We will never have any AI (LLM) component inside Matilda to avoid that situation. Bibliographic search needs deterministic, transparent and reproducible processes. AI-training robots already come to the platform, which is fine, though we need to slow them down to avoid DDOS-style crashes.

Tonight at #OSFair2025 we perform two Matilda demos. Come and let's chat around #openscience in bibliographic search www.opensciencefair.eu/demos/matild...

You can now download the book on the publisher's website: press.umich.edu/Books/P/Publ... (open access of course)

A remarkable colleague, whose work is the reason why we chose "Matilda" for our open bibliographical platform. Read not only her "Matilda effect" article, but her extensive work on the history of women in science. #STS #historyofscience

Matilda is LLM-free so that you can always trace back any data/reference/publication to trusted sources. Of course we don't build fancy bios, we just give you the link to the author's ORCID page #openscience #opendata

Not sure how many scientists here have tried Claude Code or similar command line coding assistants. I had a complicated family property tax problem that was best solved by a brute force Monte Carlo simulation approach, so I spent a few days coding up and analyzing a model with Claude Code.

The beauty of full-text search in Matilda. Among 27M papers, at least 61 use "herbal language" (instead of natural language). #openscience #researchintegrity matilda.science/search/?l=fr...

Another manuscript with authors but not readers. Who can approve "herbal language" and "face popularity" in a paper? How can IEEE not be fined for fraud and boycotted by institutions and authors? #researchintegrity #openscience

Matilda.science hits new milestones! 151.3 M works, 254.5 M texts, 12.3 M authors, all connected in an ever-growing research graph. Citations in the graph: 174.6 M total, 41.1 M from full texts. And we’re still growing! #OpenScience #FAIRData @matilda-science.bsky.social @foxcubfr.bsky.social

We keep on growing, and that includes 26M publications with full-text search. No ID needed, everything is reusable and shareable #openscience #opendata

My PhD candidate Juan Pablo Bascur submitted one of his papers to MetaROR. The reviews of his paper have now been published on the platform. Very glad to see early-career researchers embracing new approaches to scientific publishing!

Matilda has just passed two milestones: 150M different works are included and you can full-text search 25 millions of them We aim at 200M and 50M a year from now. matilda.science?l=en #scientificpublishing #opendata

Matilda already catches a few "give a positive review only" manuscripts, including one where the authors updated the ArXiv version in 2025 (publication in 2022). Note that the prompt is being put just before the "Introduction" part @samuelmoore.org @gcabanac.cpesr.fr matilda.science/search/?l=fr...

TSOSI will hopefully become a key platform to proudly show investment in #Openscience #infrastructure

We just presented the soon available pipeline from Matilda to @opencitations.bsky.social as we will be a new source for the OC infrastructure. The presentation and abstract will be on @cern.bsky.social Zenodo. #WOOC2025

Very interesting talks, many on citation extractions (including posters soon to be presented). We collectively need to be able to extract references independently from publishers #WOOC2025

In today’s @scholarlykitchen.bsky.social I interview Laetitia Bracco and Eric Jeangirard from the French #OpenScience Monitor about their work tracking open scientific outputs in scholarly literature #FOSM #OSMI scholarlykitchen.sspnet.org/2025/05/27/s...

"After going through two rounds of peer review, which is typical, the journal asked her to do something unusual: replace the word “equitably” and take out data reporting demographic makeup. “They asked us to remove that,” she said, citing compliance with an executive order." Nuts!

Our 155M are deduplicated; the full number of publications is 245 M. You can also use full-text search over 22M works without any personal ID or tracking systems, share alerts and results. matilda.science?l=en

G Scholar is more and more tainted. At Matilda, we do not have the same problem: we only rely on trusted sources (Crossref, ArXiv, Pubmed, HAL,...) for pub. and we closely control our enrichments. So less publications and less citations but legit ones #researchintegrity #Barcelonadeclaration #ORI

In Matilda, you will find already 26 "squared blunder", including some, like this one, followed by the wrong acronym. #researchintegrity matilda.science/search/?l=fr...

who says that science doesn't generate profit?

Matilda has just passed the 20 million mark for full-text search! Still a long way to go, but the goal is to make full text the standard for bibliographical search rather than metadata. matilda.science?l=en #scientificpublishing #opendata

Censorship of locally-curated collections, readers and librarians may still use commercial tools. That is why we need open data everywhere, tagging preserved in many copies in various countries #openscience #openaccess

Have you ever wondered why information in tables is so difficult to pull out of PDF files? Our paper on this topic was just released by the journal of BigDataMining, which may have editors who do not pass the Turning test, but at the reviewers were great. Thank You! doi.org/10.1186/s130...

