Natural language processing – ETO Research Almanac

ETO Research Almanac

Between 2017 and 2022, researchers published roughly

135,000

English-language articles on natural language processing

The global natural language processing literature grew

111%

between 2017 and 2022

Articles on natural language processing made up

11%

of all articles on artificial intelligence

Authors from private companies contributed to

of natural language processing articles

What counts as "natural language processing" research?•How precise are these numbers?•What are the other caveats?

Explore relevant research in the Map of Science

Data last updated May 8, 2024

Vital signs

137,446
natural language processing articles were released between 2017 and 2022 (estd.)
111%
growth in research between 2017 and 2022
11%
of all artificial intelligence articles focused on natural language processing

Citation impact

19
citations for the average natural language processing article
17
citations for the average artificial intelligence article

Countries

20%
of natural language processing articles have American authors
19%
have Chinese authors
18%
have European authors

Country	Articles (2017-2021)	Growth (2017-2021)	Global share	Citations per article
United States	21,481	102%	20%	56.7
China	20,024	167%	19%	21.7
India	7,421	146%	7%	12.7

Top-cited articles

Attention Is All You Need
2017: arXiv. 91,291 citations
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
2018: arXiv. 73,542 citations
RoBERTa: A Robustly Optimized BERT Pretraining Approach
2019: arXiv. 17,724 citations
Language Models are Unsupervised Multitask Learners
2019. 14,624 citations
Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
2019: arXiv. 12,790 citations
Transformers: State-of-the-Art Natural Language Processing
2019: arXiv. 11,800 citations
Deep contextualized word representations
2018: arXiv. 11,400 citations
Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks
2019: arXiv. 7,902 citations
BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension
2019: arXiv. 7,823 citations
Improving Language Understanding by Generative Pre-Training
2018. 7,776 citations

Research organizations

Top producers of natural language processing research (by article count)

#1 Microsoft (United States)
#2 Carnegie Mellon University (United States)
#3 Google (United States)

Organization	Articles (2017-2022)	Citations per article	Country
Microsoft	1,710	75.1	United States
Carnegie Mellon University	1,384	78.2	United States
Google	1,361	279.6	United States

Patents and industry

5%
of natural language processing articles are cited in patents
6%
of artificial intelligence articles are cited in patents
8%
of natural language processing articles had authors from private companies (vs 6% of artificial intelligence articles in general)

Top industry organizations by publication count

Company	Articles (2017-2022)	Citations per article	Country
Microsoft	1,705	75.2	United States
Google	1,361	279.6	United States
Amazon	776	37.4	United States

The Research Almanac runs on bibliometric data from Clarivate Web of Science and other sources. For details on our sourcing, methodology, and limitations, see the tool documentation.