--- title: "A Corpus-Based Approach for Building Semantic Lexicons — Riloff 1997" description: "Riloff and Shepherd address the cost of building domain-specific semantic lexicons. 14 references, cited by 69 papers here." authors: ["Ellen Riloff", "Jessica Shepherd"] year: 1997 venue: "Second Conference on Empirical Methods in Natural Language Processing" source: "https://aclanthology.org/W97-0313/" 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/riloff-1997-corpus-based-approach-building-semantic/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/riloff-1997-corpus-based-approach-building-semantic/) [home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [1997](https://md.research.elamur.ai/free/en/papers/1997/) / A Corpus-Based Approach for Building Semantic Lexicons # A Corpus-Based Approach for Building Semantic Lexicons **Authors:** [Ellen Riloff](https://md.research.elamur.ai/free/en/authors/ellen-riloff/) and Jessica Shepherd · **Year:** 1997 · **Venue:** Second Conference on Empirical Methods in Natural Language Processing · **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/W97-0313/) · [PDF](https://aclanthology.org/W97-0313.pdf) · [OpenAlex](https://openalex.org/W1930023685) **Citations:** 160 citations (OpenAlex, 2026-10-14) · [69](https://md.research.elamur.ai/free/en/papers/riloff-1997-corpus-based-approach-building-semantic/cited-by/) in our library, of the 140,884 papers whose bibliographies we read ## Summary Riloff and Shepherd address the cost of building domain-specific semantic lexicons. General resources like WordNet and Cyc cannot cover every sublanguage, and hand-coding lexicons for each application is slow. Their semi-automatic method needs only a representative corpus and about five seed words per category. It rests on the observation that members of a category tend to appear near each other in conjunctions, lists, appositives and nominal compounds. For every occurrence of a seed word used as a head noun, the system collects the nearest noun on each side. It then scores each corpus noun by how often it appears in these contexts relative to its overall frequency. In a bootstrapping loop, the top-scoring new nouns join the seed list. After several rounds, a person reviews the final ranked list. The method was tested on the MUC-4 corpus with five categories. Judges typically found about 60 category words per category in 10–15 minutes. This shows that useful core lexicons can be built from raw text without any prior semantic resources. ## Key points - **Data:** the MUC-4 development corpus (1700 texts, mostly terrorism and military reports). - **Categories:** energy, financial, military, vehicle, weapon, each starting from 5 seed words (e.g., bomb, dynamite, explosives, gun, rifle for Weapon). - **Scoring:** a word's score is its frequency inside category context windows divided by its total corpus frequency. Because windows can overlap, the score may exceed 1. - **Filtering:** stopwords (about 30 general nouns), numbers and words with frequency below 5 are removed. - **Bootstrapping:** each iteration adds the top 5 new nouns to the seed list without human checks. About eight iterations worked best. - **Evaluation:** two independent judges per category rated the top 200 words in random order on a graded association scale. - **Other results:** - Energy did relatively poorly, probably because it is sparse in the corpus. - Location looked good but was not formally evaluated. - Commercial yielded few new words, and Person gave mixed results. - Using 10 seeds instead of 5 changed little. ## Interesting ideas - **Tiny context windows:** using only the nearest noun on each side worked better than wider windows. It captures coordination and apposition patterns cheaply, so it is a quick way to bootstrap category lists in low-resource languages. - **Seeds as head nouns only:** restricting seeds to head-noun positions keeps them as the main concept of the phrase and reduces wrong-sense