Combining Czech Dependency Parsers

Holan, Tomáš; Žabokrtský, Zdeněk

doi:10.1007/11846406_12

Cited by 7 publications

(4 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As already mentioned, the first experiments integrating a segment and/or a clause identification have brought promising results in dependency syntactic parsing (namely a combination of several parsers, one of them exploiting the idea of segmentation, see Holan and Ž abokrtský 2006) and in machine translation between related languages (namely Homola and Kuboň 2010). These results encourage us in our effort to prepare a sufficient amount of reliable data analyzed at the level of a sentence structure.…”

Section: Prague Dependency Treebank and Segment Annotationmentioning

confidence: 96%

See 1 more Smart Citation

Annotation of sentence structure

Lopatková

Homola

Klyueva

2011

Lang Resources & Evaluation

View full text Add to dashboard Cite

The focus of this article is on the creation of a collection of sentences manually annotated with respect to their sentence structure. We show that the concept of linear segments-linguistically motivated units, which may be easily detected automatically-serves as a good basis for the identification of clauses in Czech. The segment annotation captures such relationships as subordination, coordination, apposition and parenthesis; based on segmentation charts, individual clauses forming a complex sentence are identified. The annotation of a sentence structure enriches a dependency-based framework with explicit syntactic information on relations among complex units like clauses. We have gathered a collection of 3,444 sentences from the Prague Dependency Treebank, which were annotated with respect to their sentence structure (these sentences comprise 10,746 segments forming 6,341 clauses). The main purpose of the project is to gain a development data-promising results for Czech NLP tools (as a dependency parser or a machine translation system for related languages) that adopt an idea of clause segmentation have been already reported. The collection of sentences with annotated sentence structure provides the possibility of further improvement of such tools.

show abstract

Section: Prague Dependency Treebank and Segment Annotationmentioning

confidence: 96%

“…Jones (1994) for English or Ohno et al (2006) for Japanese; also first results for Czech are promising (esp. a clause segmentation in a rule-based dependency parser, see Holan and Ž abokrtský 2006, or in a machine translation system for related languages, as in Homola and Kuboň 2010).…”

Section: Motivationmentioning

confidence: 97%

Annotation of sentence structure

Lopatková

Homola

Klyueva

2011

Lang Resources & Evaluation

View full text Add to dashboard Cite

show abstract

“…UAS (unlabeled LAS (labeled attachment score) attachment score) Our best system (joint prediction, BERT, Flair) 93.10% 89. 93% Holan andŽabokrtský (2006) [16] 85.84% -Novák andŽabokrtský (2007) [26] 84.69% - Koo et al (2010) [21] † 87.32% -Treex framework (using MST parser&manual rules) [39] ‡ 83.93% 77.04% PDT 2.0 subset in CoNLL 2007 shared task; manually annotated POS tags available. Nakagawa (2007) [23] 86.28% 80.19% PDT 2.0 subset in CoNLL 2009 shared task; manually annotated POS tags available.…”

Section: Systemmentioning

confidence: 99%

“…Otherwise, Spoustová et al (2009) [31] used an averaged perceptron for POS tagging. For parsing the PDT, Holan and Zabokrtský (2006) [16] and Novák and Žabokrtský (2007) [26] used a combination of non-neural parsing techniques .…”

Section: Related Workmentioning

confidence: 99%

Czech Text Processing with Contextual Embeddings: POS Tagging, Lemmatization, Parsing and NER

Straka

Straková

Hajič

2019

Text, Speech, and Dialogue

View full text Add to dashboard Cite

Contextualized embeddings, which capture appropriate word meaning depending on context, have recently been proposed. We evaluate two methods for precomputing such embeddings, BERT and Flair, on four Czech text processing tasks: part-of-speech (POS) tagging, lemmatization, dependency parsing and named entity recognition (NER). The first three tasks, POS tagging, lemmatization and dependency parsing, are evaluated on two corpora: the Prague Dependency Treebank 3.5 and the Universal Dependencies 2.3. The named entity recognition (NER) is evaluated on the Czech Named Entity Corpus 1.1 and 2.0. We report state-of-the-art results for the above mentioned tasks and corpora.

show abstract