---
title: "NPtool, a Detector of English Noun Phrases — Voutilainen 1993"
description: "It builds on the ENGCG pipeline, which works by reduction. 100% agreement. 15 references, cited by 11 papers here."
authors: ["Atro Voutilainen"]
year: 1993
venue: "Very Large Corpora: Academic and Industrial Perspectives"
source: "https://aclanthology.org/W93-0306/"
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/voutilainen-1993-nptool-detector-english-noun-phrases/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/voutilainen-1993-nptool-detector-english-noun-phrases/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [1993](https://md.research.elamur.ai/free/en/papers/1993/) / NPtool, a Detector of English Noun Phrases

# NPtool, a Detector of English Noun Phrases

**Authors:** [Atro Voutilainen](https://md.research.elamur.ai/free/en/authors/atro-voutilainen/) · **Year:** 1993 · **Venue:** Very Large Corpora: Academic and Industrial Perspectives · **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/W93-0306/) · [PDF](https://aclanthology.org/W93-0306.pdf) · [OpenAlex](https://openalex.org/W2134351768)

**Citations:** 70 citations (OpenAlex, 2026-10-14) · [11](https://md.research.elamur.ai/free/en/papers/voutilainen-1993-nptool-detector-english-noun-phrases/cited-by/) in our library, of the 149,439 papers whose bibliographies we read

## Summary
NPtool is a system that finds noun phrases in unrestricted English text. It is meant for index-term generation in information retrieval, for discovering translation units, and for corpus linguistics. It builds on the ENGCG pipeline, which works by reduction. First, every word receives all of its possible morphological readings from a detailed hand-built lexicon. Then hand-written, order-independent constraints remove readings that the context rules out. A new Constraint Syntax and a finite-state parser add shallow syntactic function tags such as @>N (premodifier) and @NH (nominal head). These tags mark head–modifier relations without building full trees. An extraction utility then matches noun-phrase patterns and labels each candidate as sure or uncertain. On about 20,000 manually checked words from several domains, recall for complex NPs is 98.5–100% and precision is roughly 95–98%. The work shows that careful linguistic rules with a deliberately shallow tagset can deliver very reliable chunking. It also shows that leftover uncertainty can be flagged for a human reviewer rather than hidden.

## Key points
- The syntax was tested on newspapers, electrical-engineering abstracts and car manuals, with 20,000–30,000 words proofread. Afterwards 93–96% of words carried a single syntactic analysis, and the error rate stayed under 1%.
- A version of the finite-state parser without NP preference was run on 25,500 words (1,495 sentences) from The Grolier Electronic Encyclopaedia:
  - 64% of sentences (960) came out unambiguous;
  - only about 2% kept more than ten readings;
  - the worst case had 72 readings.
- For comparison, after plain ENGCG parsing 23.5% of sentences were more ambiguous than NPtool's worst case.
- The ENGCG morphological disambiguator relies only on context rules:
  - it keeps the correct reading in over 99.5% of cases;
  - 2–3% of words remain ambiguous;
  - it resolves more than 97% uniquely, against roughly 90% for lexical probabilities alone (Church).
- About 95% of wrongly proposed NP candidates receive the uncertain "?:" mark rather than "ok:". Double-blind manual morphological annotation typically reaches 100% agreement between judges.

## Interesting ideas
- Parsing as elimination: the lexicon proposes every reading, constraints only delete readings, and no new structure is ever built, so the output is always a subset of what the lexicon offered.
- Rules are independent and can be applied in any order. Broad feature-level constraints sit alongside narrow word-form-level ones that refine them without conflict, which makes rules easy to add and test one at a time.
- The tagset includes only distinctions that syntax alone can resolve reliably, such as modifier direction and head status. Hard cases like adverbial attachment and modifier scope are left out, so the parser is not forced to guess.
- Each output is labelled as a sure or uncertain hit, and the uncertain label catches about 95% of the false candidates, which makes human or downstream review cheap.
- No input is rejected outright: rules are ranked by how far they may be relaxed, so ill-formed text still gets its best available analysis.

## For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

- **I1** (tokens): The tagger could keep the full reading lattice from crate en and record which rule removed each reading, so every lemma/POS decision can be explained.
- **I2** (learn): Self-induced repair rules could follow this design, each one an isolated, order-free deletion that is checked separately against negative controls.
- **I3** (tree): When converting between UD dialects or merging witness parsers, unresolvable attachments could be marked as underspecified rather than guessed, and passed downstream for meaning-based resolution.
- **I4** (answer): Readers could attach a sure/uncertain flag to each extracted entity or answer, and spend LLM-teacher checks only on the uncertain ones.
- **I5** (frames): Frame extraction in prag could fall back to ranked, relaxable constraints on noisy sentences instead of producing nothing.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

## Institutions

- [University of Helsinki](https://md.research.elamur.ai/free/en/universities/university-of-helsinki/) — University, Helsinki, Finland · 50 works · 5,318 citations

📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 16,039 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 15,891 works

🧭 Topics (1): [English](https://md.research.elamur.ai/free/en/topics/english/) · 2,263 works

## Cite

Cite as: Atro Voutilainen. 1993. NPtool, a Detector of English Noun Phrases. Very Large Corpora: Academic and Industrial Perspectives.

