How lemmatisation and derivational annotation affect productivity measures
The case of deverbal agent nouns in the Joint Corpus of Lithuanian
Bibliographic Data
| ID | 22343147 |
|---|---|
| Authors | Jurgis Pakerys (0000-0002-9944-8598, Vilnius University), Virginijus Dadurkevičius (0000-0001-8602-6591, Vytautas Magnus University), Agnė Navickaitė-Klišauskienė (Vilnius University) |
| Year | 2024 |
| Pages | 138-151 |
| Publication date | 2024-12-16 |
| Peer Reviewed | Yes |
| Open Access | Yes |
| Type | ARTICLE |
| Venue | Valoda: nozīme un forma / Language: Meaning and Form (CONFERENCE) |
| Journal identifiers | ISSN: 2255-9256 • E-ISSN: 2256-0602 |
| Publisher | University of Latvia (PUBLISHER • LV) |
| DOI | 10.22364/vnf.15.09 |
| OpenAlex | W4405423387 |
| Language | EN |
| References cited | 11 |
We discuss the automatic and manual stages of the lemmatisation and annotation of the Joint Corpus of Lithuanian (1.3 billion words) used to measure derivational productivity. As a case study, we present data of three productive deverbal agent noun suffixes in Lithuanian, -toj-, -ėj-, -ik-, and measure their realized, expanding, and potential productivity. We show that an additional semi-automatic lemmatisation and a manual derivational annotation significantly increase type and hapax counts. We also note that lemmatisation is affected by an artificially increased number of lemmas due to homographic forms unresolved by the lemmatiser. After the manual disambiguation of hapaxes, the numbers of feminine formations in -toj-(a) and -ėj-(a) were the most significantly reduced
Annotation · Economics · Linguistics · Lithuanian · Natural language processing · Noun · Productivity · Computer Science · Engineering · Natural Language Processing Techniques · Philosophy · Psychology · Speech and dialogue systems · Topic Modeling · Artificial Intelligence
| Citation velocity | historical |
|---|---|
| Highly cited | No |