matches. This simple syntactic filter could cut noise in many seed-expansion methods. - **Machine ranks, human decides:** the system outputs a ranked list, and a person chooses the true members. This avoids defining category boundaries algorithmically and makes the tool's value measurable as useful words found per minute of review. - **Blind, shuffled judging:** showing judges candidates in random order prevents runs of good words from biasing later ratings. A graded scale lets different applications pick their own membership threshold. Both tricks suit any annotation of ranked outputs. - **Watching for semantic drift:** the authors note that unchecked bootstrapping can derail when a category is poorly represented in the corpus. Checking category coverage before running, or limiting the number of iterations, is a lesson for any self-training loop. *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/) ## Institutions - [University of Utah](https://md.research.elamur.ai/free/en/universities/university-of-utah/) — University, Salt Lake City, United States · 22 works · 4,778 citations ## Cite Cite as: Ellen Riloff and Jessica Shepherd. 1997. A Corpus-Based Approach for Building Semantic Lexicons. Second Conference on Empirical Methods in Natural Language Processing. The paper itself, not this page. BibTeX from the library's record of the work. ```bibtex @inproceedings{riloff1997corpus, title = {{A Corpus-Based Approach for Building Semantic Lexicons}}, author = {Ellen Riloff and Jessica Shepherd}, booktitle = {Second Conference on Empirical Methods in Natural Language Processing}, year = {1997}, url = {https://aclanthology.org/W97-0313/} } ``` - [← previous A Fast and Portable Realizer for Text Generation Systems](https://md.research.elamur.ai/free/en/papers/lavoie-1997-fast-portable-realizer-text-generation/) - [next → IGTree: Using Trees for Compression and Classification in Lazy Learning Algorithms](https://md.research.elamur.ai/free/en/papers/daelemans-1997-igtree-trees-compression-classification-lazy/) - [↑ back to papers of 1997](https://md.research.elamur.ai/free/en/papers/1997/) ## References 14 references, 2 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet. 1. Berwick, Robert C. 1989. Learning Word Meanings from Examples. In Semantic Structures: Advances in Natural Language Processing. Lawrence Erlbaum Associates, chapter 3, pages 89-124. [find](https://scholar.google.com/scholar?q=Berwick%2C+Robert+C%2E+1989%2E+Learning+Word+Meanings+from+Examples%2E+In+Semantic+Structures%3A+Advances+in+Natural+Language+Processing%2E+Lawrence+Erlbaum+Associates%2C+chapter+3%2C+pages+89%2D124%2E) 2. Brill, E. 1994. Some Advances in Rule-based Part of Speech Tagging. In Proceedings of the Twelfth National Conference on Artificial Intelligence, pages 722-727. AAAI Press/The MIT Press. [find](https://scholar.google.com/scholar?q=Brill%2C+E%2E+1994%2E+Some+Advances+in+Rule%2Dbased+Part+of+Speech+Tagging%2E+In+Proceedings+of+the+Twelfth+National+Conference+on+Artificial+Intelligence%2C+pages+722%2D727%2E+AAAI+Press%2FThe+MIT+Press%2E) 3. Carbonell, J. G. 1979. Towards a Self-Extending Parser. In Proceedings of the 17th Annual Meeting of the Association for Computational Linguistics, pages 3-7. [find](https://scholar.google.com/scholar?q=Carbonell%2C+J%2E+G%2E+1979%2E+Towards+a+Self%2DExtending+Parser%2E+In+Proceedings+of+the+17th+Annual+Meeting+of+the+Association+for+Computational+Linguistics%2C+pages+3%2D7%2E) 4. Cardie, C. 1993. A Case-Based Approach to Knowledge Acquisition for Domain-Specific Sentence Analysis. In Proceedings of the Eleventh National Conference on Artificial Intelligence, pages 798-803. AAAI Press/The MIT Press. [find](https://scholar.google.com/scholar?q=Cardie%2C+C%2E+1993%2E+A+Case%2DBased+Approach+to+Knowledge+Acquisition+for+Domain%2DSpecific+Sentence+Analysis%2E+In+Proceedings+of+the+Eleventh+National+Conference+on+Artificial+Intelligence%2C+pages+798%2D803%2E+AAA) 5. Church, K. 1989. A