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@inproceedings{voutilainen1993nptool,
  title = {{NPtool, a Detector of English Noun Phrases}},
  author = {Atro Voutilainen},
  booktitle = {Very Large Corpora: Academic and Industrial Perspectives},
  year = {1993},
  url = {https://aclanthology.org/W93-0306/}
}
```

- [← previous Text-Translation Alignment](https://md.research.elamur.ai/free/en/papers/kay-1993-text-translation-alignment/)
- [next → Book Reviews: Corpus Linguistics and the Automatic Analysis of English](https://md.research.elamur.ai/free/en/papers/briscoe-1993-book-reviews-corpus-linguistics-automatic/)
- [↑ back to papers of 1993](https://md.research.elamur.ai/free/en/papers/1993/)

## References

15 references, 3 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Bourigault, D. 1992. Surface grammatical analysis for the extraction of terminological noun phrases. In Proceedings of the fifteenth International Conference on Computational Linguistics. COLING-g2, Vol. IIL Nantes, France. 977-981. [find](https://scholar.google.com/scholar?q=Bourigault%2C+D%2E+1992%2E+Surface+grammatical+analysis+for+the+extraction+of+terminological+noun+phrases%2E+In+Proceedings+of+the+fifteenth+International+Conference+on+Computational+Linguistics%2E+COLING%2Dg2%2C+V)
2. Church, K. 1988. [A Stochastic Parts Program and Noun Phrase Parser for Unrestricted Text](https://md.research.elamur.ai/free/en/papers/church-1988-stochastic-parts-program-noun-phrase/). Proceedings of the Second Conference on Applied Natural Language Processing, A CL. 136-143.
3. Church, K. 1992. Current Practice in Part of Speech Tagging and Suggestions for the Future, in Simmons (ed.), Sboruik praci: In Honor of Henry Ku~era. Michigan Slavic Studies. [find](https://scholar.google.com/scholar?q=Church%2C+K%2E+1992%2E+Current+Practice+in+Part+of+Speech+Tagging+and+Suggestions+for+the+Future%2C+in+Simmons+%28ed%2E%29%2C+Sboruik+praci%3A+In+Honor+of+Henry+Ku%7Eera%2E+Michigan+Slavic+Studies%2E)
4. van der Eijk, P. 1993. Automat,-ing the acquisition of bilingual terminology. Proceedings of EACL'93. Utrecht, The Netherlands. [find](https://scholar.google.com/scholar?q=van+der+Eijk%2C+P%2E+1993%2E+Automat%2C%2Ding+the+acquisition+of+bilingual+terminology%2E+Proceedings+of+EACL%2793%2E+Utrecht%2C+The+Netherlands%2E)
5. Karlsson, F. 1990. [Constraint Grammar as a Framework for Parsing Running Text](https://md.research.elamur.ai/free/en/papers/karlsson-1990-constraint-grammar-as-framework-parsing/). In H. Karlgren (ed.), Papers presented to the 13th International Conference on Computational Linguistics, Vol. 3. Helsinki. 168-173. [↗](https://doi.org/10.3115/991146.991176)
6. Koskenniemi, K. (1990). Finite-state parsing and disambiguation. In Karlgren, H. (ed.) COLING-90. Papers presented to the 13th International Conference on Computational Linguistics, Vol. 2. Helsinki, Finland. 229-232. [Koskenniemi, Tapanainen and Voutilainen, 1992] Koskenniemi, K., Tapanainen, P. and Voutilainen, A. (1992). Compiling and using finite-state syntactic rules. In Proceedings of the fifteenth later-national Conference on Computational Linguistics. COLING-9~, Vol. L Nantes, France. 156-162. [Quirk, Greenbaum, Leech and Svartvik, 1985] Quirk, R., Greenbanm, S., Leech, G. and Svartvik, J. 1985. A Comprehensive Grammar of the English Language. London & New York: Longman. [find](https://scholar.google.com/scholar?q=Koskenniemi%2C+K%2E+%281990%29%2E+Finite%2Dstate+parsing+and+disambiguation%2E+In+Karlgren%2C+H%2E+%28ed%2E%29+COLING%2D90%2E+Papers+presented+to+the+13th+International+Conference+on+Computational+Linguistics%2C+Vol%2E+2%2E+Helsinki%2C+)
