about

Please call me Seva

(SEH-vah: “SEH” as in “set”, and “vah” rhymes with “spa”);.

here is my CV.

I study how to improve human-AI teamwork in teams and organizations. I combine the lenses of analytical sociology and human-computer interaction to understand the mechanisms of computer-supported collaborative work. To do that, I build virtual lab experiments where I can look at digital traces and ask ego-centric questions. Finally, I use social network analysis to untangle the social structure in human-AI teams. I am also passionate about the Open Science movement, especially the idea of integrative experiments. Recently, I have worked as a replicator for a massive open collaboration project and participated in a replication game at Northwestern.

Bio

In Summer 2022, I joined Technology and Social Behaviour at Northwestern University where I work as a research assistant at SONIC. I use social network analysis models to understand mechanism of interaction and communication in human-AI teams. Recently, I have been engaging with open science initiatives, like replication games, and massive collaboration projects.

Before that, I worked as a research assistant at Universitetet i Bergen. Together with Mohammad Khalil at SLATE, we created CERES – a virtuall lab to study course recommendation system. We even published a paper about the first iteration, but we also managed to launch and analyze a pilot experiment with real students. Furthermore, I worked on analyzing classroom intervention with mixed models and SEMs. (sumbitted to a Journal of Internet and Higher Education)

Also, I was a HCI student at HSE University. I worked with Ilya Musabirov, and my thesis received a research grant from Toloka to explore the effect of self nudges on motivation.

Last, but not by the amount of time it required, I worked as (1) an invited lecturer at Data Science minor at the same HSE University. The course was mostly a mix of CSS and intro to data science for 200 non-STEM students. We are talking about tidyverse, tidymodels, and shiny. In addition, I worked as (2, 3) an invited lecturer for a graduate-level courses: Small Data Science for Political Science students, and Intro to Experiments for HCI. Finally, (4) an invited co-lecturer for a Data Visualization course with Anastasia Kuznetsova at European University at St. Petersburg. (At some point, some materials will be available somewhere here)

Test of visnetwork

require(visNetwork, quietly = TRUE)

Warning: package 'visNetwork' was built under R version 4.3.3

# minimal example
nodes <- data.frame(id = 1:9,
label = c("Seva", "Social Network Analysis", "Analytical Sociology", "Virtual Labs",
"Deployment", "Design of Experiments", "ERGMs", "DyNAMs", "Micro-Macro"),
level = c(1, 2, 2, 2, 3, 3, 3, 3, 3))
edges <- data.frame(from = c(1,1, 1, 4, 4, 2, 2, 3), to = c(2, 3, 4, 5, 6, 7, 8, 9))
visNetwork(nodes, edges, width = "100%") |> 
visEdges(arrows = "to") |> 
visHierarchicalLayout()