Refine
Document Type
- Conference Proceeding (3)
- Part of a Book (1)
Has Fulltext
- yes (4)
Keywords
- Korpus <Linguistik> (3)
- Annotation (2)
- Automatische Klassifikation (1)
- Corpus annotation (1)
- Corpus query language (1)
- Corpus technology (1)
- Datenbanksystem (1)
- Grammatik (1)
- HPSG (1)
- Head-driven phrase structure grammar (1)
Publicationstate
Reviewstate
- Peer-Review (2)
- (Verlags)-Lektorat (1)
In this paper, we deal with register-driven variation from a probabilistic perspective, as proposed in Schäfer, Bildhauer, Pankratz, Müller (2022). We compare two approaches to analyse this variation within HPSG. On the one hand, we consider a multiple-grammar approach and combine it with the architecture proposed in the CoreGram project Müller (2015) - discussing its advantages and disadvantages. On the other hand, we take into account a single-grammar approach and argue that it appears to be superior due to its computational efficiency and cognitive plausibility.
In this paper, I present the COW14 tool chain, which comprises a web corpus creation tool called texrex, wrappers for existing linguistic annotation tools as well as an online query software called Colibri2. By detailed descriptions of the implementation and systematic evaluations of the performance of the software on different types of systems, I show that the COW14 architecture is capable of handling the creation of corpora of up to at least 100 billion tokens. I also introduce our running demo system which currently serves corpora of up to roughly 20 billion tokens in Dutch, English, French, German, Spanish, and Swedish
Eine reichhaltige Auszeichnung mit Metadaten ist für alle Arten von Korpora für die linguistische Forschung wünschenswert. Für große Korpora (insbesondere Webkorpora) müssen Metadaten automatisch erzeugt werden, wobei die Genauigkeit der Auszeichnung besonders kritisch ist. Wir stellen einen Ansatz zur automatischen Klassifikation nach Themengebiet (Topikdomäne) vor, die auf dem lexikalischen Material in Texten basiert. Dazu überführen wir weniger gut interpretierbare Ergebnisse aus einer so genannten Topikmodellierung mittels eines überwachten Lernverfahrens in eine besser interpretierbare Kategorisierung nach 13 Themengebieten. Gegenüber (automatisch erzeugten) Klassifikationen nach Genre, Textsorte oder Register, die zumeist auf Verteilungen grammatischer Merkmale basieren, erscheint eine solche thematische Klassifikation geeigneter, um zusätzliche Kontrollvariablen für grammatische Variationsstudien bereitzustellen. Wir evaluieren das Verfahren auf Webtexten aus DECOW14 und Zeitungstexten aus DeReKo, für die jeweils getrennte Goldstandard-Datensätze manuell annotiert wurden.
In this paper, we describe preliminary results from an ongoing experiment wherein we classify two large unstructured text corpora—a web corpus and a newspaper corpus—by topic domain (or subject area). Our primary goal is to develop a method that allows for the reliable annotation of large crawled web corpora with meta data required by many corpus linguists. We are especially interested in designing an annotation scheme whose categories are both intuitively interpretable by linguists and firmly rooted in the distribution of lexical material in the documents. Since we use data from a web corpus and a more traditional corpus, we also contribute to the important field of corpus comparison and corpus evaluation. Technically, we use (unsupervised) topic modeling to automatically induce topic distributions over gold standard corpora that were manually annotated for 13 coarse-grained topic domains. In a second step, we apply supervised machine learning to learn the manually annotated topic domains using the previously induced topics as features. We achieve around 70% accuracy in 10-fold cross validations. An analysis of the errors clearly indicates, however, that a revised classification scheme and larger gold standard corpora will likely lead to a substantial increase in accuracy.