7. Bausch, B., Norrback, R., and Svensson, T. 1992. Excerper-ing av nominMfraser ur 15pande text. Manuscript. Stockholms universitet, Institutionen F6r Lingvistik. [find](https://scholar.google.com/scholar?q=Bausch%2C+B%2E%2C+Norrback%2C+R%2E%2C+and+Svensson%2C+T%2E+1992%2E+Excerper%2Ding+av+nominMfraser+ur+15pande+text%2E+Manuscript%2E+Stockholms+universitet%2C+Institutionen+F6r+Lingvistik%2E)
8. SMton, G. and McGill, M. 1983. Introduction to Modern Information Retrieval. McGraw-Hill, Auckland. [find](https://scholar.google.com/scholar?q=SMton%2C+G%2E+and+McGill%2C+M%2E+1983%2E+Introduction+to+Modern+Information+Retrieval%2E+McGraw%2DHill%2C+Auckland%2E)
9. Sampson, G. 1987. Probabilistic Models of Analysis. In Garside, Leech and Sampson (eds.) 1987. 16-29. [find](https://scholar.google.com/scholar?q=Sampson%2C+G%2E+1987%2E+Probabilistic+Models+of+Analysis%2E+In+Garside%2C+Leech+and+Sampson+%28eds%2E%29+1987%2E+16%2D29%2E)
10. Sampson, G. 1987. The grammatical database and parsing scheme. In Garside, Leech and Sampson (eds.) 1987.82-96. [find](https://scholar.google.com/scholar?q=Sampson%2C+G%2E+1987%2E+The+grammatical+database+and+parsing+scheme%2E+In+Garside%2C+Leech+and+Sampson+%28eds%2E%29+1987%2E82%2D96%2E)
11. Tapanainen, P. 1991..~.grellisin~. automaatteina esitettyjen kielioppis~£~.ntSjen so-veltaminen luonnollisen kielen j~ent~.j~.s.sg (Natural language parsing with finite-state syntactic rules). Master's thesis. Dept. of computer science, University of Helsinki. [find](https://scholar.google.com/scholar?q=Tapanainen%2C+P%2E+1991%2E%2E%7E%2Egrellisin%7E%2E+automaatteina+esitettyjen+kielioppis%7E%C2%A3%7E%2EntSjen+so%2Dveltaminen+luonnollisen+kielen+j%7Eent%7E%2Ej%7E%2Es%2Esg+%28Natural+language+parsing+with+finite%2Dstate+syntactic+rules%29%2E+Master%27)
12. Taylor, L., Graver, C. and Briscoe, T. 1989. [The Syntactic Regularity of English Noun Phrases](https://md.research.elamur.ai/free/en/papers/taylor-1989-syntactic-regularity-english-noun-phrases/). In Proceedings of the Fourth Conference of the European Chapter of the ACL. 256-263.
13. Voutilainen, A. (forthcoming 1993). Designing a parsing grammar. [find](https://scholar.google.com/scholar?q=Voutilainen%2C+A%2E+%28forthcoming+1993%29%2E+Designing+a+parsing+grammar%2E)
14. Voutilainen, A., Heikkil~, J. and Anttila, A. (1992). Constraint Grammar of English: A Performance.Oriented Introduction. Publication No. 21, Department of General Linguistics, University of Helsinki. [find](https://scholar.google.com/scholar?q=Voutilainen%2C+A%2E%2C+Heikkil%7E%2C+J%2E+and+Anttila%2C+A%2E+%281992%29%2E+Constraint+Grammar+of+English%3A+A+Performance%2EOriented+Introduction%2E+Publication+No%2E+21%2C+Department+of+General+Linguistics%2C+University+of+Helsinki%2E)
15. Voutilninen, A. and Tapanainen, P. 1993. Ambiguity resolution in a reductionistic parser. Proceedings of EACL'93. Utrecht, Holland. APPENDIX Here is given the NPtool analysis of a small sample from the CACM text collection. - Here is the input text: The b i n a r y number system o f f e r s many advantages o v e r a decimal r e p r e s e n t a t i o n for a high-performance, general-purpose computer. The greater simplicity of a binary arithmetic unit and the greater [find](https://scholar.google.com/scholar?q=Voutilninen%2C+A%2E+and+Tapanainen%2C+P%2E+1993%2E+Ambiguity+resolution+in+a+reductionistic+parser%2E+Proceedings+of+EACL%2793%2E+Utrecht%2C+Holland%2E+APPENDIX+Here+is+given+the+NPtool+analysis+of+a+small+sample+from+th)

*Updated 1 November 2026*