iiinova — where your digital self keeps evolving

Enes Altuncu @enesaltuncu.com.tr · 13/07/2025

It presents a universal post-processing approach to boost AKE methods. Our experiments on 17 benchmark datasets showed significant improvements across 5 AKE methods. Thanks all the co-authors (@jasonnurse.bsky.social,Yang Xu,Jie Guo, @hooklee.bsky.social) for their effort and collaboration.

Enes Altuncu @enesaltuncu.com.tr · 13/07/2025

📢 Happy to share that our work "Improving Performance of Automatic Keyword Extraction Methods Using PoS Tagging and Enhanced Semantic-Awareness" is now published in the special issue Information Extraction and Language Discourse Processing of MDPI's Information journal.

132

Enes Altuncu @enesaltuncu.com.tr · 01/07/2025

🎉 Starting a new chapter at Istanbul University-Cerrahpaşa! Delighted to join the Department of Computer Engineering as an academic. Looking forward to contributing to research and teaching in this vibrant academic community.

Enes Altuncu @enesaltuncu.com.tr · 21/05/2025

🎙 Akademik çalışmalarım artık sade bir dille podcast formatında! Sade-ce Bilim, yapay zeka ile oluşturulan iki kişilik sohbetlerle dahil olduğum araştırmaları daha anlaşılır bir biçimde sunuyor. 🎧İlk bölüm Spotify'da yayında: bit.ly/4mvKM4o #Bilimİletişimi #Podcast #SadeceBilim #BilimHerkesİçin

open.spotify.com

Sade-ce Bilim

Podcast · Dr. Enes Altuncu · Sade-ce Bilim, içerisinde yer aldığım akademik araştırmaların, herkesin anlayabileceği şekilde sadeleştirilerek sunulduğu bir podcast kanalıdır. Yapay zekâ destekli iki ki...

020

Our paper is a dataset paper with example applications to show its usefulness. So, we used an existing SOTA political bias classifier (based on previous research and used in other relevant studies) for simplicity. How to validate/improve the performance of this model was beyond our scope.

The predictions of the model based on previously published fact-checking reports says so. But, it's important to note that the reported findings are limited by the accuracy of the model.

100

We used a publicly available BERT-based pretrained model for political bias classification.

100

Those bars refer to the standard deviation of predicted bias scores. If a bar is shown on both sides of the centre line, it means the corresponding organisation published both left- and right-leaning reports.

000

🛠️ Example applications enabled by the dataset: • Detailed stats for 39 orgs + ~2k authors (activity span, avg. word counts, % unique claims) • Predicted political bias scores (via a pre-trained model) • Credibility assessment based on historical signals

300

✅ ~118k English fact-checking reports from 39 organisations: ✔️ All are active signatories of @factchecknet.bsky.social and/or EFCSN 🗓️ Spanning 1995–2025 💬 Includes 7k+ claims investigated by multiple organisations

100

Enes Altuncu @enesaltuncu.com.tr · 23/04/2025

🚨 The preprint and dataset are now available! 🔎 Here's what FACTors offers — a quick thread on the dataset and example applications 🧵⤵️

100

Another example of how a "Large Language Model" can be used as a "Large Lying Machine" ⤵️

Enes Altuncu @enesaltuncu.com.tr · 21/04/2025

This new decentralised verification approach on Bluesky sounds a good idea although its success depends on how Trusted Verifiers will be selected. I think it could be enhanced to enable organisational accounts to verify their employee's accounts with a separate badge/label showing the affiliation.

000

Enes Altuncu @enesaltuncu.com.tr · 05/04/2025

Many thanks to the co-authors (Can Başkent, Sanjay Bhattacherjee, Shujun Li, and Dwaipayan Roy) for their effort and collaboration.

000

Enes Altuncu @enesaltuncu.com.tr · 05/04/2025

Excited to share that our work "FACTors: A New Dataset for Studying the Fact-checking Ecosystem" has been accepted for #SIGIR2025 as a Resource & Reproducibility paper. More details coming soon! #academicresearch #factchecking #dataset sigir2025.dei.unipd.it

sigir2025.dei.unipd.it

SIGIR 2025, Padua, 13-18 July | Home

The 48th International ACM SIGIR Conference on Research and Development in Information Retrieval | July 13-18, 2025 in Padua, Italy

Enes Altuncu @enesaltuncu.com.tr · 04/04/2025

X'teki dezenformasyondan bıkıp kendini burada bulan yeni Bluesky göçmenleri için tekrar paylaşma gereği hissettim🙂

000

Enes Altuncu @enesaltuncu.com.tr · 19/01/2025

With new additions, including @johnfocook.bsky.social, @emmalbriant.bsky.social, @sanjanah.bsky.social, @misinforeview.bsky.social, and @zakird.com, the starter pack is now full! If you haven't checked it recently, a final look might be useful to discover more accounts 🙂

042

Enes Altuncu @enesaltuncu.com.tr · 04/01/2025

A worth reading threat to understand why misinformation research is needed ⤵️

042

Reposted by Enes Altuncu

David Lazer @davidlazer.bsky.social · 20/12/2024

Selecting a few of the recommendations from our National Academies report on science misinformation to highlight. Recommendation 2 is simply that universities & researchers need better procedures for publicizing research. There should be accountability for overhyping science.