🚀 Matilda.science, developed by @foxcubfr.bsky.social, keeps growing! With 144.7M deduplicated works, 241.9M texts, and 11.8M mapped authors, we’re shaping the future of bibliometrics. #OpenScience #FAIRData @matilda-science.bsky.social

We don't know whether more than 35k science papers with "Donald Trump" in it is a sign of hope or despair #standupforscience matilda.science/search/?l=fr...

I just reviewed two papers. One was almost certainly written with heavy assistance from genAI. The other was full of missing, misleading and incorrect references, many apparently planted. I suspected as much from the abstracts. I hate that I have to spend time policing science. #researchintegrity

Matilda is among the organizations supporting the Barcelona Declaration on Open Research Information. A very fine collection of #openinfra and #openscience stakeholders barcelona-declaration.org/signatories/

The price of success for an open infrastructure: since Sunday, we have robots touching 60K works/hour and 15K author pages/hour. LLM machines? other stuff? #openscience #openinfra

Fake papers are contaminating the world’s scientific literature, fueling a corrupt industry and slowing legitimate lifesaving medical research @joelving.bsky.social @gcabanac.cpesr.fr and Cyril Labbé have the stories, the data, and this receipts in our partnership with @theconversation.com

Thrilled to announce this: Retraction Watch retractions now in the Crossref API.

Matilda embraces these criteria (the fourth to be implemented in 2026). Users should choose what is relevant to them, not databases. Providing open data, making everything reusable and exposable, giving you access to the freshest unfiltered academic literature.

Matilda was in maintenance yesterday. Back to normal with more than 82K new works to search & discover, more than 95K new full texts being ingested today and more than 1M new reference links. #openscience #opendata matilda.science?l=en

There are more then 300 academic articles which title contains "Happy new year" in Matilda. And some are not editorials (do you want to learn about the "Happy new year heart attacks?") matilda.science/search/?l=fr...

Side interest of full-text search. An example of a query to find LLM AI traces in 10M+ published academic texts @gcabanac.cpesr.fr @pubpeer.com #metascience (you can copy the permanent url or create an alert through the RSS tag to receive new results in Zotero) matilda.science/search/?l=fr...

OA increases availability and doesn't discriminate against those without wealth. Everybody benefits.

You can also discover that these vey odd papers are sometimes cited, which makes the citing papers a subject of inquiry @gcabanac.cpesr.fr

We wish to include pubpeer links in Matilda. Until then, if you don't know it, go and discover this very useful platform (and comment if you wish/can).

We made the open-data and transparent Google Scholar at Matilda: matilda.science?l=en. We will never use LLM AI to obfuscate results.

From an idea of Dan Brockington (thanks)

A starter pack on #openscience. Just some of the great researchers and experts I follow around here who are promoting open science/scholarship/research: go.bsky.app/U46tU6

We know that "search engines are not neutral," yet one can never be too careful, *especially* when one's health and wellbeing are at stake. A word of caution from "A Body Made of Glass" by Caroline Crampton:

We hope you can join us for our next free online seminar of the Wayfinders series - #4: a deep-dive into S2O, on 6 November, 4pm GMT. Full details can be found here: bit.ly/4f3xDuY Registration: bit.ly/registration... #openaccess #oa

It's also discussed in this 2015 paper Serenko, A. and Dumay, J. "Citation classics published in Knowledge Management journals. Part II: studying research trends and discovering the Google Scholar Effect", Journal of Knowledge Management, Vol. 19 No. 6, pp. 1335-1355. doi.org/10.1108/JKM-...

The deadline for these two postdoc positions is approaching. One job for someone with expertise in evidence review and another for a philosopher of science:

You will never find that kind of feature in Matilda. It is not made to give you (wrong) answers, but to lead you to read new papers, to discover the science you need. No "hallucinations", no fake abstract, no mixed authors. Just organised open data to help you find the right papers and share them.

As a cardinal example of linguistic diversity, follow the literature on the Matilda effect in eight European languages #bibliodiversity (feel free to add more in the search key) matilda.science/search/?l=fr...

Thed van Leeuwen new professor by special appointment for Monitoring Open Science Policies and Practices http://dlvr.it/TCj4tR