--- title: "For the sake of simplicity — Yatskar 2010" description: "The paper tackles automatic discovery of lexical simplifications, meaning word- or phrase-level rewrites into easier wording, such as \"collaborate\" →…" authors: ["Mark Yatskar", "Bo Pang", "Cristian Danescu-Niculescu-Mizil", "Lillian Lee"] year: 2010 venue: "Human Language Technologies: The 2010 Annual Conference of the North American Chapter of the Association for Computational Linguistics" source: "https://aclanthology.org/N10-1056/" licence: "CC BY-NC-SA 3.0 (ACL Anthology, before 2016)" licence_url: "https://creativecommons.org/licenses/by-nc-sa/3.0/" summary_licence: "Elamur Cosmic License" canonical: "https://research.elamur.ai/free/en/papers/yatskar-2010-sake-simplicity-unsupervised-extraction-lexical/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/yatskar-2010-sake-simplicity-unsupervised-extraction-lexical/) [home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2010](https://md.research.elamur.ai/free/en/papers/2010/) / For the sake of simplicity: Unsupervised extraction of lexical simplifications from Wikipedia # For the sake of simplicity: Unsupervised extraction of lexical simplifications from Wikipedia **Authors:** [Mark Yatskar](https://md.research.elamur.ai/free/en/authors/mark-yatskar/), [Bo Pang](https://md.research.elamur.ai/free/en/authors/bo-pang/), [Cristian Danescu-Niculescu-Mizil](https://md.research.elamur.ai/free/en/authors/cristian-danescu-niculescu-mizil/), [Lillian Lee](https://md.research.elamur.ai/free/en/authors/lillian-lee/) · **Year:** 2010 · **Venue:** Human Language Technologies: The 2010 Annual Conference of the North American Chapter of the Association for Computational Linguistics · **Licence:** [CC BY-NC-SA 3.0 (ACL Anthology, before 2016)](https://creativecommons.org/licenses/by-nc-sa/3.0/) **Links:** [Publisher's page](https://aclanthology.org/N10-1056/) · [PDF](https://aclanthology.org/N10-1056.pdf) · [OpenAlex](https://openalex.org/W1638924786) **Citations:** 127 citations (openalex-snapshot · 2026-10-14) · [82](#cited-by) in our library, of the 140884 papers whose bibliographies we read ## Summary The paper tackles automatic discovery of lexical simplifications, meaning word- or phrase-level rewrites into easier wording, such as "collaborate" → "work together". Instead of hand-written syntactic rules, the authors mine the revision histories of Simple English Wikipedia (SimpleEW). The main difficulty is that most SimpleEW edits are not simplifications. Many are grammar or factual fixes, or spam. Two unsupervised strategies address this noise: - **Edit model.** This probabilistic model treats each observed change as a mixture of operations. Fix rates are estimated from regular English Wikipedia (ComplexEW), where simplification is assumed to be rare. These rates are then subtracted from SimpleEW change rates to isolate simplification probabilities. - **SIMPL.** This method uses revision comments matching *simpl* to select trusted edits, then ranks them by PMI. In human evaluation of the top 100 pairs, the edit model reached 77% precision and SIMPL reached 66%. Both far exceeded frequency and random baselines (17%). Both also found many good simplifications missing from a manually compiled list. The work shows that noisy collaborative edit logs can yield directional lexical resources without supervision. ## Key points - **Data:** about 1.5M revisions of 81,733 SimpleEW articles (November 2009 snapshot, only about 30% with textual changes) and about 16M revisions of 19,407 ComplexEW articles (January 2008 snapshot). - **Edit extraction:** - Adjacent article versions were sentence-aligned with tf-idf, following Nelken and Shieber. - Lexical edit instances were taken as the longest differing word segments. - Pairs with more than five words on either side were discarded. - **Edit model estimate:** the simplification probability of a phrase is estimated as max(0, f_S(A) − α·f_C(A)). Here f is the fraction of articles containing the phrase in which it gets edited, and α is a damping factor between 0 and 1. Spam is assumed to have zero probability in this initial version. - **Precision@100 (native-speaker majority labels):** SpList 86%, edit model 77%, SIMPL 66%, FREQUENT 17%, RANDOM 17%. The methods produced 1,079 (edit model) and 2,970 (SIMPL) pairs, versus 2,000 in SpList. - **Annotation:** three native and three non-native English speakers labelled the pairs. Agreement was κ = 0.69 among native speakers and κ = 0.49 among non-native speakers. The non-native annotators also reported needing a dictionary. ## Interesting ideas - **Use a "control" corpus to subtract background noise.