---
title: "Learning to Parse Database Queries Using Inductive Logic Programming — Zelle 1996"
description: "The authors address the problem of automatically constructing a natural language interface to a database. 16 references, cited by 358 papers here."
authors: ["John M. Zelle", "Raymond J. Mooney"]
year: 1996
venue: "AAAI 1996"
source: "https://www.cs.utexas.edu/~ml/publications/"
licence: "author's copy — read only"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/zelle-1996-learning-parse-database-queries-using/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/zelle-1996-learning-parse-database-queries-using/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [1996](https://md.research.elamur.ai/free/en/papers/1996/) / Learning to Parse Database Queries Using Inductive Logic Programming

# Learning to Parse Database Queries Using Inductive Logic Programming

**Authors:** John M. Zelle and [Raymond J. Mooney](https://md.research.elamur.ai/free/en/authors/raymond-mooney/) · **Year:** 1996 · **Venue:** AAAI 1996 · **Licence:** author's copy — read only

**Links:** [Publisher's page](https://www.cs.utexas.edu/~ml/publications/) · [PDF](https://www.cs.utexas.edu/~ml/papers/chill-aaai-96.pdf)

**Citations:** [358](https://md.research.elamur.ai/free/en/papers/zelle-1996-learning-parse-database-queries-using/cited-by/) in our library, of the 140,884 papers whose bibliographies we read

## Summary
The authors address the problem of automatically constructing a natural language interface to a database. An English question must be mapped directly onto an executable query. To this end, they use Chill, a system that frames acquiring a parser as searching for rules that govern a shift-reduce parser implemented in Prolog. First, an overly general parser is built from the training examples; it produces many spurious parses. Next, the contexts in which each operation is appropriate or inappropriate are collected from the examples. Inductive Logic Programming (ILP) methods learn relational rules from these contexts, and the rules are then incorporated back into the parser. The domain is U.S. geography. For it, the authors developed the logical query language Geoquery and a corpus of sentences, each paired with a query. The existing hand-built system Geobase served as the baseline for comparison. The learned parser answered more accurately than Geobase when given at least 175 training examples. The significance of the work is that a corpus-based approach learned a semantic mapping rather than syntax alone and improved a complete application. Moreover, annotating a corpus with ready-made queries is far easier than producing detailed linguistic annotation.

## Key points
- The Geobase database originates from an example supplied with Turbo Prolog 2.0 and contains approximately 800 Prolog facts about U.S. states: population, area, capitals, neighboring states, major cities, rivers, and each state's highest and lowest elevations.
- Geoquery is a first-order logical form with meta-predicates (`answer`, `largest`, `count`, `most`, etc.). The initial version of the language was designed on the basis of a sample of 50 sentences. A logical form was chosen over SQL because it allows a simpler compositional mapping from natural language.
- Evaluation was based on 10 runs with random data splits, each test set containing 25 sentences. Chill outperforms Geobase starting from 175 training examples. In the best run, a parser of approximately 1100 lines of Prolog correctly answered 84% of novel questions.
- Spurious parses, i.e., queries that yield an incorrect answer, are rare: 3.2% with 175 examples and 2.3% with 200. For Geobase, this figure is 3.8%.
- Unlike systems for the ATIS benchmark, where only part of a hand-written grammar was learned, Chill maps a string of words directly onto a query, without intermediate syntactic analysis. Only the lexicon is still constructed separately by hand.

*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: [AI & machine learning](https://md.research.elamur.ai/free/en/shelves/ai/)

## Cite

Cite as: John M. Zelle and Raymond J. Mooney. 1996. Learning to Parse Database Queries Using Inductive Logic Programming. AAAI 1996.

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

```bibtex
@inproceedings{zelle1996learning,
  title = {{Learning to Parse Database Queries Using Inductive Logic Programming}},
  author = {John M. Zelle and Raymond J. Mooney},
  booktitle = {AAAI 1996},
  year = {1996},
  url = {https://www.cs.utexas.edu/~ml/publications/}
}
```

- [← previous A Stochastic Finite-State Word-Segmentation Algorithm for Chinese](https://md.research.elamur.ai/free/en/papers/sproat-1996-stochastic-finite-state-word-segmentation/)
- [next → MBT: A Memory-Based Part of Speech Tagger-Generator](https://md.research.elamur.ai/free/en/papers/daelemans-1996-mbt-memory-based-part-speech/)
- [↑ back to papers of 1996](https://md.research.elamur.ai/free/en/papers/1996/)

