoy hip hop and kpop dance, playing the piano, re-learning lambda calculus every few months, writing blog posts, biking around unfamiliar cities, doing theater and acting (recently, I was in a production about air guitarists!), ranting about transit systems, and memory sports. In the future, I hope to do some combination of deep tech entrepreneurship and academia on top of working in policy (especially in the realms of tech ethics & education).

Skip testimonial carousel

"Being a Davidson Fellow means that I can have a chance to learn the technical skills needed to help better global systems and to improve the impact this and future projects can have on the world while being inspired by the passion the people in the Davidson community have."

About Michael

I’m Michael. I’m also attending MIT next fall to study Computer Science. I love anything in math, physics, computer science, robotics, politics, and philosophy. I enjoy thinking about abstract ideas and systems. And sometimes, I put those ideas into words in my personal newsletter. I also love to build things like tiny video games and various computer science projects.

Outside of academic arenas, I love to play chess, go down philosophy rabbit holes, and sometimes dabble in writing—I blog on Daylight Reveries. In the future, I plan to continue dabbling in theoretical computer science research, but I would also love to build more practical tech products.

Skip testimonial carousel

"I am honored to be selected as a Davidson Fellow. This talented young community always inspires me to be the best version of myself and use my privilege to contribute back to this world. Davidson Fellowship gives me the invaluable opportunity to learn from this community and navigate my future endeavor with financial freedom."

Project Description

Our project develops a fast data mining algorithm that discovers densely connected communities within networks. These networks can represent social media communities, disease genetics connections, protein-protein interactions, political connections, and so on. Our algorithm leverages parallelism to improve efficiency over existing data mining solutions. Our algorithm can be used to detect fraudsters/spammers, identify network communities, and can be used in content recommendation algorithms, etc.

Deeper Dive

The algorithm we developed identifies a hierarchy of densely connected subgraphs within the larger bipartite graph. This can be used to detect fraudsters and bot accounts in social networks by identifying anomalous connectivities. It can be used in content recommendation algorithms and as a subprocess for discovering densely connected communities of vertices in general graphs.

To develop a parallel algorithm for mining bipartite graphs, we needed a deep understanding of parallel computing, general graph mining algorithms, and specific properties of bipartite graphs. Most of these contents are not covered in a traditional computer science curriculum because they are relatively novel or else specific fields. We benefited from the MIT PRIMES program, which matched us with our mentors, Jessica Shi and Prof. Julian Shun from the MIT Computer Science and Artificial Intelligence Lab. Under the guidance of our mentors, we studied parallel computing using lectures from MIT OCW and brought ourselves up to speed with pioneering research in general graph mining algorithms by reading papers. During the pandemic, we divided our work and met through Zoom to synchronize progress.

Our algorithm can process bipartite graphs several times faster than existing algorithms. This will enable researchers to analyze the density hierarchy structure of gene-disease correlation networks, protein-protein interaction networks, and social networks much more quickly. We expect our algorithm, when applied, to be able to speed up research in genetics, biochemistry, and social studies. Our algorithm can also be applied to detect spam bots or fraudsters in social networks quickly. This can conceivably help social media companies improve the user experience of their products. The ideas developed in our work could conceivably also have wider applications to other parallel computing research. The ideas used to parallelize our algorithm could conceivably inspire research to parallelize other similar algorithms. Increasing the parallelism of algorithms is important in the modern era as it becomes difficult to increase the clock speed of CPU cores due to the generated heat. But Moore’s Law permits us to fit ever more transistors in one CPU. So, chip manufacturers fitted the increasing number of transistors into a larger number of cores instead. With the number of cores in a commercial CPU increasing at an exponential rate, it’s critical for algorithms to have significant parallelism to take advantage of a large number of cores, hence the general importance of research that focuses on developing algorithms with high parallelism.

Our work has many specific applications we have described before, such as detecting spammers or better understanding the relationship between people in social networks and different proteins. Beyond that, there are many general applications based on the optimization methods we’ve used for parallelism (i.e., scheduling or multithreading) and graph decomposition beyond just bipartite graphs. This is super useful as the amount of data any piece of software will have to process as an increasing number of people rely on social networks or discover new proteins.


What is your favorite food?

Claire: Coffee 🙂 in all forms: cold brew, celsius, lattes… you name it, I love it.

Michael: Dumplings

What is your favorite tradition or holiday?

Claire: Christmas! basic answer, but I just *love* winter—It’s got everything. My birthday, my sister’s birthday, all the celebration of winter & new years… it’s a happy month.

Michael: Favorite holiday has gotta be summer holiday!

What are the top three foreign countries you’d like to visit?

Claire: Australia, Portugal, and Estonia

Michael: Japan, Britain, New Zealand

Click image to download high resolution files

In The News

Boston – The Davidson Fellows Scholarship Program has announced the 2023 scholarship winners. Among the honorees are Michael Huang, 18, Claire Wang, 18, and Anna Du, 17, of Andover. Only 21 students across the country are recognized as 2023 scholarship winners.

Download the full press release here