Ahrabhi Kathirgamalingam's Avatar

Ahrabhi Kathirgamalingam

@ahrabhikat.bsky.social

she/ her | Researcher at Center for Advanced Internet Studies (CAIS) @cais-research.bsky.social | PhD candidate at @compcommlab.bsky.social at U of Vienna | πŸ”Ž marginalization, bias, racism and polarization in media and CSS methods. ahrkat.github.io

492 Followers  |  545 Following  |  50 Posts  |  Joined: 12.10.2023  |  2.4036

Latest posts by ahrabhikat.bsky.social on Bluesky

Very excited to be at the Center for Advanced Internet Studies @cais-research.bsky.social this week to talk about bias in CSS.

Thanks to @ahrabhikat.bsky.social for hosting us πŸ‘©β€πŸ’».

08.08.2025 08:39 β€” πŸ‘ 18    πŸ” 6    πŸ’¬ 0    πŸ“Œ 0

I had a great time at @cais-research.bsky.social, everyone was so welcoming, and it was a pleasure to visit and work on this incredibly important and inspiring paper with @ahrabhikat.bsky.social, @aytalina.bsky.social, and @valeriehase.bsky.social !

08.08.2025 09:59 β€” πŸ‘ 10    πŸ” 3    πŸ’¬ 0    πŸ“Œ 0

It was a truly precious experience to talk and write about social bias in CSS in a safe and inclusive space @cais-research.bsky.social generously provided πŸ™
Grateful to these badass scholars @ahrabhikat.bsky.social, @janabernhard.bsky.social, and @valeriehase.bsky.social for the amazing collab πŸ™Œ

08.08.2025 13:54 β€” πŸ‘ 10    πŸ” 3    πŸ’¬ 0    πŸ“Œ 0
The foreground text reads: β€œStory – The AI Revolution Comes With the Exploitation of Gig Workers.” Background: Illustration showing people working at computers in a watercolor-like, artistic style. The individuals appear tired or focused, seated at desks with monitors and headsets.  Image credit: Kathryn Conrad & Digit – Better Images of AI  – CC BY 4.0.

The foreground text reads: β€œStory – The AI Revolution Comes With the Exploitation of Gig Workers.” Background: Illustration showing people working at computers in a watercolor-like, artistic style. The individuals appear tired or focused, seated at desks with monitors and headsets. Image credit: Kathryn Conrad & Digit – Better Images of AI – CC BY 4.0.

The Industrial Revolution produced not only technical innovation, but also social disruption. What is the situation with the current β€œAI revolution”? This report by our fellows examines working conditions in this context:
EN: algorithmwatch.org/en/ai-revolu...
DE: algorithmwatch.org/de/die-ki-re...

09.08.2025 07:26 β€” πŸ‘ 7    πŸ” 4    πŸ’¬ 0    πŸ“Œ 0

2 new job opportunities at the @cais-research.bsky.social:
- PhD student, team "Research Data & Methods" stellenangebote.cais-research.de/de?id=f50b76
- Researcher, DFG project "SocEnRep - Automation of Reproducibility in Economics and Social Sciences" stellenangebote.cais-research.de/de?id=f50b76

25.07.2025 09:59 β€” πŸ‘ 16    πŸ” 13    πŸ’¬ 1    πŸ“Œ 0

πŸ’―πŸ’―πŸ’―

08.08.2025 12:38 β€” πŸ‘ 2    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0
Preview
Question Form Matters: Examining Trust in Government through Open and Closed Survey Items Abstract. How we ask questions in surveys significantly impacts the answers. While previous research has examined differences between open-ended and closed

🚨 New publication! 🚨 Do people really trust the government? It depends on how you ask.
Happy to share that our paper on why Question Form Matters with KatharinaPfaff (@stawi-univie.bsky.social‬) is now out in @jssam.bsky.social πŸ“ŠπŸ‡¦πŸ‡Ή
πŸ”— academic.oup.com/jssam/advanc...

08.08.2025 10:30 β€” πŸ‘ 21    πŸ” 8    πŸ’¬ 2    πŸ“Œ 0
Post image

Felt very welcomed today @cais-research.bsky.social β˜€οΈ Thank you for inviting @ahrabhikat.bsky.social πŸ™Œ

06.08.2025 15:46 β€” πŸ‘ 6    πŸ” 3    πŸ’¬ 1    πŸ“Œ 0

Some feeling i can't shake: when we (as researchers) work with LLMs in the pipeline, there seems to be the idea - even among critical people - that LLMs have reasoning capabilities, we just need to understand in which way these abilities differ from human reasoning. I think this is potentially

05.08.2025 12:31 β€” πŸ‘ 4    πŸ” 1    πŸ’¬ 2    πŸ“Œ 0
Post image

Anti-immigrant bias in the German criminal justice system: who would have thought?

osf.io/preprints/so...

