Refine
Year of publication
- 2018 (2)
Document Type
- Part of a Book (2) (remove)
Language
- English (2)
Has Fulltext
- yes (2)
Is part of the Bibliography
- yes (2)
Keywords
- Terminologie (2)
- Automatische Sprachverarbeitung (1)
- CLARIN (1)
- Deutsch (1)
- Grammatik (1)
- Grammis (1)
- ISO (1)
- Informationsmanagement (1)
- Linguistik (1)
- Standardisierung (1)
Publicationstate
- Veröffentlichungsversion (2) (remove)
Reviewstate
- Peer-Review (2)
In this paper, we present our approach to automatically extracting German terminology in the domain of grammar using texts from the online information system grammis as our corpus. We analyze existing repositories of German grammatical terminology and develop Part-of-speech patterns for our extraction thereby showing the importance of unigrams in this domain. We contrast the results of the automatic extraction with a manually extracted standard. By comparing the performance of well-known statistical measures, we show how measures based on corpus comparison outperform alternative methods.
The present submission reports on a pilot project conducted at the Institute for the German Language (IDS), aiming at strengthening the connection between ISO TC37SC4 “Language Resource Management” and the CLARIN infrastructure. In terminology management, attempts have recently been made to use graph-theoretical analyses to get a better understanding of the structure of terminology resources. The project described here aims at applying some of these methods to potentially incomplete concept fields produced over years by numerous researchers serving as experts and editors of ISO standards. The main results of the project are twofold. On the one hand, they comprise concept networks dynamically generated from a relational database and browsable by the user. On the other, the project has yielded significant qualitative feedback that will be offered to ISO. We provide the institutional context of this endeavour, its theoretical background, and an overview of data preparation and tools used. Finally, we discuss the results and illustrate some of them.