** ComplexEW serves as a reference where only ordinary fixes happen, so its edit rates can be removed from SimpleEW rates to expose the target operation. The same contrastive trick could isolate any style-specific edit type wherever a comparable "neutral" edit stream exists. - **Count documents, not revisions.** Edit propensity is measured over whole articles rather than individual versions. A phrase that survives many minor revisions is therefore not wrongly judged stable. This is a useful normalization for any revision-history mining. - **Treat editor comments as weak supervision.** A single regex over revision comments cheaply selects high-precision training edits. Commit messages or edit summaries in other wikis and repositories could play a similar role for other edit intents. - **Show annotation pairs in random direction.** Annotators saw each pair in either orientation and chose among "simpler", "more complex", "equal" and similar labels. This hides the system's claimed direction and reduces annotation bias. The approach suits evaluating any directional relation, such as entailment or formality. - **Treat a manual list as a comparison point, not a gold standard.** The hand-built SpList was used for agreement rather than recall, because it was known to be incomplete. Mixing its samples into the blind evaluation gave a human-quality reference point for the system scores. *Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.* Fields: [Natural language processing](https://md.research.elamur.ai/free/en/shelves/nlp/) ## Cited by Cited by 82 papers of our library cite it (of the 140884 whose bibliographies we read); 13 of them have a page here, 69 not yet. - [Putting it Simply: a Context-Aware Approach to Lexical Simplification](https://md.research.elamur.ai/free/en/papers/biran-2011-putting-it-simply-context-aware/) - [Evaluating prose style transfer with the Bible](https://md.research.elamur.ai/free/en/papers/carlson-2017-evaluating-prose-style-transfer-bible/) - [Learning to Simplify Sentences Using Wikipedia](https://md.research.elamur.ai/free/en/papers/coster-2011-learning-simplify-sentences-wikipedia/) - [Simple English Wikipedia: A New Text Simplification Task](https://md.research.elamur.ai/free/en/papers/coster-2011-simple-english-wikipedia-new-text/) - [Simplifying Lexical Simplification: Do We Need Simplified Corpora?](https://md.research.elamur.ai/free/en/papers/glavas-2015-simplifying-lexical-simplification-do-we/) - [Learning a Lexical Simplifier Using Wikipedia](https://md.research.elamur.ai/free/en/papers/horn-2014-learning-lexical-simplifier-wikipedia/) - [Improving Text Simplification Language Modeling Using Unsimplified Text Data](https://md.research.elamur.ai/free/en/papers/kauchak-2013-improving-text-simplification-language-modeling/) - [Data-Driven Sentence Simplification: Survey and Benchmark](https://md.research.elamur.ai/free/en/papers/manchego-2020-data-driven-sentence-simplification-survey/) - [Linguistic Models for Analyzing and Detecting Biased Language](https://md.research.elamur.ai/free/en/papers/recasens-2013-linguistic-models-analyzing-detecting-biased/) - [SemEval-2012 Task 1: English Lexical Simplification](https://md.research.elamur.ai/free/en/papers/specia-2012-semeval-2012-task-1-english/) - [Learning to Simplify Sentences with Quasi-Synchronous Grammar and Integer Programming](https://md.research.elamur.ai/free/en/papers/woodsend-2011-learning-simplify-sentences-quasi-synchronous/) - [Sentence Simplification by Monolingual Machine Translation](https://md.research.elamur.ai/free/en/papers/wubben-2012-sentence-simplification-monolingual-machine-translation/) - [Paraphrasing for Style](https://md.research.elamur.ai/free/en/papers/xu-2012-paraphrasing-style/) ## Cite The paper itself, not