30.07.2025 11:45 β€” πŸ‘ 71    πŸ” 23    πŸ’¬ 3    πŸ“Œ 2
Preview
Non-White scientists appear on fewer editorial boards, spend more time under review, and receive fewer citations | PNAS Disparities continue to pose major challenges in various aspects of science. One such aspect is editorial board composition, which has been shown t...

"Non-White scientists appear on fewer editorial boards, spend more time under review, and receive fewer citations"

www.pnas.org/doi/10.1073/...

01.08.2025 23:47 β€” πŸ‘ 16    πŸ” 14    πŸ’¬ 3    πŸ“Œ 4
Preview
Wie mit KI-Videos Rassismus verbreitet wird TikTok wird mit KI-generierten Videos ΓΌberschwemmt - die Fakes erreichen teils Millionen Menschen.

Wie mit KI-Videos Rassismus verbreitet wird #KI #Rassismus

02.08.2025 06:26 β€” πŸ‘ 29    πŸ” 13    πŸ’¬ 5    πŸ“Œ 3

πŸ‘‹ #ACL2025NLP πŸ‡¦πŸ‡Ή @marlutz.bsky.social and I are presenting our poster on demographic representativeness of LLMs today!

πŸ•¦ 10:30-12:00
πŸ“ Hall X5 (board 1 or 14 according to different sources 🧐)

Here’s the paper on ACL anthology: aclanthology.org/2025.finding...

Drop by!

29.07.2025 07:31 β€” πŸ‘ 20    πŸ” 7    πŸ’¬ 0    πŸ“Œ 1
Preview
Gesichtserkennung und KI: Innenminister Dobrindt plant neues Sicherheitspaket Die Polizei soll bald auch Gesichter-Suchmaschinen wie Clearview und Überwachungs-KI wie Palantir nutzen. Wir verâffentlichen das neue Sicherheitspaket des Innenministeriums. Die Zivilgesellschaft kri...

Nicht nur die Polizei soll mit Gesichtserkennung VerdΓ€chtige, Kontaktpersonen, Opfer und Zeugen suchen. Auch das Bundesamt fΓΌr Migration und FlΓΌchtlinge soll Fotos, die es von Asylsuchenden aufnimmt, mit Gesichter-Suchmaschinen im Internet abgleichen. netzpolitik.org/2025/gesicht...

25.07.2025 16:17 β€” πŸ‘ 73    πŸ” 34    πŸ’¬ 7    πŸ“Œ 4
Post image

Thinking about using #CSS methods to study #racism, #stereotypes or #hate speech in text? πŸ“

πŸ‘‰ Check out my first dissertation paper co-authored by @fabiennelind.bsky.social and @hajoboo.bsky.social just published in Annals of the ICA! @icahdq.bsky.social πŸ₯³

πŸ”— doi.org/10.1093/annc...

21.07.2025 06:43 β€” πŸ‘ 86    πŸ” 37    πŸ’¬ 2    πŸ“Œ 1
Preview
Mapping and Measuring Listening in Public Communication Settings | Article | Media and Communication Tariq Choucair

Congratulations to @tariqchoucair.bsky.social on his newly published paper, "Mapping and Measuring Listening in Public Communication Settings" in Media & Communication.

Tariq outlines the key elements of listening and how it is critical for political communication.

www.cogitatiopress.c...

24.07.2025 05:14 β€” πŸ‘ 11    πŸ” 4    πŸ’¬ 0    πŸ“Œ 0

Before heading to ACL, I'm excited to be at #IC2S2 this week! 🌞

I'll present a related working paper on validating LLM social simulations at the ABM session on Tuesday (11 AM, Vingen 7): indiiigo.github.io/files/GABM_V...

(w/ @wanlo.bsky.social @mstrohm.bsky.social and @janalasser.bsky.social)

21.07.2025 20:00 β€” πŸ‘ 17    πŸ” 6    πŸ’¬ 0    πŸ“Œ 0

Merci! β€οΈπŸ™ƒ

22.07.2025 17:37 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0
Post image

3.2/ Very simplified illustration of the four measurement pipelines!

