💻 computer science

Machine Learning Research in India: Institutions, Citation Patterns, and Research Trajectories Among Top-Cited Scholars on Google Scholar

This paper introduces the Scholar-India-ML bibliometric dataset to analyze the citation patterns and research trajectories of 478 top-cited Indian machine learning scholars, revealing that the Indian Institute of Science and the Indian Statistical Institute outperform the IITs in citation impact while industry labs are emerging as significant contributors to the nation's AI ecosystem.

Original authors: Kunal Dhanda

Published 2026-06-26

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Kunal Dhanda

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine India's Machine Learning (ML) research scene as a massive, bustling orchestra. For a long time, everyone assumed the loudest, most famous section was the "IITs" (the Indian Institutes of Technology), which are like the prestigious, high-salary soloists everyone knows.

However, this paper acts like a sound engineer who just finished recording the whole orchestra and analyzing the audio. The engineer (Dr. Kunal Dhanda) looked at 478 top musicians (researchers) in India who play the "Machine Learning" instrument, checked how many people have listened to their music (citations), and found some surprising things about who is actually making the most noise.

Here is the breakdown of what the paper found, using simple analogies:

1. The Data Collection: "The Google Scholar Playlist"

The researcher didn't just guess who was famous. He went to Google Scholar (a giant library of academic papers) and asked it to list everyone in India who tagged themselves as a "Machine Learning" expert. He sorted them by how many times their work was cited (like counting how many times a song has been streamed).

The Cleanup: He started with 480 names but threw out two "fake" or "misplaced" entries: one was a particle physicist whose high numbers came from huge team projects (not ML), and one looked like someone trying to game the system.
The Result: He was left with a clean list of 478 real researchers with a combined 1.52 million citations.

2. The "Rich Get Richer" Effect (Gini Coefficient)

The paper measured inequality using something called the Gini coefficient (a score from 0 to 1).

The Metaphor: Imagine a pie. If everyone gets an equal slice, the score is 0. If one person eats the whole pie, it's 1.
The Finding: India's ML pie has a score of 0.53. This means the "rich get richer" is happening here, but it's not extreme. The top 5% of researchers (about 24 people) are responsible for nearly 31% of all the attention (citations). It's a bit like a music festival where a few headliners get most of the applause, but there are still plenty of other bands getting noticed.

3. The Big Surprise: Who is Actually the Loudest?

This is the paper's biggest twist. Everyone thinks the IITs are the kings of Indian research.

The Reality Check: The paper found that the Indian Institute of Science (IISc) actually outperforms the IITs in terms of total citations.
- IISc: Has 26 researchers but grabs 8.9% of the total citations.
- IITs: Has 16 researchers but only grabs 4.1% of the citations.
Why? The paper suggests IISc is more like a specialized research lab where everyone focuses purely on research, whereas IITs are massive universities with huge teaching loads. Also, many top IIT graduates leave India for jobs abroad, so the "stars" aren't always staying in the IIT system.

4. The "Legacy Giants": The Indian Statistical Institute (ISI)

There is a small group called ISI (Indian Statistical Institute).

The Metaphor: Think of them as veteran jazz musicians. They are a tiny group (only 8 researchers), but they contribute 6.1% of all citations.
Why? They have been doing this for decades, long before modern "Deep Learning" became popular. Their work on older, foundational topics (like pattern recognition) has been cited thousands of times over the years. They punch way above their weight because of their long history.

5. The Corporate Players: Big Tech Labs

The paper also looked at researchers working for companies like Google and Microsoft.

The Finding: Even though they make up a small part of the group (about 6%), they account for 7.7% of the citations.
The Metaphor: These are the Hollywood studios of the research world. They have big budgets and produce high-quality "blockbusters" (papers) that get a lot of attention. Google and Microsoft India are the main players here.

6. What Are They Playing? (Research Interests)

When the researcher looked at what these people actually study:

Computer Vision (teaching computers to "see" images) is the most popular genre.
Image Processing and Data Mining are close behind.
Interestingly, the "newest" trends like Large Language Models (the tech behind chatbots) aren't the top cited yet. This is likely because the people with the most citations have been working in the field for a long time, and their "classic hits" (older papers) still get the most plays.

7. The Limitations (What the Microphone Didn't Catch)

The author is honest about what this study can't tell us:

Self-Selection: This only counts people who bothered to make a Google Scholar profile. If a brilliant researcher in India doesn't use Google Scholar, they aren't in this orchestra.
Past vs. Present: The data shows where people work now, but not where they worked when they wrote their famous papers. A researcher might be at a university now but got famous while working at a company.
No Cause-and-Effect: We know IISc has high citations, but we don't know if IISc caused the high citations, or if high-citation researchers just chose to go there.

Summary

The paper paints a picture of India's Machine Learning scene that is more diverse than the public realizes. While the IITs are famous for their brand, the IISc and the ISI are actually the heavy hitters in terms of research impact. Meanwhile, Big Tech labs are quietly becoming major players. It's a landscape where a few institutions and companies hold a lot of the spotlight, but there is a healthy mix of legacy experts and modern innovators.