Bio

I am currently an Applied Research Scientist at Snapchat. I solve Data Science research problems for the company. My interests lie in Applied Machine Learning (model design), Data Science (insight generation), and Economics (causal explanation). I just completed my Ph.D. in Computer Science from the University of Illinois at Urbana–Champaign (UIUC)). For my doctoral thesis, I studied the successes and failures of content-based online platforms, such as Stack Exchange and Reddit. My research received an honorable mention award at ACM CSCW 2019.

I am enthusiastic about exploring the data science landscape in tech industry. I have worked as a research intern at Adobe (Summer 2016), Visa (Summer 2018), and Snap (Summer 2019). At Adobe, I designed a pattern mining method for extracting user tasks from Photoshop logs. At Visa, I built a deep learning workflow for generating restaurant profiles from Visa transactions. At Snap, I developed survival models for studying the adoption and diffusion of Snapchat lenses. The outcomes of these internships include patents, publications, and software products.

In my previous life, I was a lecturer in Computer Science and Engineering at Bangladesh University of Engineering and Technology (BUET), having completed my B.Sc. from the same department. I received several honors for my undergraduate research including The Undergraduate Awards [world's largest academic awards program for undergraduates], invited young researcher at Heidelberg Laureate Forum [prestigious meeting with the Abel, Fields, and Turing Laureates], and National ICT Achievement Award [prestigious recognition in ICT by the Government of the People's Republic of Bangladesh]

Research


Graduate Projects
Why did Yahoo Answers fail? Why is Stack Overflow declining? Will Quora survive in 2020? — When are community question answering websites like StackExchange, sustainable? In our recent paper at the Web conference, we investigate CQA sustainability from an economic perspective–providing insights on their successes and failures through an interpretable model.
Read more
Content (post, comment, blog, wiki, photo, video, etc.) production and subsequent consumption is a key part of the social web experience. Every second, on average, several thousand tweets are tweeted on Twitter, which corresponds to several hundred million tweets per day and several hundred billion tweets per year.
Read more
Internship Projects
I developed survival models for studying the adoption dynamics of app gadgets, such as photo filters, augmented reality (AR) lenses, and in-situ games. The models provided insights into the adoption (micro-level) and diffusion (macro-level) of Snapchat lenses. The outcome of this project is currently under preparation.
I developed a deep learning workflow for profiling restaurants from payment card transactions. The workflow achieved a 49% improvement over baseline in identifying the cuisine types of US restaurants from Visa transactions. The outcome of this project includes a patent and a full paper at IEEE DSAA 2020.
I designed an outlier-aware pattern mining method for extracting user tasks from application logs. The method achieved a 38% improvement over baseline in identifying user tasks from Photoshop logs. The outcome of this project includes a workshop paper at IEEE VIS 2016 and a full paper at ACM IUI 2017.
Undergraduate Projects
I developed a crowd-sourced testbed that enables a social media user to get a real-time evaluation of his/her proposed post, before broadcasting it in the actual platform. The outcome of this project includes a poster at ACM CHI 2014, and a full paper at SocInfo 2015.
My undergraduate thesis involved community detection and privacy. I designed a community detection method based on active interactions among users. I also designed a method for privacy-preserving community detection. The outcome of this project includes two posters at ACM SIGMOD 2014, and a full paper at DASFAA 2014.
I designed a multi-provider cloud architecture to prevent data-mining based privacy attacks in clouds. I also designed a logging system that records computational provenance in clouds. The outcome of this project includes a workshop paper at SC 2012, and a workshop paper at DASFAA 2014.

Publications

  1. Profiling US Restaurants from Billions of Payment Card Transactions
    Himel Dev, and Hossein Hamooni
    The 7th IEEE International Conference on Data Science and Advanced Analytics (DSAA), Sydney, Australia, October, 2020.

  2. Quantifying Voter Biases in Online Platforms: An Instrumental Variable Approach   ★ Best Paper Honorable Mention
    Himel Dev, Karrie Karahalios, and Hari Sundaram
    The 22nd ACM Conference on Computer-Supported Cooperative Work and Social Computing (CSCW), Austin, USA, November, 2019.

  3. A Generative Model for Discovering Action-Based Roles and Community Role Compositions on Community Question Answering Platforms
    Chase Geigle, Himel Dev, Hari Sundaram, and ChengXiang Zhai
    The 13th International AAAI Conference on Web and Social Media (ICWSM), Munich, Germany, June, 2019.

  4. Avoiding Drill-down Fallacies with VisPilot: Assisted Exploration of Data Subsets
    Doris Jung-Lin Lee, Himel Dev, Huizi Hu, Hazem Elmeleegy, and Aditya Parameswaran
    The 24th ACM International Conference on Intelligent User Interfaces (IUI), Los Angeles, USA, March, 2019.

  5. The Size Conundrum: Why Online Knowledge Markets Can Fail at Scale
    Himel Dev, Chase Geigle, Qingtao Hu, Jiahui Zheng, and Hari Sundaram
    The 27th International Conference on World Wide Web (WWW), Lyon, France, April, 2018.

  6. Identifying Frequent User Tasks from Application Logs
    Himel Dev, and Zhicheng Liu
    The 22nd ACM International Conference on Intelligent User Interfaces (IUI), Limassol, Cyprus, March, 2017.

  7. A Real-Time Crowd-Powered Testbed for Content Assessment of Potential Social Media Posts
    Himel Dev, Mohammed Eunus Ali, Jalal Mahmud, Tanmoy Sen, Madhusudan Basak, and Rajshakhar Paul
    The 7th International Conference on Social Informatics (SocInfo), Beijing, China, December, 2015.

  8. User Interaction Based Community Detection in Online Social Networks
    Himel Dev, Mohammed Eunus Ali, and Tanzima Hashem
    The 19th International Conference on Database Systems for Advanced Applications (DASFAA), Bali, Indonesia, April, 2014.

Contact

Email: himeldev@gmail.com; hdev3@illinois.edu