Interview with Prof. Sarika Jalan

  1. Brief self-introduction

This is Sarika Jalan from IIT Indore. Currently, I am at Physics of Complex Systems @ IBS for my sabbatical. Broad areas of my research are complex networks including social networks, synchronization, spatio-temporal chaos, spectral graph theory and applications of RMT.

  1. Could you explain your recent research?

Recently, we have been interested in understanding how dynamical properties of nonlinear units coupled through a network can be controlled or tuned by multiplexing this network with another network forming a multilayer network.  Remarkably, we have found that by appropriate multiplexing one can induce first order phase transition to synchronization, referred as explosive synchronization, in networks which are incapable to exhibit the same in isolation.

  1. What is your future research plan?

To relate structural properties of the underlying network structure with dynamical properties, such as cluster synchronization and explosive synchronization. I have two immediate future goals: The first one, which is more mathematically inclined, aims to relate localization properties of underlying adjacency matrices with the complexity of corresponding networks. Second one is of more interdisciplinary in nature combining machine learning techniques to achieve the goal of relating network structure representing a complex system with dynamical properties of that complex system.

  1. What motivated you to research with IBS, and what is the current research in IBS?

IBS has a vibrant international atmosphere with a great scope of inter-disciplinary research and stimulating discussions. The physics of complex systems (PCS) group has scientists with expertise in Anderson localization, and being in IBS offers a unique opportunity for advancing my research on localization properties of complex networks.

The second part of my work which pertains to machine learning applications aiming to relate network structure with dynamical properties of corresponding complex systems is being carried out in collaboration with the data science center at IBS. As a first step, we have initiated a collaborative project on “Twitter networks” by utilizing COVID19 data.

  1. What are the most challenging and favorite parts of your researches?

Networks provide a very simple framework to model and predict various emerging properties of a diverse range of real-world complex systems. The real challenge, also the favourite part of my research, is to identify the crucial properties of an individual complex system and incorporate them in the corresponding “coupled dynamics on networks” model for a better understanding and predictions of the system’s behaviours. Interestingly, a variety of complex systems having underlying network structures exhibit universal features and mechanisms making this direction of research extremely fascinating.

  • Contact: sarika (at) iiti.ac.in

Interview with Dr. Inho Hong

  1. Brief self-introduction

Inho Hong is a postdoctoral researcher at Asia Pacific Center for Theoretical Physics. His research focuses on understanding social systems through complex systems analysis, with particular interests in urban science, human mobility, scaling theory, economic complexity, and science of science. Inho received his Ph.D. in physics at Pohang University of Science and Technology (POSTECH) in 2019. Previously, he worked at Northwestern Institute on Complex Systems and Kellogg School of Management at Northwestern University as a visiting predoctoral fellow.

  1. Could you explain your recent research?

I am recently working on two main projects, the urban recapitulation and the gravity model on urban landscapes. The urban recapitulation project is about explaining the longitudinal changes of employment and population in US cities using the urban scaling law. The work on the gravity model is about understanding the unique characteristics of human mobility on centralized urban landscapes. Through these projects, I am trying to answer how spatial constraints shape some regular patterns in social systems and human behaviors.

  1. What is your future research plan?

I will start working as a postdoctoral researcher at Max Planck Institute for Human Development from March 2020. The main theme of the position is the Future of Work which explains how scientific and technological developments interplay with the labor market and society. Through this work, I hope to find some key drivers of societal changes and show the blueprint of the future of society. Specifically, I wonder how the recent growth of AI and machines will impact urban labor markets and what will be the optimal strategies for these changes. In addition to this work, I am also open to any project on social systems and human behaviors such as cities, human mobility, scaling, and economic complexity.

  1. What motivated you to research with IBS, and what is the current research in IBS?

Understanding cities and spatial interactions are one of the main parts of my research. Therefore, recent achievements at IBS using satellite imagery are fascinating that it can provide great spatiotemporal information of unknown regions. I started to collaborate with IBS on the relation of urban greenspace and the happiness of developed countries. I believe that this work will attract great attention from researchers and general audiences.

  1. What are the most challenging and favorite parts of your researches?

Finding high-quality data has always been the most challenging part. Even if I have a great model, it is not proved until it is supported by data. In this sense, I believe that collaborating with the Data Science Group at IBS has great potential to provide impactful scientific findings. The favorite part of my research is the integration of complex systems frameworks and data analysis. Sometimes, a result of data analysis shows a good agreement with an intuitive model, and then I feel like touching the hidden nature of the world! Understanding the mechanism of our world is a great joy to me as a physicist.

Interview with Dr. Kunwoo Park

  1. 본인의 연구를 간략히 설명해주세요.

가장 최근에 한 연구는 가짜 뉴스와 관련이 있습니다. 가짜 뉴스는 사람이 판별하는데 한계가 있기 때문에 이를 잘 포착하는 방법론을 제시함으로써 사회에 기여하고자 노력하고 있습니다. 그 외에도 전반적인 뉴스나 온라인 소셜 미디어가 어떤 특성을 가지는지, 어떤 방법으로 사람들 사이에서 퍼지는지에 대해 관심을 갖고 연구를 진행하고 있습니다.

  1. 지금의 연구과제를 선택하게 된 계기는 무엇인가요?

