Steiner, Petra ; Ruppenhofer, Josef
Hrsg.: Fisseni, Bernhard ; Schröder, Bernhard ; Zesch, Torsten (2015)
Growing Trees from Morphs: Towards Data-Driven Morphological Parsing.
International Conference of the German Society for Computational Linguistics and Language Technology (GSCL 2015). Duisburg, Germany (30.09.2015-02.10.2015)
Konferenzveröffentlichung, Bibliographie
Kurzbeschreibung (Abstract)
We present a quantitative approach to disambiguating flat morphological analyses and producing more deeply structured analyses. Based on existing morphological segmentations, possible combinations of resulting word trees for the next level are filtered first by criteria of linguistic plausibility and then by weighting procedures based on the geometric mean. The frequencies for weighting are derived from three different sources (counts of morphs in a lexicon, counts of largest constituents in a lexicon, counts of token frequencies in a corpus) and can be used either to find the best analysis on the level of morphs or on the next higher constituent level. The evaluation shows that for this task corpus-based frequency counts are slightly superior to counts of lexical data
| Typ des Eintrags: | Konferenzveröffentlichung |
|---|---|
| Erschienen: | 2015 |
| Herausgeber: | Fisseni, Bernhard ; Schröder, Bernhard ; Zesch, Torsten |
| Autor(en): | Steiner, Petra ; Ruppenhofer, Josef |
| Art des Eintrags: | Bibliographie |
| Titel: | Growing Trees from Morphs: Towards Data-Driven Morphological Parsing |
| Sprache: | Englisch |
| Publikationsjahr: | 2015 |
| Verlag: | Gesellschaft für Sprachtechnologie and Computerlinguistik e.V. |
| Buchtitel: | Proceedings of the International Conference of the German Society for Computational Linguistics and Language Technology |
| Veranstaltungstitel: | International Conference of the German Society for Computational Linguistics and Language Technology (GSCL 2015) |
| Veranstaltungsort: | Duisburg, Germany |
| Veranstaltungsdatum: | 30.09.2015-02.10.2015 |
| URL / URN: | https://konvens.org/proceedings/2015/index.html |
| Kurzbeschreibung (Abstract): | We present a quantitative approach to disambiguating flat morphological analyses and producing more deeply structured analyses. Based on existing morphological segmentations, possible combinations of resulting word trees for the next level are filtered first by criteria of linguistic plausibility and then by weighting procedures based on the geometric mean. The frequencies for weighting are derived from three different sources (counts of morphs in a lexicon, counts of largest constituents in a lexicon, counts of token frequencies in a corpus) and can be used either to find the best analysis on the level of morphs or on the next higher constituent level. The evaluation shows that for this task corpus-based frequency counts are slightly superior to counts of lexical data |
| Fachbereich(e)/-gebiet(e): | Zentrale Einrichtungen Zentrale Einrichtungen > Universitäts- und Landesbibliothek (ULB) |
| Hinterlegungsdatum: | 19 Jun 2023 09:23 |
| Letzte Änderung: | 19 Jun 2023 09:23 |
| PPN: | |
| Export: | |
| Suche nach Titel in: | TUfind oder in Google |
 |
Frage zum Eintrag |
Optionen (nur für Redakteure)
 |
Redaktionelle Details anzeigen |
Drucken
Impressum