this page. BibTeX from the library's record of the work. ```bibtex @inproceedings{yatskar2010sake, title = {{For the sake of simplicity: Unsupervised extraction of lexical simplifications from Wikipedia}}, author = {Yatskar, Mark and Pang, Bo and Danescu-Niculescu-Mizil, Cristian and Lee, Lillian}, booktitle = {Human Language Technologies: The 2010 Annual Conference of the North American Chapter of the Association for Computational Linguistics}, year = {2010}, url = {https://aclanthology.org/N10-1056/} } ``` - [← previous Fine-Grain Morphological Analyzer and Part-of-Speech Tagger for Arabic Text](https://md.research.elamur.ai/free/en/papers/lingbulk-2010-sawalha-fine-grain-morphological-analyzer-part/) - [next → FrameNet Translation Using Bilingual Dictionaries with Evaluation on the English-French Pair](https://md.research.elamur.ai/free/en/papers/mouton-2010-framenet-translation-bilingual-dictionaries-evaluation/) - [↑ back to papers of 2010](https://md.research.elamur.ai/free/en/papers/2010/) ## References 10 references, 0 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet. 1. R. Chandrasekar, B. Srinivas. Automatic induction of rules for text simplification. Knowledge-Based Systems, 1997. [find](https://scholar.google.com/scholar?q=R%2E+Chandrasekar%2C+B%2E+Srinivas%2E+Automatic+induction+of+rules+for+text+simplification%2E+Knowledge%2DBased+Systems%2C+1997%2E) 2. L. Deléger, P. Zweigenbaum. Extracting lay paraphrases of specialized expressions from monolingual comparable medical corpora. Workshop on Building and Using Comparable Corpora, 2009. [find](https://scholar.google.com/scholar?q=L%2E+Dele%CC%81ger%2C+P%2E+Zweigenbaum%2E+Extracting+lay+paraphrases+of+specialized+expressions+from+monolingual+comparable+medical+corpora%2E+Workshop+on+Building+and+Using+Comparable+Corpora%2C+2009%2E) 3. S. Devlin, J. Tait. The use of a psycholinguistic database in the simplification of text for aphasic readers. In Linguistic Databases, 1998. [find](https://scholar.google.com/scholar?q=S%2E+Devlin%2C+J%2E+Tait%2E+The+use+of+a+psycholinguistic+database+in+the+simplification+of+text+for+aphasic+readers%2E+In+Linguistic+Databases%2C+1998%2E) 4. N. Elhadad, K. Sutaria. Mining a lexicon of technical terms and lay equivalents. Workshop on BioNLP, 2007. [find](https://scholar.google.com/scholar?q=N%2E+Elhadad%2C+K%2E+Sutaria%2E+Mining+a+lexicon+of+technical+terms+and+lay+equivalents%2E+Workshop+on+BioNLP%2C+2007%2E) 5. B. Beigman Klebanov, K. Knight, D. Marcu. Text simplification for information-seeking applications. OTM Conferences, 2004. [find](https://scholar.google.com/scholar?q=B%2E+Beigman+Klebanov%2C+K%2E+Knight%2C+D%2E+Marcu%2E+Text+simplification+for+information%2Dseeking+applications%2E+OTM+Conferences%2C+2004%2E) 6. R. Nelken, S. M. Shieber. Towards robust context-sensitive sentence alignment for monolingual corpora. EACL, 2006. [find](https://scholar.google.com/scholar?q=R%2E+Nelken%2C+S%2E+M%2E+Shieber%2E+Towards+robust+context%2Dsensitive+sentence+alignment+for+monolingual+corpora%2E+EACL%2C+2006%2E) 7. R. Nelken, E. Yamangil. Mining Wikipedia’s article revision history for training computational linguistics algorithms. WikiAI, 2008. [find](https://scholar.google.com/scholar?q=R%2E+Nelken%2C+E%2E+Yamangil%2E+Mining+Wikipedia%E2%80%99s+article+revision+history+for+training+computational+linguistics+algorithms%2E+WikiAI%2C+2008%2E) 8. E. Shnarch, L. Barak, I. Dagan. Extracting lexical reference rules from Wikipedia. ACL, 2009. [find](https://scholar.google.com/scholar?q=E%2E+Shnarch%2C+L%2E+Barak%2C+I%2E+Dagan%2E+Extracting+lexical+reference+rules+from+Wikipedia%2E+ACL%2C+2009%2E) 9. A. Siddharthan, A. Nenkova, K. McKeown. Syntactic simplification for improving content selection in multi-document summarization. COLING, 2004. [find](https://scholar.google.com/scholar?q=A%2E+Siddharthan%2C+A%2E+Nenkova%2C+K%2E+McKeown%2E+Syntactic+simplification+for+improving+content+selection+in+multi%2Ddocument+summarization%2E+COLING%2C+2004%2E) 10. D. Vickrey, D. Koller. Sentence simplification for semantic role labeling/ ACL, 2008. [find](https://scholar.google.com/scholar?q=D%2E+Vickrey%2C+D%2E+Koller%2E+Sentence+simplification+for+semantic+role+labeling%2F+ACL%2C+2008%2E)