## References

16 references, 1 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Abramson, H., and Dahl, V. 1989. Logic Grammars . New York: Springer-Verlag. Black, E.; Jelineck, F.; La erty, J.; Magerman, D.; Mercer, R.; and Roukos, S. 1993. Towards history-based grammars: Using richer models for probabilistic parsing. In Proceedings of the 31st Annual Meeting of the Associatio… [find](https://scholar.google.com/scholar?q=Abramson%2C+H%2E%2C+and+Dahl%2C+V%2E+1989%2E+Logic+Grammars+%2E+New+York%3A+Springer%2DVerlag%2E+Black%2C+E%2E%3B+Jelineck%2C+F%2E%3B+La+erty%2C+J%2E%3B+Magerman%2C+D%2E%3B+Mercer%2C+R%2E%3B+and+Roukos%2C+S%2E+1993%2E+Towards+history%2Dbased+grammars%3A+Using+)
2. Borland International. 1988. Turbo Prolog 2.0 Reference Guide . Scotts Valley, CA: Borland International. [find](https://scholar.google.com/scholar?q=Borland+International%2E+1988%2E+Turbo+Prolog+2%2E0+Reference+Guide+%2E+Scotts+Valley%2C+CA%3A+Borland+International%2E)
3. Brill, E. 1993. Automatic grammar induction and parsing free text: A transformation-based approach. In Proceedings of the 31st Annual Meeting of the Association for Computational Linguistics , 259{265. [find](https://scholar.google.com/scholar?q=Brill%2C+E%2E+1993%2E+Automatic+grammar+induction+and+parsing+free+text%3A+A+transformation%2Dbased+approach%2E+In+Proceedings+of+the+31st+Annual+Meeting+of+the+Association+for+Computational+Linguistics+%2C+259%7B265)
4. Charniak, E., and Carroll, G. 1994. Context-sensitive statistics for improved grammatical language models. In Proceedings of the Twelfth National Conference on Arti - cial Intelligence . [find](https://scholar.google.com/scholar?q=Charniak%2C+E%2E%2C+and+Carroll%2C+G%2E+1994%2E+Context%2Dsensitive+statistics+for+improved+grammatical+language+models%2E+In+Proceedings+of+the+Twelfth+National+Conference+on+Arti+%2D+cial+Intelligence+%2E)
5. Kuhn, R., and De Mori, R. 1995. The application of semantic classi cation trees to natural language understanding. IEEE Transactions on Pattern Analysis and Machine Intelligence 17(5):449{460. Lavrac, N., and Dzeroski, S., eds. 1994. Inductive Logic Programming: Techniques and Applications . Ellis… [find](https://scholar.google.com/scholar?q=Kuhn%2C+R%2E%2C+and+De+Mori%2C+R%2E+1995%2E+The+application+of+semantic+classi+cation+trees+to+natural+language+understanding%2E+IEEE+Transactions+on+Pattern+Analysis+and+Machine+Intelligence+17%285%29%3A449%7B460%2E+Lavra%15c)
6. Magerman, D. M. 1994. Natrual Lagnuage Parsing as Statistical Pattern Recognition . Ph.D. Dissertation, Stanford University. [find](https://scholar.google.com/scholar?q=Magerman%2C+D%2E+M%2E+1994%2E+Natrual+Lagnuage+Parsing+as+Statistical+Pattern+Recognition+%2E+Ph%2ED%2E+Dissertation%2C+Stanford+University%2E)
7. Miller, S.; Bobrow, R.; Ingria, R.; and Schwartz, R. 1994. Hidden understanding models of natural language. In Proceedings of the 32nd Annual Meeting of the Association for Computational Linguistics , 25{32. [find](https://scholar.google.com/scholar?q=Miller%2C+S%2E%3B+Bobrow%2C+R%2E%3B+Ingria%2C+R%2E%3B+and+Schwartz%2C+R%2E+1994%2E+Hidden+understanding+models+of+natural+language%2E+In+Proceedings+of+the+32nd+Annual+Meeting+of+the+Association+for+Computational+Linguistics+%2C)
8. Muggleton, S. H., ed. 1992. Inductive Logic Programming . New York, NY: Academic Press. [find](https://scholar.google.com/scholar?q=Muggleton%2C+S%2E+H%2E%2C+ed%2E+1992%2E+Inductive+Logic+Programming+%2E+New+York%2C+NY%3A+Academic+Press%2E)
