Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

A universal information theoretic approach to the identification of stopwords


One of the most widely used approaches in natural language processing and information retrieval is the so-called bag-of-words model. A common component of such methods is the removal of uninformative words, commonly referred to as stopwords. Currently, most practitioners use manually curated stopword lists. This approach is problematic because it cannot be readily generalized across knowledge domains or languages. As a result of the difficulty in rigorously defining stopwords, there have been few systematic studies on the effect of stopword removal on algorithm performance, which is reflected in the ongoing debate on whether to keep or remove stopwords. Here we address this challenge by formulating an information theoretic framework that automatically identifies uninformative words in a corpus. We show that our framework not only outperforms other stopword heuristics, but also allows for a substantial reduction of document size in applications of topic modelling. Our findings can be readily generalized to other bag-of-words-type approaches beyond language such as in the statistical analysis of transcriptomics, audio or image corpora.

This is a preview of subscription content, access via your institution

Access options

Rent or buy this article

Prices vary by article type



Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Using entropy as a universal measure to quantify the information content of a word.
Fig. 2: Identification of stopwords by thresholding of information content.
Fig. 3: Removal of information theoretic stopwords makes the topic model more accurate and stable.
Fig. 4: Universal improvement of topic model inference for different language corpora.
Fig. 5: Robustness of supervised classification accuracy with respect to removal of information theoretic stopwords.
Fig. 6: Application to data from scRNA-seq reveals ‘stopgenes’.

Similar content being viewed by others

Data availability

The text data are available in the public repository

Code availability

The code for this Article, along with an accompanying computational environment, is available in the public repository and is executable online as a Code Ocean capsule. Code for the calculation of the information theoretic measure \(I\) and for the experiments with topic models can be found at


  1. Manning, C. D. & Schütze, H. Foundations of Statistical Natural Language Processing (MIT Press, 1999).

  2. Evans, J. A. & Aceves, P. Machine translation: mining text for social theory. Ann. Rev. Sociol. 42, 21–50 (2016).

    Article  Google Scholar 

  3. Rebholz-Schuhmann, D., Oellrich, A. & Hoehndorf, R. Text-mining solutions for biomedical research: enabling integrative biology. Nat. Rev. Genet. 13, 829–839 (2012).

    Article  Google Scholar 

  4. García, S., Luengo, J. & Herrera, F. Data Preprocessing in Data Mining (Springer, 2014).

  5. Dasu, T. & Johnson, T. Exploratory Data Mining and Data Cleaning (John Wiley & Sons, 2003).

  6. Schoenfeld, B., Giraud-Carrier, C., Poggemann, M., Christensen, J. & Seppi, K. Preprocessor selection for machine learning pipelines. Preprint at (2018).

  7. Blei, D. M. Probabilistic topic models. Commun. ACM 55, 77–84 (2012).

    Article  Google Scholar 

  8. Boyd-Graber, J., Hu, Y. & Mimno, D. Applications of topic models. Found. Trends Inf. Retr. 11, 143–296 (2017).

    Article  Google Scholar 

  9. Luhn, H. P. The automatic creation of literature abstracts. IBM J. Res. Dev. 2, 159–165 (1958).

    Article  MathSciNet  Google Scholar 

  10. Rasmussen, E. in Encyclopedia of Database Systems (eds Liu, L. & Özsu, M. T.) (2009).

  11. McCallum, A. K. Mallet: a machine learning for language toolkit. (2002).

  12. Nothman, J., Qin, H. & Yurchak, R. Stop word lists in free open-source software packages. In Proc. Workshop for NLP Open Source Software (NLP-OSS) (eds Park, E. L. et al.) 7–12 (Association for Computational Linguistics, 2018).

  13. Lo, R. T.-W., He, B. & Ounis, I. Automatically building a stopword list for an information retrieval system. J. Digit. Inf. Manag. 5, 17–24 (2005).

  14. Zou, F., Wang, F. L., Deng, X., Han, S. & Wang, L. S. Automatic construction of Chinese stop word list. In Proc. 5th WSEAS International Conference on Applied Computer Science (ACOS’06) (Huang, W. et al.) 1009–1014 (World Scientific and Engineering Academy and Society, 2006).

  15. Salton, G. & Yang, C. S. On the specification of term values in automatic indexing. J. Doc. 29, 351–372 (1973).

    Article  Google Scholar 

  16. Blei, D. M., Ng, A. Y. & Jordan, M. I. Latent Dirichlet allocation. J. Mach. Learn. Res. 3, 993–1022 (2003).

    MATH  Google Scholar 

  17. Wang, C., Paisley, J. & Blei, D. M. Online variational inference for the hierarchical Dirichelet process. In Proc. 14th International Conference on Artificial Intelligence and Statistics, Proceedings of Machine Learning Research Vol. 15, 752–760 (AISTAT, 2011).

  18. Hoffman, M. D., Blei, D. M. & Bach, F. Online learning for latent dirichlet allocation. In Advances in Neural Information Processing Systems 23 (NIPS 2010) (eds Lafferty, J. D. et al.) 1–9 (Neural Information Processing Systems Foundation, 2010).

  19. Blei, D. M., Griffiths, T. L. & Jordan, M. I. The nested Chinese restaurant process and Bayesian nonparametric inference of topic hierarchies. J. ACM 57, 1–30 (2010).

    Article  MathSciNet  Google Scholar 

  20. Blei, D. M. & Mcauliffe, J. D. Supervised topic models. In Advances in Neural Information Processing Systems (eds Platt J. C. et al.) vol. 20, 121–128 (NIPS 2007).