21.07.2025 06:52 β€” πŸ‘ 4    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

@compcommlab.bsky.social @ipk-univie.bsky.social @cais-research.bsky.social

21.07.2025 06:51 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

5/ We offer many recommendations (check out Table 5) such as:

- Use clear terminology and theoretical grounding
- Embrace intersectionality, and go comparative
- Go beyond X and English
- Share own data (when safe)
- Explore existing datasets (check out Table A1)
- And of course: validate! πŸ€“

21.07.2025 06:44 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

4/ Some other key findings:

- Variety of concepts or variety of terminology?
- Limited intersectionality
- English data and social media dominate
- Heavy use of a few secondary datasets
- Lack of validation of data collection methods
- Strong preference for supervised classification

21.07.2025 06:44 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

3/ Based on our review, we identified 4 common measurement pipelines:

πŸ”Ή Top-down classification using predefined labels
πŸ”Ή Sentiment analysis paired with out-group keywords
πŸ”Ή Inductive clustering through topic modeling
πŸ”Ή Association-based approach using word co-occurrence or embeddings

21.07.2025 06:43 β€” πŸ‘ 3    πŸ” 0    πŸ’¬ 2    πŸ“Œ 0

2/ We mapped:

- What concepts were measured 🧠
- What data was used and how it was selectedπŸ“±
- What methods were applied πŸ› οΈ
- How validation was done βœ”οΈ

21.07.2025 06:43 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

1/ Racism, stereotypes & hate speech are increasingly studied using comp methods. But how do scholars actually measure these concepts in text? We reviewed 115 publications to find out. By that, we fill a critical gap and provide a list of recommendations on conceptualization, data, and methods! πŸ“š

21.07.2025 06:43 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0
Post image

Thinking about using #CSS methods to study #racism, #stereotypes or #hate speech in text? πŸ“

πŸ‘‰ Check out my first dissertation paper co-authored by @fabiennelind.bsky.social and @hajoboo.bsky.social just published in Annals of the ICA! @icahdq.bsky.social πŸ₯³

πŸ”— doi.org/10.1093/annc...

21.07.2025 06:43 β€” πŸ‘ 86    πŸ” 37    πŸ’¬ 2    πŸ“Œ 1
Preview
DHS Faces New Pressure Over DNA Taken From Immigrant Children The US government has added the DNA of approximately 133,000 migrant children and teens to a criminal database, which critics say could mean police treat them like suspects β€œindefinitely.”

The US government has added the DNA of approximately 133,000 migrant children and teens to a criminal database, which critics say could mean police treat them like suspects β€œindefinitely.”

20.07.2025 20:45 β€” πŸ‘ 999    πŸ” 574    πŸ’¬ 117    πŸ“Œ 79
Post image

New #HSR Special Issue
β€œTowards an Empirical Theory of Science?” looks at how science is evolving across disciplines and cultures. From objectivity to knowledge transfer, this issue reflects on science's shifting norms.
Read more πŸ‘‰ www.gesis.org/hsr/ak...
#historialsocialresearch #GESISpublications

16.07.2025 07:00 β€” πŸ‘ 2    πŸ” 1    πŸ’¬ 0    πŸ“Œ 0
Preview
Features of disinformation: an expert interview study on the perception of disinformation among political, governmental, media and business elites in Germany To curb the spread of disinformation, societal decision-makers are working intensively on the design of effective intervention strategies. Primarily elites from the sectors of politics, administrat...

✨New Publication in JEPOP✨
How do societal stakeholders understand and detect disinformation?

www.tandfonline.com/doi/full/10....

14.07.2025 15:49 β€” πŸ‘ 8    πŸ” 7    πŸ’¬ 1    πŸ“Œ 0
DHS Twitter account: 
FIRE UP THE DEPORTATION PLANES!

Image of a skeleton doing squats with the text MY BODY IS A
THAT TURNS
MACHINE
ICE FUNDING
INTO
MASS DEPORTATIONS

DHS Twitter account: FIRE UP THE DEPORTATION PLANES! Image of a skeleton doing squats with the text MY BODY IS A THAT TURNS MACHINE ICE FUNDING INTO MASS DEPORTATIONS

This is a real tweet from the U.S. Department of Homeland Security.

09.07.2025 22:29 β€” πŸ‘ 6668    πŸ” 2271    πŸ’¬ 882    πŸ“Œ 759

@ahrabhikat is following 20 prominent accounts