Hostility on #Twitter increases after Jihadist terror attacks.
New study w/ @gorodzeisky in the Journal of #computationalSocialScience , analyzing ~4.5M Tweets from ~1.2M users before and after ten major attacks across five European countries. Available in #openaccess 🔓 at:
Our paper just won "Best Paper Engaged in Quantitative Description on an Under-studied Phenomenon" from JQDM for 2023! 🏆 I'm no longer in #ComputationalSocialScience work, but it's still dear to my heart, and this paper was a blast :) Spoiler: We learned a whole lot about #kpop
"Fame and Ultrafame: Measuring and comparing daily levels of ‘being talked about’ for United States’ presidents, their rivals, God, countries, and K-pop." https://doi.org/10.51685/jqd.2022.004
If you are curious about the novel results published in Science and Nature by the collaboration btw Meta and US academics, you might want to read #ComplexityThoughts
I try to build some background for non-experts of #ComputationalSocialScience and summarize the main findings.
Finally, I dig into the ongoing debate, covering existing documents and my chats with Sandra González-Bailón, Sander van der Linden, Pierluigi Sacco and others
This debate might be of high interest for self-organized decentralized platforms, such as #Mastodon and the #Fediverse in general
In my class on data visualization in the fall, I'll be introducing students to machine learning with text analysis. The first example uses album reviews to train a classifier by genre (rock, jazz, hip hop, folk). I contrast deductive (keyword coding) and inductive (token vectorization) approaches. The latter returns a classifier with 98% accuracy. This is the first LDA plot to demonstrate the classification.
Datasci.social is a server for researchers & practitioners in human-centric data science, broadly defined, like network science, computational social science, geospatial data science:
As part of a summer course, I'm teaching a few weeks about reproducible research workflows.
I'm planning on talking about Makefiles + READMEs + Git + file organization - what else should I include, and what are your favorite resources about reproducible research?