--- title: "Expanding the Lexicon for a Resource-Poor Language Using a" description: "The approach offers a cheap way to expand lexicons for morphologically rich, under-resourced languages. 84% precision. 8 references." authors: ["Michael Gasser"] year: 2010 venue: "Proceedings of the Seventh International Conference on Language Resources and Evaluation (LREC'10)" source: "https://aclanthology.org/L10-1629/" licence: "CC BY-NC 4.0 (ELRA, LREC proceedings)" licence_url: "https://creativecommons.org/licenses/by-nc/4.0/" summary_licence: "Elamur Cosmic License" canonical: "https://research.elamur.ai/free/en/papers/gasser-2010-expanding-lexicon-resource-poor-language/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/gasser-2010-expanding-lexicon-resource-poor-language/) [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/page/2/) / Expanding the Lexicon for a Resource-Poor Language Using a Morphological Analyzer and a Web Crawler # Expanding the Lexicon for a Resource-Poor Language Using a Morphological Analyzer and a Web Crawler **Authors:** [Michael Gasser](https://md.research.elamur.ai/free/en/authors/michael-gasser/) · **Year:** 2010 · **Venue:** [Proceedings of the Seventh International Conference on Language Resources and Evaluation (LREC'10)](https://md.research.elamur.ai/free/en/venues/lrec/) ([LREC 2010](https://md.research.elamur.ai/free/en/venues/lrec/2010/)) · **Publisher:** [European Language Resources Association (ELRA)](https://md.research.elamur.ai/free/en/publishers/european-language-resources-association-elra/) · **Licence:** [CC BY-NC 4.0 (ELRA, LREC proceedings)](https://creativecommons.org/licenses/by-nc/4.0/) **Links:** [Publisher's page](https://aclanthology.org/L10-1629/) · [PDF](https://aclanthology.org/L10-1629.pdf) · [OpenAlex](https://openalex.org/W3789944) **Citations:** 5 citations (OpenAlex, 2026-10-14) ## Summary Tigrinya is an Ethio-Semitic language with very few digital resources. Its only usable digital lexicon, the online dictionary of Zacarias, yielded just 598 verb roots. Gasser shows how a morphological analyzer can enlarge this lexicon from raw web text, with no syntactic parser. A web crawler gathered more than 200,000 Tigrinya wordform types. These were run through HornMorpho, a finite-state analyzer built on transducers weighted with feature structures. When the lexicon-based analysis failed, a lexicon-free "guesser" analyzer proposed a root. The candidates were then filtered in two steps: - They were checked against a profile of how known roots typically behave morphosyntactically. - Confusable candidates were grouped into competition sets, and a trigram model over root consonants chose one winner per set. Of the 417 surviving candidates, 275 (66%) were confirmed by a printed dictionary. Treating confusions that differ only in root category as a single root raises precision to 84%. The same analyses also show which derivational categories known roots accept. The approach offers a cheap way to expand lexicons for morphologically rich, under-resourced languages. ## Key points - The starting lexicon had 598 Tigrinya verb roots. The crawler list contained over 200,000 wordform types, analyzed by HornMorpho, whose lexical analyzer is backed by a root "guesser". - One verb root can surface in hundreds of thousands of wordforms, from 32 stem templates (4 TAM × 8 derivational categories) plus many affixes. Ge'ez script does not mark gemination, so several root categories become ambiguous. - Root selection used competition sets: - 678 sets were formed, each holding 2 to 19 roots. - The trigram-preferred root won each set, leaving 417 candidates. - 275 of these (66%) were verified in Kane (2000). - Of the 142 errors, 75 were the right consonants with the wrong root category. Counting those as correct gives 84% precision. Other errors came from: - typos in very frequent roots; - a bug in handling root-final y; - orthographic variants not yet covered; - 7 cases that remain unexplained. - Subject and object agreement and derivational forms seen in the data reveal root-specific constraints. For example, some roots occur almost only in the transitive/causative form, and the split between intransitive and transitive roots is not clear-cut. ## Interesting ideas - A lexicon-free "guesser" runs only when the lexical analyzer fails, so unknown words still get a full structured analysis rather than nothing. - Instead of forcing a choice between confusable hypotheses that the data cannot tell apart, the paper suggests keeping them merged as one entry until disambiguating evidence appears. - Raw-text morphology (agreement markers, derivational forms) is mined to learn which valency and derivation patterns each lexeme allows. - A "typical profile" built from trusted known items is used as a filter for new candidates. - The error analysis sorts mistakes by cause (category confusion, typos, a single analyzer bug, missing variants), turning failures into concrete fixes. ## For Mova How these ideas could serve Mova, our small open language model — its files, stages and tests: - **I1** (tokens): add a fallback that guesses lemma and features from affix rules