Andreu Casas

🚨 One more week to apply to the PhD position at the London Social Media Observatory. If you are interested in empirical, quantitative, and/or computational approaches to studying social media and politics, apply! 😊

🎉 We're really excited to be collaborating with our excellent partners on this, as part of WP3 of the What-If Horizon project, hosted at the Observatory.

🎯 Understanding what interventions affect citizen misinformation effects is crucial to build evidence based policy in the area!

More insights 🔜

The candidate will also have the opportunity to work in other international projects, with societal partners and policymakers, network with top scholars in the field, and the opportunity to teach in our department from the second year onwards.

The candidate will join a very exciting work environment, work side by side with me and the team in our central London location, and have access to funding for conferences, training, and data collection, on top of the stipend. Help me spread the word!

🚨Hiring a fully funded (3.5 years) PhD for the @ldnsocmedobs.bsky.social to research social media and politics. Candidates should have quantitative/computational skills and/or be interested in content curation/moderation. UK home candidates only unfortunately. www.royalholloway.ac.uk/media/hquftp...

All EU countries appear to be gone in the Google political ad library

By @lizcarolan.com

I hope I’m wrong and the UK government is still committed to transparency, academic access to tech/platform data, and independent/external scrutiny of the impact of these products on society. Time will tell I suppose. Meanwhile, we’ll keep advocating for what we think is right.

The Memorandum of Understanding from the UK government about the deal I fear confirms some of these suspicions, with language such as “advancing pro-innovation AI policy frameworks and efforts to support U.S. and UK-led AI technology adoption” (full memo here: lnkd.in/dw3NYvGz).

The US-UK tech deal last week left me worried. There was already a feeling the UK government was abandoning the idea of regulating tech companies/platforms, to facilitate trade deals with the US, incl legislation for better academic data access, and external scrutiny of tech products and platforms.

We went to Berlin this week to learn more from our peers about ways to leverage DSA to access platfrom data and conduct relevant research, and to keep building a coalition to advocate for this topic in the EU, the UK, and everywhere.

Very happy to have people from our @ldnsocmedobs.bsky.social attend the DSA40 Data Access Days organized by @dsa40collaboratory.bsky.social. With threats to our online and political safety in the rise, access to platform data for independent and academic researchers is more important than ever.

What a pity that my usual partner in crime on this one couldn’t make it this time around 🥺 @feloe.bsky.social we really missed you!

And as usual, the highlight of the week was to learn about the participants’ projects and to discuss with them how to take them to the next level. Here a few pictures from our traditional final session “Slides with Slices” 🍕 where participants presented the projects they worked on during the week.

Thanks @gesistraining.bsky.social for having me again. @marlenemauk.bsky.social @verenakunz.bsky.social and the team always make me feel at home in Mannheim for one week a year. Is Mannheim growing on me? 🤔

🥰I had so much fun again teaching computer vision and other computational methods to these bright social scientists this week. (Un)supervised classification, image encoding, content similarity, object detection, face detection/analysis, muldimodal modeling, video data, etc., you name it!

To junior scholars, the Observatory is the perfect home for computational social scientists to grow and shine! I’ll be soon recruiting a data engineer, PhD, and 2 postdocs, for our new offices in Bedford Square. To academics/orgs working in this area, we are eager to collaborate, please reach out!

Additionally, the Observatory will engage with key stakeholders in this area (@ofcom.bsky.social, @electoralcommission.org.uk, platforms, and others) to promote evidence-based policymaking in this area. Stay tuned for a first public event that we are putting together for the first week of December.

🚨Very honored to receive the Future Leaders Fellowship from @ukri.org I’ll use this 2.3 million £ investment to create the London Social Media Observatory, for cutting-edge research on social media and politics, and novel computational tools for social science research. www.ukri.org/news/ukri-an...

📣 Last Call!
We still have spots left for our #GESISfallseminar course on Computer Vision for Image and Video Data Analysis with @andreucasas.bsky.social.

Join us before it’s too late!⏳
➡️ t1p.de/ImageVideoAn...

You'll be able to master all of these and more, with a combination of lectures, hands on sessions where code is provided, and with one-on-one consulting sessions. Students will also work on a project throughout the week. Bring your own data and project, and use that week to take it to the next level

I will cover a wide range of relevant topics: video/image processing, supervised classification, unsupervised classification, facial analysis, multimodal modeling, Large Visual Language Models, cloud computing, etc.

Do you want to use large amounts of visual data (or text + visuals) in your research but don't know how? Join my "Computer Vision for Image and Video Analysis" course at GESIS this September (15-19th) to learn about machine learning for automatic visual analysis: training.gesis.org?site=pDetail...

This is all work in progress ( /!\ ), so all feedback welcome. Come and share your thoughts!

Friday (11:20-13:00: 1A.12) Large Visual Language Models for Supervised Multimodal Classification in Political Science Research. We compare 2 VLMs to 6 other supervised classifiers (incl. LLMs: Llama2&3), across 2 datasets and 10 tasks. VLMs outperform all models across all tasks, w. some caveats.

Thursday (13:10-14:50: 1A.13): Moderation of Political Content on YouTube during the 2024 US Election. I tracked 20k channels and 6mil videos. Conservatives suspended more often, but also post most of hateful/misinfo. No difference in visibility, w. higher engagement for conservatives.

On my way to Madrid-EPSA! Please join in if you want to hear about (political biases? in) content moderation on YouTube during the 2024 US Election, and/or about using open-source Large Visual Language Models (VLMs) for supervised multimodal classification in political science research.

Interestingly, we do not see Presidents to have much of an influence (neither Biden nor Trump).
See above a great summary from the CSMaP comms team. Many years in the making. Thank you so much to the great co-authors, specially Oscar Stuhler (@oms279.bsky.social), my partner-in-crime on this one!

🚨New Publication @thejop.bsky.social We study who sets the issue agenda of state policymakers in the US. We find a mixed picture: they respond to shifts in attention by members of Congress, but also to the constituents in their states, particularly their own party supporters.

@thejop.bsky.social

We discuss the problems that emerge for research, and put forward solutions. Beyond academia, the findings emphasize that AI models trained on human data can replicate cultural, ideological, gender, etc. biases; and call for closer attention and more diversity in AI development.