Stochastic Parts Program and Noun Phrase Parser for Unrestricted Text. In Proceedings off the Second Conference on Applied Natural Language Processing. [our card](https://md.research.elamur.ai/free/en/papers/church-1988-stochastic-parts-program-noun-phrase/) 6. Granger, R. H. 1977. FOUL-UP: A Program that Figures Out Meanings of Words from Context. In Proceedings of the Fifth International Joint Conference on Artificial Intelligence, pages 172-178. [find](https://scholar.google.com/scholar?q=Granger%2C+R%2E+H%2E+1977%2E+FOUL%2DUP%3A+A+Program+that+Figures+Out+Meanings+of+Words+from+Context%2E+In+Proceedings+of+the+Fifth+International+Joint+Conference+on+Artificial+Intelligence%2C+pages+172%2D178%2E) 7. Hastings, P. and S. Lytinen. 1994. The Ups and Downs of Lexical Acquisition. In Proceedings of the Twelfth National Conference on Artificial Intelligence, pages 754-759. AAAI Press/The MIT Press. [find](https://scholar.google.com/scholar?q=Hastings%2C+P%2E+and+S%2E+Lytinen%2E+1994%2E+The+Ups+and+Downs+of+Lexical+Acquisition%2E+In+Proceedings+of+the+Twelfth+National+Conference+on+Artificial+Intelligence%2C+pages+754%2D759%2E+AAAI+Press%2FThe+MIT+Press%2E) 8. Jacobs, P. and U. Zernik. 1988. Acquiring Lexical Knowledge from Text: A Case Study. In Proceedings of the Seventh National Conference on Artificial Intelligence, pages 739-744. [find](https://scholar.google.com/scholar?q=Jacobs%2C+P%2E+and+U%2E+Zernik%2E+1988%2E+Acquiring+Lexical+Knowledge+from+Text%3A+A+Case+Study%2E+In+Proceedings+of+the+Seventh+National+Conference+on+Artificial+Intelligence%2C+pages+739%2D744%2E) 9. Lehnert, W., C. Cardie, D. Fisher, J. McCarthy, E. Riloff, and S. Soderland. 1992. University of Massachusetts: Description of the CIR-CUS System as Used for MUC-4. In Proceedings of the Fourth Message Understanding Conference (MUC-~), pages 282-288, San Mateo, CA. Morgan Kaufmann. [find](https://scholar.google.com/scholar?q=Lehnert%2C+W%2E%2C+C%2E+Cardie%2C+D%2E+Fisher%2C+J%2E+McCarthy%2C+E%2E+Riloff%2C+and+S%2E+Soderland%2E+1992%2E+University+of+Massachusetts%3A+Description+of+the+CIR%2DCUS+System+as+Used+for+MUC%2D4%2E+In+Proceedings+of+the+Fourth+Messag) 10. Lenat, D. B., M. Prakash, and M. Shepherd. 1986. CYC: Using Common Sense Knowledge to Overcome Brittleness and Knowledge-Acquisition Bottlenecks. AI Magazine, 6:65-85. [find](https://scholar.google.com/scholar?q=Lenat%2C+D%2E+B%2E%2C+M%2E+Prakash%2C+and+M%2E+Shepherd%2E+1986%2E+CYC%3A+Using+Common+Sense+Knowledge+to+Overcome+Brittleness+and+Knowledge%2DAcquisition+Bottlenecks%2E+AI+Magazine%2C+6%3A65%2D85%2E) 11. Miller, G. 1990. Wordnet: An On-line Lexical Database. International Journal of Lexicography, 3(4). [find](https://scholar.google.com/scholar?q=Miller%2C+G%2E+1990%2E+Wordnet%3A+An+On%2Dline+Lexical+Database%2E+International+Journal+of+Lexicography%2C+3%284%29%2E) 12. MUC-4 Proceedings. 1992. Proceedings of the Fourth Message Understanding Conference (MUC-4). Morgan Kaufmann, San Mateo, CA. [find](https://scholar.google.com/scholar?q=MUC%2D4+Proceedings%2E+1992%2E+Proceedings+of+the+Fourth+Message+Understanding+Conference+%28MUC%2D4%29%2E+Morgan+Kaufmann%2C+San+Mateo%2C+CA%2E) 13. Weischedel, R., M. Meteer, R. Schwartz, L. Ramshaw, and J. Palmucci. 1993. Coping with Ambiguity and Unknown Words through Probabilistic Models. Computational Linguistics, 19(2):359-382. [find](https://scholar.google.com/scholar?q=Weischedel%2C+R%2E%2C+M%2E+Meteer%2C+R%2E+Schwartz%2C+L%2E+Ramshaw%2C+and+J%2E+Palmucci%2E+1993%2E+Coping+with+Ambiguity+and+Unknown+Words+through+Probabilistic+Models%2E+Computational+Linguistics%2C+19%282%29%3A359%2D382%2E) 14. Yarowsky, D. 1992. Word sense disambiguation using statistical models of Roget's categories trained on large corpora. In Proceedings of the Fourteenth International Conference on Computational Linguistics (COLING-92), pages 454-460. [our card](https://md.research.elamur.ai/free/en/papers/yarowsky-1992-word-sense-disambiguation-statistical-models/)