for out-of-vocabulary English words, marking the result as guessed so later steps can see where the analysis came from. - **I2** (learn): when witness parsers disagree in ways the evidence cannot settle, store an underspecified silver label rather than an arbitrary pick, and resolve it later from more data. - **I3** (world): collect counts of which argument structures and alternations each English verb shows in parsed text, and use them to check or extend VerbNet-style verb classes. - **I4** (learn): build behavioural profiles from gold entries (frequency, contexts, feature spread) and reject induced lexical items or repair rules that fall far outside them. - **I5** (learn): label evaluation errors by source so each class can produce a targeted repair rule or a bug report rather than just a lower score. *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 - [Indiana University Bloomington](https://md.research.elamur.ai/free/en/universities/indiana-university-bloomington/) — University, Bloomington, United States · 18 works · 2,320 citations - [Indiana University](https://md.research.elamur.ai/free/en/universities/indiana-university/) — University, Bloomington, United States · 24 works · 2,995 citations 📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 9,086 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 8,889 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 9,296 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 9,206 works 🧭 Topics (3): [Semitic languages](https://md.research.elamur.ai/free/en/topics/semitic-languages/) · 1,484 works · [🧩 Morphology](https://md.research.elamur.ai/free/en/topics/morphology/) · 277 works · [🌱 Low-resource languages](https://md.research.elamur.ai/free/en/topics/low-resource-languages/) · 166 works ## Cite Cite as: Michael Gasser. 2010. Expanding the Lexicon for a Resource-Poor Language Using a Morphological Analyzer and a Web Crawler. Proceedings of the Seventh International Conference on Language Resources and Evaluation (LREC'10). The paper itself, not this page. BibTeX from the library's record of the work. ```bibtex @inproceedings{gasser2010expanding, title = {{Expanding the Lexicon for a Resource-Poor Language Using a Morphological Analyzer and a Web Crawler}}, author = {Michael Gasser}, booktitle = {Proceedings of the Seventh International Conference on Language Resources and Evaluation (LREC'10)}, year = {2010}, doi = {10.63317/5bnxfwtiwm6m}, url = {https://doi.org/10.63317/5bnxfwtiwm6m} } ``` - [← previous A Lucene and Maximum Entropy Model Based Hedge Detection System](https://md.research.elamur.ai/free/en/papers/chen-2010-lucene-maximum-entropy-model-based/) - [next → Towards multi-lingual summarization: A comparative analysis of sentence extraction methods on English and Hebrew corpora](https://md.research.elamur.ai/free/en/papers/litvak-2010-multi-lingual-summarization-comparative-analysis/) - [↑ back to papers of 2010](https://md.research.elamur.ai/free/en/papers/2010/page/2/) ## References 8 references, 1 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet. 1. Jan Amtrup. 2003. Morphology in machine translation systems: Efficient integration of finite state transducers and feature structure descriptions. Machine Translation, 18:213–235. [↗](https://doi.org/10.1007/s10590-004-2476-5) 2. Kenneth R. Beesley and Lauri Karttunen. 2003. Finite State Morphology. CSLI Publications, Stanford, CA, USA. [find](https://scholar.google.com/scholar?q=Kenneth+R%2E+Beesley+and+Lauri+Karttunen%2E+2003%2E+Finite+State+Morphology%2E+CSLI+Publications%2C+Stanford%2C+CA%2C+USA%2E) 3. Yitna Firdyiwek and Daniel Yaqob. 1997. The system for Ethiopic representation in ASCII. URL: citeseer.ist.psu.edu/56365.html. [find](https://scholar.google.com/scholar?q=Yitna+Firdyiwek+and+Daniel+Yaqob%2E+1997%2E+The+system+for+Ethiopic+representation+in+ASCII%2E+URL%3A+citeseer%2Eist%2Epsu%2Eedu%2F56365%2Ehtml%2E) 4. Michael Gasser. 2009. [Semitic morphological analysis and generation using finite state transducers with feature structures](https://md.research.elamur.ai/free/en/papers/gasser-2009-semitic-morphological-analysis-generation-finite/). In Proceedings of the 12th Conference of the European Chapter of the ACL, pages 309–317, Athens, Greece. 5. Biniam Gebremichael. 2009. Wordlist and spell checking for Amharic and Tigrigna. Available online at [http://www.cs.ru.nl/](http://www.cs.ru.nl/) biniam/geez/crawl.php. [↗](http://www.cs.ru.nl/) 6. Thomas L. Kane. 2000. Tigrinya-English Dictionary. Dunwoody Press, Springfield, VA, USA. [find](https://scholar.google.com/scholar?q=Thomas+L%2E+Kane%2E+2000%2E+Tigrinya%2DEnglish+Dictionary%2E+Dunwoody+Press%2C+Springfield%2C+VA%2C+USA%2E) 7. Wolf Leslau. 1941. Documents Tigrigna: Grammaire et Textes. Libraire C. Klincksieck, Paris. [find](https://scholar.google.com/scholar?q=Wolf+Leslau%2E+1941%2E+Documents+Tigrigna%3A+Grammaire+et+Textes%2E+Libraire+C%2E+Klincksieck%2C+Paris%2E) 8. Efrem Zacarias. 2009. Memhir.org dictionaries (English-Tigrinya, Hebrew-Tigrinya dictionaries). Available at [http://www.memhr.org/dic/](http://www.memhr.org/dic/). [↗](http://www.memhr.org/dic/) *Updated 24 October 2026*