Paul Bose
pbose.bsky.social
Paul Bose
@pbose.bsky.social
Postdoc at University of Rome Tor Vergata. Interested in Applied Microeconomics with a focus on Political Economy.

http://www.paulbose.com/
Reposted by Paul Bose
Looks amazing!
To borrow another example, taken from the `dbreg` README: github.com/grantmcdermo...

Here I am running a fixed-effects regression on 180 million(!) row parquet dataset... and it completes **< 2 seconds**... on my laptop 🤯

This is powered by @duckdb.org under the hood.

#rstats #econsky
August 27, 2025 at 8:42 PM
How is economics doing in terms of representation of women and researchers with a region of origin outside North America or Europe?
I looked at the top 5% authors on RePEc in the last 12 years.
- Only 12% are women
- 25% a region of origin other than US/Canada/Europe

#EconSky #AcademicSky #PoliSky
July 2, 2025 at 3:06 PM
Need to predict people's gender or region of origin from their name for your research? Check out my python package "nametrace" which provides a simple modern API to do just that.
June 26, 2025 at 3:50 PM
It was great to be in Clermont Ferrand to present my work with @econom.bsky.social on local social media activity after refugee arrival. Thanks so much to the organizers and the other participants. It was a super interesting workshop!
June 25, 2025 at 3:22 PM
Reposted by Paul Bose
🚨3 days left to apply!
June 5, 2025 at 12:26 PM
Olivier Marie is presenting our paper (joint with Renske Stans) on social media salience of refugees and election effects in Linnaeus today. I am super excited that this paper ready to be presented more widely!

We study local twitter discourse around the timing of refugee arrival during 2015/2016.
May 27, 2025 at 9:22 AM
Are you using LLMs for your research and want to classify millions of text? This can be a very slow and expensive process. But it doesn't have to be. In my blogpost I explain how to use multiple GPUs and vLLM to analyse thousands of texts with an LLM super fast!

#econsky #polisky

lnkd.in/di9AbciU
May 20, 2025 at 4:38 PM
Want to learn how to finetune a large language model for your specific needs? I wrote a short post on how to train LLAMA for a classification task:
www.paulbose.com/thisandthat/...

#econsky #polisky
Finetuning an LLM to do classification | Paul Bose
LLMs are ubiquitous at this point. ChatGPT for example is constantly used to generate labels for unstructured data such as text. However using ChatGPT might be too expensive or not perform very well f...
www.paulbose.com
May 7, 2025 at 4:30 PM
Want to learn how to run your own version of Deepseek R1 or Meta's LLAMA model on a remote high-performance computing server? I wrote a brief blog post explaining how you can install and run the models using ollama.

www.paulbose.com/thisandthat/...

#econsky #polisky
February 26, 2025 at 10:15 AM
Reposted by Paul Bose
Last chance to submit a paper (Deadline Jan 31). 3rd CEPR Workshop on Media, Technology, Politics, and Society cepr.org/events/3rd-b...
3rd BOCCONI - CEPR Workshop on Media, Technology, Politics, and Society
Search the site
cepr.org
January 29, 2025 at 10:19 AM
Can we learn something about gender and age from a person's tweets? Yes! In a new project with
Lorenzo Luro, Mahyar Habibi, Dirk Hovy and Carlo Schwarz, we provide a labeled dataset of 20k Twitter users. We show that demographic classifiers using tweets outperform traditional tools.
March 14, 2024 at 12:10 PM