42512

Reposted by Enes Altuncu

Maarten Schenk @mschenk.bsky.social · 16/12/2024

Check out listifications.app (aka @listifications.app). If you follow that account it automatically sends you a DM every time someone adds you to a list or Starter Pack. Fun to see randos blocking you as "fash" or "scum" while others add you to lists of "journalists" or "fact checkers" :-)

listifications.app

Listifications (@listifications.app)

Follow me to receive notifications whenever you're added to a starter pack, block list, feed or when someone blocks your account. Follow @unsub.blocks.listifications.app to opt out of direct block no...

182

Enes Altuncu @enesaltuncu.com.tr · 11/12/2024

If you're following @politifact.bsky.social, it's not the real account of Politifact ⤵️

011

Enes Altuncu @enesaltuncu.com.tr · 05/12/2024

I like the idea of using "information integrity" as an umbrella term. One reason is that it emphasises that information disorder is a cyber security issue since integrity is a key principle of cyber security.

030

Reposted by Enes Altuncu

Brent Lee @brentlee.bsky.social · 05/12/2024

Misinformation? Disinformation? What's the difference? A very simple and festive explainer. 🎅 Credit: 🎨 @felixuncia.bsky.social 🧠 @brentlee.bsky.social

A festive two tile comic strip illustrating the difference between misinformation and disinformation.

Tile A: Disinformation.
Scene: Christmas morning in the living room. A child looking at wrapped Christmas present which seems to have the outline of a bicycle. Parents stood in the background are telling the child "It's from Santa Claus!".

Tile B: Misinformation
Scene: Later that day, the child is riding his new bike outside. While passing his three friends standing on the sidewalk, he shouts "It's from Santa Claus!".

Credit: This comic strip is a collaboration between @brentlee.bsky.social & artist @felixuncia.bsky.social

371352245

Reposted by Enes Altuncu

Shayan Sardarizadeh @shayan86.bsky.social · 04/12/2024

THREAD: How to check if the screenshot of a social media post is real or fake? Fake screnshots of online posts regularly mislead people. So, here's a simple guide on how you can quickly check if a screenshot of a post attributed to an account is genuine or fake before falling for or sharing it.

66791455

Finally, I recommend following two relevant initiatives here for more resources: 1️⃣ @vera-ai.bsky.social provides AI-based tools that can be useful to spot deepfakes. 2️⃣ @deepfakestracker.org is a research portal offered by @socialmedialab.ca with relevant tools, games, and educational resources.

172

Pelosi videos manipulated to make her appear drunk are being shared on social media

⚠️Despite the deepfakes' high potential to harm, don't forget cheapfakes (or shallow fakes)! They are still more common (doi.org/10.1177/1940...) and convincing (doi.org/10.1093/ijpo...). Below shows a famous cheapfake produced solely by slowing down an existing video ⤵️ youtu.be/sDOo5nDJwgA?...

youtu.be

YouTube video by Washington Post

Deepfakes for good? How synthetic media is transforming business

While deepfakes are powerful tools for deceiving and misleading people, they are not necessarily harmful. Deepfakes have a variety of positive applications, ranging from education to marketing ⤵️ techinformed.com/deepfakes-fo...

techinformed.com

Despite the bad press, deep fakes do have positive applications. Dan Patterson reports on how the tech is transforming certain sectors.

How to Distinguish AI-Generated Images from Authentic Photographs

Detecting deepfakes is a cat-and-mouse game. As deepfakes become more advanced, so do the detection methods. This gives an overview of the available methods 👉 doi.org/10.1002/widm... Besides, here's a very useful guide from @medialab.bsky.social on how to spot deepfakes⤵️ arxiv.org/abs/2406.08651

arxiv.org

The high level of photorealism in state-of-the-art diffusion models like Midjourney, Stable Diffusion, and Firefly makes it difficult for untrained humans to distinguish between real photographs and A...

Nowadays, it is quite simple to generate deepfakes for specific purposes. For example, the image below was generated with thispersondoesnotexist.com and is totally fake! Such images are commonly used in fake profiles to convince other users of the profiles' authenticity.

A fake image generated using thispersondoesnotexist.com

100

Frontiers | Deepfake: definitions, performance metrics and standards, datasets, and a meta-review

Deepfake is a portmanteau of "deep learning" and "fake," referring to the use of deep learning algorithms to generate fake multimedia (images, videos, audio, and even text!). See our recent review for more discussions on the definition and scope of deepfake ⤵️ doi.org/10.3389/fdat...

doi.org

Recent advancements in AI, especially deep learning, have contributed to a significant increase in the creation of new realistic-looking synthetic media (vid...

100