🏴☠️ We just released a new version of Zeeschuimer, fixing various capture bugs.
⚠️ If you are already using Zeeschuimer it will unfortunately NOT be able to automatically update to this new version. Please re-install from tools.digitalmethods.net/zeeschuimer instead. Apologies for the inconvenience!
A small 4CAT update just in time for the holidays - v1.52!
📊 Processors for regression analysis and dataset statistics
🧭 Test LLM prompts from our prompt library with the PromptCompass processor
🤖 Use Deepseek and Gemini 3 APIs with LLM processors
🐛 Many bug fixes!
➡️ github.com/digitalmetho...
4CAT now contains everything for quite sophisticated classification pipelines with humans and (local) LLMs!
I myself used it to classify what types of migration were mentioned in Dutch news articles (see www.groene.nl/artikel/onde...).
See all changes below 👇
Just released: 4CAT 1.51 “The R is in the month”
Highlights:
📊 Classification evaluation: subpar F1 scores, now in 4CAT
🤖 New LLM processor options
#️⃣ Search by #hashtag on Telegram
🖼️ Image hasher: check for duplicate images!
🪰 A LOT of small bug fixes and improvements
github.com/digitalmetho...
Hi Baptiste, yes, you can scrape messages from open Telegram groups with 4CAT's Telegram data source. You'll find more information when going to "Create dataset" and selecting "Telegram" as a data source. This requires a valid API key.
🌍 Web interface enhancements:
• Extension management via the web app
• New 'Jobs' page in the control panel
• Faster page loads for large datasets
• Dataset parameters are now copyable
• Expanded API functionality
⚙️ Misc processor improvements:
• More processors can now write annotations, like the Whisper processor for audio to text.
• Annotations are carried over when filtering
• Improved proxy handling for image and metadata downloads
• Updated language resources for tokenisation
🎞️ New: 'Video Wall' processor
Create collages from video datasets.
Ideal for exploration and presentation of audiovisual materials (and to overstimulate yourself with Italian Brainrot memes).
📈 Other new processors:
• 'Confusion Matrix' for evaluation
• 'Replace Text' to change any text value
🧠 New: 'LLM Prompter' processor
A much more robust and flexible way to analyse datasets with LLMs
• More APIs: OpenAI, Google, Anthropic, and Mistral
• Support for LM Studio and Ollama to use local (and free) LLMs
• Batching of input values to increase speed and reduce cost
• Structured outputs
🐈⬛ 4CAT 1.50 is here! 🐈⬛
This release brings under-the-hood improvements, a faster web interface, new processors, better LLM analyses, improved annotations, and lots of bugfixes.
🔗 Full changelog: github.com/digitalmetho...
🔗 How to upgrade: github.com/digitalmetho...
4CAT Explorer
🔎The Explorer lets you examine datasets in a rich environment where social media posts are "reassembled" to look like their original appearance.
You can sort data, make manual annotations like field notes, or let (LLM-)processors add categories and notes.
How to use: github.com/digitalmetho...
🔔New 4CAT version: 1.49🔔
🔎 Explorer revamp: Investigate datasets through an interface that mimics the original data source. Sort your data, make manual field notes, or add processor-made annotations.
🧰 Data source updates
🦜 Zeeschuimer import fixes
Full changelog: github.com/digitalmetho...
Hi John, you can upload JSON files as custom datasets. Many processors also output JSON. Hope that helps!
🦜 New Zeeschuimer release 🦜
📌📕 New modules for Pinterest and RedNote/Xiaohongshu!
Capture data from these platforms directly with Zeeschuimer—and export them to the newest version of 4CAT (also out now) for further analysis.
github.com/digitalmetho...
Updates automatically if installed already!
🔔 We just released a new 4CAT version, v1.47!
Some highlights:
📥 Data capture from Bluesky, Pinterest and RedNote/Xiaohongshu
🤖 Prompt-based coding and annotation of data via the OpenAI API
🛠️ New & updated data fields for datasets from TikTok, Instagram, & co
Read more: github.com/digitalmetho...
We're pleased to announce the launch of KingsCAT 🐱, a new @4cat.nl instance to support social media research at King's created by
@digitalmethods.net @oilab.bsky.social
hosted by @kingsdh.bsky.social @digitalculturekcl.bsky.social @kingsartshums.bsky.social www.kcl.ac.uk/news/gather-... 🎊✨
This extensive website on computational social science by Michael Achmann-Denkler provides 4CAT and Zeeschuimer tutorials alongside other great tips & tricks.
Check it out 👇
social-media-lab.net
Screenshot from the workshop titled “how does TikTok organise its data”, with the subheadings “for you page”, “user”, “hashtag”, “keyword search” and “sound”
Fascinating workshop with @4cat.nl walking us through their social media capture and analysis tool, talking through capabilities and possibilities!
@kingsdh.bsky.social
KingsCAT😸👑
🔔 New version out for Zeeschuimer, our tool for capturing social media data from the browser! This brings it up to speed with Instagram's latest updates, fixing the capture of posts from Location and Tag pages.
💾 Get it here: github.com/digitalmetho...
⬆️ Already installed it? It will auto-update!
and... you see these Bluesky updates on the "About" page! Feel free to get in touch.
🍂 Release v1.46: Autumn Additions 🍂
- Add custom data sources and processors as 'extensions' (see t.ly/WQc4Z)
- Import Threads data from Zeeschuimer
- Import and export datasets as zip files
- Crawl Telegram based on mentions and links
- LLM processor for text classification
And more: t.ly/psaqe
4CAT's original look (circa 2018).
4CAT's original look (circa 2018).
[testing our new bsky integration...]
Hellooo world! We'll use this profile for informal updates and questions regarding the 4CAT: Capture and Analysis Toolkit.
If you want to check out what's new with 4CAT outside of BlueSky, check out the GitHub releases github.com/digitalmetho...
Stay tuned for exciting new features 🐈🐈🐈🐈
