Every summer, Megagon Labs gets busy as we work with top talented interns from universities around the world. Unlike last summer, when having the complete remote internship in the middle of the pandemic, the summer 2021 interns came to the Bay Area and worked in the office every day with their mentors, lab members, and internship mates. They also enjoyed the summer of California and even a team offsite event in the Santa Cruz mountain.
In this article, our interns are invited to discuss their projects, their path to Megagon Labs, internship experience, and more!
Ting-Yao Hsu
University:
Pennsylvania State University, Ph.D. Program (Advisor: Prof. C. Lee Giles, Co-Advisor: Prof. Ting-Hao K. Huang)
What was your project?
During my internship, I worked on a project to help humans quickly and better obtain useful information from E-commerce platforms. More specifically, we would like to extract both subjective and factual information from large-scale community question-answering to generate short essential and precise summaries. (Project team members: Xiaolan Wang and Yoshi Suhara)
How did you make the path to Megagon Labs?
I am a Ph.D. student working on NLP-related research. I had a great interview experience with Megagon Labs and my labmate also interned there last year. After reading some publications by the members in Megagon Labs, I found several research topics that I would be interested in. Even though it didn’t completely match my works in school, it was still a good opportunity to extend my NLP knowledge and get inspired for my current research.
What was your favorite part of the internship?
I am glad that I interned at Megagon Labs this summer. Everyone was unexpectedly nice and intelligent there. Especially my mentors were supportive and gave me much useful advice, not only about the project but general research. They always helped me to think more and deep about the research without giving me too much pressure. I was happy to collaborate with them and learn quickly in a short time. Moreover, I had a great working environment with other researchers there. We often shared research and life experience opinions with each other, which was also my delighted part of the internship.
Tianyu Jiang
University:
University of Utah, Ph.D. Program (Advisor: Prof. Ellen Riloff)
What was your project?
My project was about behavior testing and data augmentation. We aimed to provide an insight into whether different data augmentation techniques would lead to different model behavior even when they had similar accuracy or f1 scores. (Project team members: Nikita Bhutani, Yoshi Suhara, Jinfeng Li, and Estevam Hruscka)
How did you make the path to Megagon Labs?
When I attended the EMNLP 2020, I came across Megagon team members and found out that we shared research interests, such as affective events and commonsense learning. So I applied for the summer internship and was fortunate to have the opportunity to work with the team.
What was your favorite part of the internship?
My favorite part of the internship was the collaboration of the team. During the weekly team meeting with my mentors Nikita and Yoshi, as well as Estevam and Jinfeng, I always received insightful comments and feedback from a variety of perspectives, which often include ideas or clues that were ignored or under-estimated by myself. Nikita and Yoshi not only gave me freedom and flexibility to explore the problem but also closely followed up on the progress and were always supportive when I sought help. The friendly and relaxed atmosphere is not limited within the research team — Makiko and Anita helped make things much easier under the pandemic situation.
Zafeiria Moumoulidou
University:
University of Massachusetts, Amherst, Ph.D. Program (Advisor: Prof. Alexandra Meliou)
What was your project?
I worked on understanding the semantics of structured datasets in data lakes. The goal was to create a semantically labeled dataset for evaluating the performance of learning-based methods on related tasks. (Project team members: Xiaolan Wang, Yoshi Suhara, and Yuliang Li)
How did you make the path to Megagon Labs?
I am a Ph.D. student at UMass Amherst working on algorithmic fairness and data diversification. I learned about Megagon Labs from alumni of our lab who did their internships at Megagon and are now working full-time here. Although my area of research is not directly related to my intern project, I was happy to get to know a new area of research and learn new skills. I am grateful I had the opportunity to be part of the team!
What was your favorite part of the internship?
Meeting and working with great people at Megagon Labs! I really enjoyed being part of a team and collaborating with my mentors. I also liked the fact that I got the chance to meet other interns from different universities and learn about their research experiences.
Yutong Shao
University:
University of California, San Diego, Ph.D. Program (Advisor: Prof. Ndapa Nakashole)
What was your project?
I worked on seed-based knowledge base construction. In detail, given the KB taxonomy (concepts and relation types), a small seed entity set for each concept, and an unlabeled corpus, we aimed to construct a KB with expanded entity sets and relations between entities. (Project team members: Nikita Bhutani, Sajjadur Rahman, Behzad Golshan, and Estevam Hruschka)
How did you make the path to Megagon Labs?
I’m a Ph.D. student at UCSD. My advisor Ndapa Nakashole has visited Megagon Labs before. She helped me connect with Estevam, Interim Head of Research here, and he was one of my mentors on my internship project. After talking with Estevam and checking out projects and publications from Megagon Labs, I feel that there is a great overlap between the research going on here and my own research interest, mostly under the high-level topic of NLP & KB. Therefore I applied for the internship program, and fortunately, here I came!
What was your favorite part of the internship?
People at Megagon Labs were very nice. My mentors were always accessible, supportive, and willing to answer any questions, either high-level ideas, low-level technical details, or non-academic ones. They also spent a lot of effort on my project (besides discussions) to help make progress. Other people were also very friendly and helpful. It felt like everyone was willing to talk to others, know others and help others that made an enjoyable atmosphere at the lab.
Runhui Wang
University:
Rutgers University, Ph.D. Program (Advisor: Prof. Dong Deng)
What was your project?
I was working on applying self-supervised learning for data integration. The specific data integration tasks include entity matching, data cleaning, and domain discovery. The goal was to provide an end-to-end solution that involves little to no human efforts for these tasks. (Project team members: Yuliang Li and Jin Wang)
How did you make the path to Megagon Labs?
I got to know Megagon Labs during the VLDB 2020 conference. After learning about the projects in the lab, I got very interested and applied for the internship on Megagon Labs’ website. Then, I got an interview and had a very nice conversation about research ideas and projects with my interviewer Xiaolan Wang. After several weeks, I received an offer!
What was your favorite part of the internship?
There were many aspects that I enjoyed in the lab: people are friendly and inclusive, and I always get constructive suggestions from them. The best part of the internship was that I could always learn new things, explore new ideas, and perform experiments quickly with the help of my mentors, Yuliang Li and Jin Wang. They were very responsive and gave great feedback that helped me a ton on the project.
Huayi Zhang
University:
Worcester Polytechnic Institute, Ph.D. Program (Advisor: Prof. Elke A. Rundensteiner)
What was your project?
In my intern project, I worked on the table understanding project. We aimed to develop a technique that leverages pre-trained language models to analyze the semantic type of each column in a table. (Project team members: Yoshi Suhara, Yuliang Li, Dan Zhang, and Jinfeng Li)
How did you make the path to Megagon Labs?
I read a few papers published by researchers at Megagon Labs. I find them really interesting and close to my research topic. I then decided to apply for the internship here.
What was your favorite part of the internship?
I really enjoyed working with knowledgeable researchers here. All my team members were very friendly and helpful. I was also very fortunate to have the freedom to define problems and find solutions for my project. Also, I was very impressed with the working culture here. Everyone not only does a great job on research but also has a good work-life balance.
If you would like to learn more about our internship program, or to apply, check out our internship page.
Written by our summer 2021 interns and Megagon Labs