9. Periera, F., and Shabes, Y. 1992. Inside-outside reestimation from partially bracketed corpora. In Proceedings of the 30th Annual Meeting of the Association for Computational Linguistics , 128{135. [our card](https://md.research.elamur.ai/free/en/papers/pereira-1992-inside-outside-reestimation-partially-bracketed/)
10. Pieraccini, R.; Tzoukermann, E.; Z. Gorelov, J. L. G.; Levin, E.; Lee, C. H.; and Wilpon, J. 1992. A speech understanding system based on statistical representation of semantics. In Proceedings ICASSP 92 . I{193{I{196. [find](https://scholar.google.com/scholar?q=Pieraccini%2C+R%2E%3B+Tzoukermann%2C+E%2E%3B+Z%2E+Gorelov%2C+J%2E+L%2E+G%2E%3B+Levin%2C+E%2E%3B+Lee%2C+C%2E+H%2E%3B+and+Wilpon%2C+J%2E+1992%2E+A+speech+understanding+system+based+on+statistical+representation+of+semantics%2E+In+Proceedings+ICASSP)
11. Thompson, C. A. 1995. Acquisition of a lexicon from semantic representations of sentences. In Proceeding of the 33rd Annual Meeting of the Association for Computational Linguistics , 335{337. [find](https://scholar.google.com/scholar?q=Thompson%2C+C%2E+A%2E+1995%2E+Acquisition+of+a+lexicon+from+semantic+representations+of+sentences%2E+In+Proceeding+of+the+33rd+Annual+Meeting+of+the+Association+for+Computational+Linguistics+%2C+335%7B337%2E)
12. Warren, D. H. D., and Pereira, F. C. N. 1982. An ecient easily adaptable system for interpreting natural language queries. American Journal of Computational Linguistics 8(3-4):110{122. [find](https://scholar.google.com/scholar?q=Warren%2C+D%2E+H%2E+D%2E%2C+and+Pereira%2C+F%2E+C%2E+N%2E+1982%2E+An+e%0Ecient+easily+adaptable+system+for+interpreting+natural+language+queries%2E+American+Journal+of+Computational+Linguistics+8%283%2D4%29%3A110%7B122%2E)
13. Zelle, J. M., and Mooney, R. J. 1993. Learning semantic grammars with constructive inductive logic programming. In Proceedings of the Eleventh National Conference on Arti cial Intelligence , 817{822. [find](https://scholar.google.com/scholar?q=Zelle%2C+J%2E+M%2E%2C+and+Mooney%2C+R%2E+J%2E+1993%2E+Learning+semantic+grammars+with+constructive+inductive+logic+programming%2E+In+Proceedings+of+the+Eleventh+National+Conference+on+Arti+cial+Intelligence+%2C+817%7B822%2E)
14. Zelle, J. M., and Mooney, R. J. 1994. Inducing deterministic Prolog parsers from treebanks: A machine learning approach. In Proceedings of the Twelfth National Conference on Arti cial Intelligence , 748{753. [find](https://scholar.google.com/scholar?q=Zelle%2C+J%2E+M%2E%2C+and+Mooney%2C+R%2E+J%2E+1994%2E+Inducing+deterministic+Prolog+parsers+from+treebanks%3A+A+machine+learning+approach%2E+In+Proceedings+of+the+Twelfth+National+Conference+on+Arti+cial+Intelligence+%2C+7)
15. Zelle, J., and Mooney, R. 1996. Comparative results on using inductive logic programming for corpus-based parser construction. In Wermter, S.; Rilo , E.; and Scheler, G., eds., Symbolic, Connectionist, and Statistical Approaches to Learning for Natural Language Processing . Springer Verlag. [find](https://scholar.google.com/scholar?q=Zelle%2C+J%2E%2C+and+Mooney%2C+R%2E+1996%2E+Comparative+results+on+using+inductive+logic+programming+for+corpus%2Dbased+parser+construction%2E+In+Wermter%2C+S%2E%3B+Rilo+%2C+E%2E%3B+and+Scheler%2C+G%2E%2C+eds%2E%2C+Symbolic%2C+Connectionist)
16. Zelle, J. M. 1995. Using Inductive Logic Programming to Automate the Construction of Natural Language Parsers . Ph.D. Dissertation, University of Texas, Austin, TX. available via [http://cs.utexas.edu/users/ml](http://cs.utexas.edu/users/ml). [↗](http://cs.utexas.edu/users/ml)