  21. Achakulvisut, T., Acuna, D. E., Ruangrong, T. & Kording, K. Science concierge: A fast content-based recommendation system for scientific publications. PLoS ONE 11, e0158423 (2016).

    Article  Google Scholar 

  22. Schofield, A., Magnusson, M. & Mimno, D. Pulling out the stops: rethinking stopword removal for topic models. In Proc. 15th Conference of the European Chapter of the Association for Computational Linguistics (eds Lapata, M. et al.) Vol. 2, 432–436 (Association for Computational Linguistics, 2017).

  23. Montemurro, M. A. & Zanette, D. H. Towards the quantification of the semantic information encoded in written language. Adv. Complex Syst. 13, 135–153 (2010).

    Article  Google Scholar 

  24. Gries, S. T. Dispersions and adjusted frequencies in corpora. Int. J. Corpus Linguist. 13, 403–437 (2008).

    Article  Google Scholar 

  25. Zipf, G. K. Human Behaviour and the Principle of Least Effort (Addison-Wesley, 1949).

  26. Fan, A., Doshi-Velez, F. & Miratrix, L. Prior matters: simple and general methods for evaluating and improving topic quality in topic modeling. Preprint at (2017).

  27. Schofield, A. & Mimno, D. Comparing apples to apple: the effects of stemmers on topic models. Trans. Assoc. Comput. Linguist. 4, 287–300 (2016).

    Article  Google Scholar 

  28. Shi, H., Gerlach, M., Diersen, I., Downey, D. & Amaral, L. A new evaluation framework for topic modeling algorithms based on synthetic corpora. In Proc. Machine Learning Research Vol. 89 (eds. Chaudhuri, K. & Sugiyama, M.) 816–826 (PMLR, 2019).

  29. Peel, L., Larremore, D. B. & Clauset, A. The ground truth about metadata and community detection in networks. Sci. Adv. 3, e1602548 (2017).

    Article  Google Scholar 

  30. Lancichinetti, A. et al. High-reproducibility and high-accuracy method for automated topic classification. Phys. Rev. X 5, 011007 (2015).

    Google Scholar 

  31. Aggarwal, C. C. & Zhai, C. in Mining Text Data (eds. Aggarwal, C. C. & Zhai, C.) 77–128 (Springer, 2012).

  32. Uysal, A. K. & Gunal, S. The impact of preprocessing on text classification. Inf. Process. Manag. 50, 104–112 (2014).

    Article  Google Scholar 

  33. Skinnider, M. A., Squair, J. W. & Foster, L. J. Evaluating measures of association for single-cell transcriptomics. Nat. Methods 16, 381–386 (2019).

    Article  Google Scholar 

  34. Bravo González-Blas, C. et al. Cistopic: cis-regulatory topic modeling on single-cell ATAC-seq data. Nat. Methods 16, 397–400 (2019).

    Article  Google Scholar 

  35. Alberts, B. et al. Molecular Biology of the Cell Sixth International Student Edition (W. W. Norton & Co., 2014).

  36. Zheng, C. et al. Landscape of infiltrating T cells in liver cancer revealed by single-cell sequencing. Cell 169, 1342–1356.e16 (2017).

    Article  Google Scholar 

  37. Solé-Boldo, L. et al. Single-cell transcriptomes of the aging human skin reveal loss of fibroblast priming. Preprint at bioRxiv (2019).

  38. Mikolov, T., Sutskever, I., Chen, K., Corrado, G. S. & Dean, J. Distributed representations of words and phrases and their compositionality. In Advances in Neural Information Processing Systems 26 (eds Burges, C. J. C. et al.) 3111–3119 (Curran Associates, 2013).

  39. Pritchard, J. K., Stephens, M. & Donnelly, P. Inference of population structure using multilocus genotype data. Genetics 155, 945–959 (2000).

    Google Scholar 

  40. Broderick, T., Mackey, L., Paisley, J. & Jordan, M. I. Combinatorial clustering and the beta negative binomial process. IEEE Trans. Pattern Anal. Mach. Intell. 37, 290–306 (2015).

    Article  Google Scholar 

  41. Yan, X., Jeub, L. G. S., Flammini, A., Radicchi, F. & Fortunato, S. Weight thresholding on complex networks. Phys. Rev. E 98, 042304 (2018).

    Article  Google Scholar 

  42. Gerlach, M., Shi, H. & Amaral, L. A. N. Stopwords-filtering. Code Ocean (2019).

Download references


L.A.N.A. acknowledges a John and Leslie McQuown Gift to NICO and support from the Department of Defense Army Research Office (grant number W911NF-14-1-0259). M.G. thanks T. Stoeger and Z. Ren for insightful discussion on scRNA-seq.

Author information

Authors and Affiliations



M.G. and L.A.N.A. conceptualized the study. M.G. and H.S. obtained all data and conducted all analysis. M.G. and L.A.N.A. wrote the first draft. M.G., H.S. and L.A.N.A. edited and revised the manuscript.

Corresponding authors

Correspondence to Martin Gerlach or Luís A. Nunes Amaral.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Supplementary methods, notes, figures and references.

Rights and permissions

Reprints and Permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Gerlach, M., Shi, H. & Amaral, L.A.N. A universal information theoretic approach to the identification of stopwords. Nat Mach Intell 1, 606–612 (2019).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:

This article is cited by


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing