Hi there! I'm Upasana.
My research interests are primarily in computational social science, network science, and complex systems. Broadly, I am interested in developing computational models to understand what governs human behavior in social systems, such as the production and consumption of news, team and opinion dynamics, etc. I'm also interested in studying factors that lead to collective behavior, such as the spread of information, beliefs, norms, and diseases.
When I'm not working, I like practicing music, or trying out new recipes. And I love dogs! Thanks for stopping by :)
P.S. I am from India. My first name is pronounced like Oo-paa-sa-naa.
Publications
Sampling random graphs with specified degree sequences
Upasana Dutta, Bailey K. Fosdick, Aaron Clauset
Preprint https://arxiv.org/abs/2105.12120
Analyzing Twitter Users’ Behavior Before and After Contact by Russia’s Internet Research Agency
Upasana Dutta*, Rhett Hanscom*, Jason Shuo Zhang, Richard Han, Tamara Lehman, Qin Lv, Shivakant Mishra
In Proceedings of the ACM on Human-Computer Interaction, CSCW. PACM-HCI 2021, https://doi.org/10.1145/3449164, (* equal contribution), (* equal contribution)
Projects
Underrepresentation of rural undergraduate students in CU Boulder
Worked with CU Boulder Office of Data Analytics and CU Rural Network to analyse how undergraduate students from rural communities and small towns are underrepresented in CU Boulder.
Comparison of Dynamic and Linear Programming using Weighted Job Scheduling Problem
Compared Dynamic Programming and Linear Programming in solving the Weighted Job scheduling problem with increasing jobs and increasing hardness of the problem.
Study of user activity on Question-Answering Platform : Stack Exchange
Leveraged questioning-answering activities of users on Stack Exchange platform to study the latent community structure between various Q&A websites using Stochastic Block Modeling.
Movie Recommendation System : Kaggle Dataset
Using user demographics and the movies they have watched and rated, built a model to recommend movies to the users which they have not watched yet but are very likely to rate highly.