지금까지 다양한 연구를 했었습니다. 그 중에는 처음부터 관심이 있어서 한 연구가 많지만 가짜 뉴스 연구는 reinforcement learning을 공부하다 이를 text style transfer에 활용하면 좋겠다는 아이디어에서 출발했습니다. 그 이후로 제 연구 동료들과 이야기하다보니 뉴스 헤드라인을 트윗용 헤드라인으로 바뀌고 sns상에 공유되는 과정을 자동화할 수 있으면 좋을 거 같아 이 과제를 시작했습니다.  그러다보니 자연스레 클릭베이트나 신문사별 기사 특성에도 관심을 가지게 되어 여기까지 오게 되었습니다

  1. 앞으로 어떤 연구를 하고 싶으신가요?

미디어들도 자기 나름의 사용하는 전략들이 있을거잖아요? 제가 미디어 기업과 협업해서 제가 만든 모델과 현재 미디어가 사용하는 전략을 서로 비교 분석해보고 싶습니다. 또 지금까지는 사용자에 대한 단기적인 영향을 보았었는데 장기적인 반응을 연구하는 것도 흥미로울 듯 합니다. 예를 들어 클릭 베이트가 사용자의 이목을 끄는데 좋지만 장기적으로 신문사의 인기를 하락시키는 등의 영향을 줄텐데 앞으로 글쓰기 스타일에 따라 사용자에게 어떤 영향을 주는지 장기적으로 보고 싶습니다.

  1. 2018년 말부터 지금까지 여기 데이터사이언스 그룹에서 연구를 하셨다고 들었습니다. 그동안 연구를 하면서 저희 연구실만이 가지는 특별한 것이 있었다면 한말씀 부탁드립니다.

일단 다른 전산 관련 연구실은 기술 자체를 연구하는데 더 관심을 가지는 것 같습니다. 얼마나 빠르고 효율적으로 계산을 할 수 있는지에 관심을 가지고 있고, 문제에 모델을 응용할 때도 문제를 해결한다는 것보다는 모델의 우수성을 입증하기 위한 경우가 많습니다. 하지만 여기는 그런 모델을 이용해 사회를 어떻게 이롭게 할 수 있는지에 초점을 맞춘 연구를 진행하더라구요. 아직 우리나라에서는 이런 관점으로 연구를 하는 곳이 많지 않다고 생각합니다.

  1. 마지막으로 우리 연구실에 대학원생으로, 혹은 인턴으로 지원하고 싶어할 학생들에게 한말씀 부탁드립니다.

여기에 와서 무언가를 배우고 싶은 생각이 있다면 망설이지 말고 지원을 하셨으면 좋겠어요. 도전할 수 있을 때까지는 도전하는게 좋으니까요. Barabasi라는 complex system 분야에서는 유명한 교수님이 있으신데, 그분이 어떤 사람이 성공하는가를 분석했더니 꼭 높은 자리에 있는 사람이 성공하는것이 아니라 높은 자리에 지원하는 사람이 성공을 한데요. 그러니 여러분이 관심이 있다면 주저없이 지원해보세요.

Interview with Dr. Jing Ma

Dr. Jing Ma was a final-year Ph.D. student at the Chinese University of Hong Kong when she was visiting us in January 2020 (She graduated in February 2020). Her research topics include Natural Language Processing, Rumor Detection, Stance Detection, Social Media Analytics. She has published around ten papers about fake news detection at prestigious venues like ACL, IJCAI, WWW. She was also a visitor to the NLP group at NTU for more than eight months. Previously, she worked as a research intern at companies like Tencent AI Lab, Baidu NLP.

  1. Could you explain your recent research?

My recent research topic is fake news detection. Claim verification is generally a task of verifying the veracity of a given claim, which is critical to many downstream applications. Any arbitrary claims need to be verified. We need to determine the truthfulness of each claim by using NLP techniques. In this way, we can make the internet environment to stop the fake news or rumor propagation further that mislead people’s leading in social media or news media. 

  1. What is your future research plan? 

Fake news detection still has many potential problems to solve. So I am going to make an in-depth analysis of fake news detection or claim verification. I also plan to do another interesting NLP task like hate speech or Wikipedia contents quality evaluation. 

  1. What motivated you to research with IBS, and what is the current research in IBS?

Mia is a very excellent researcher, and I would like to work with Mia. It is the main reason I research with IBS. IBS is a fantastic workplace. IBS provides enormous support to researchers and interns. After I joined IBS, I would like to find a potential research topic in which I can give some help or join the project to achieve significant progress. For now, I lead the hate speech detection, fake news detection, and Wikipedia project. Especially hate speech project in which I work with IBS is an up-and-coming project. As social media develops, more and more people express their own opinion, including hate speech, which harmful to children. We also can create a lot of works contributions based on hate speech projects. So I think hate speech project is a very significant task in the future. Studying with IBS data science group members offers many ideas about my future researches. I want to continue to study with IBS.

  1. What are the most challenging and favorite parts of your researches?

I think the most challenging problem in the NLP task is to find interesting research topics. I need to predict whether my project can survive for quite a long time. I always consider that my project to be a “seed” project so that it can bring out a lot of relevant follow-up researches. For example, in the case of my fake news detection project, few people researched this project previously. I lead a fake news detection project and give some fundamental contributions to this project so that others can do further development based on my research. I think this is my favorite and challenging part of my studies.