Kaja Dobrovoljc

Kaja Dobrovoljc

Auto-generated profile, based on publicly available information

Kaja Dobrovoljc

Faculty of Arts, University of Ljubljana, Jožef Stefan Institute

Research Focus & Metrics

Academic Career Length

13Years (2014-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications

1 January 2017
Verified
Conference Paper
Closed access
KDKaja Dobrovoljc
Kaja Dobrovoljc
TETomaž Erjavec
Tomaž Erjavec
SKSimon Krek
Simon Krek

22 December 2025 in Contributions to Contemporary History
Verified
Journal Article
Open access
PDF file
LTLuka Terčon
Luka Terčon
KDKaja Dobrovoljc
Kaja Dobrovoljc
NLNikola Ljubešić
Nikola Ljubešić
We present CLASSLA-Stanza, a pipeline for automatic linguistic annotation of South Slavic languages, which is based on the Stanza natural language processing pipeline. We describe the main improvements in CLASSLA-Stanza with respect to Stanza and give a detailed description of the model training process for the latest 2.2 release of the pipeline. We also report performance scores produced by the pipeline for different languages and language varieties. CLASSLA-Stanza exhibits consistently high performance across all the supported languages and outperforms its parent pipeline Stanza at all the supported tasks. We also present the pipeline’s new functionality that enables efficient processing of web data and describe the efficiency of the pipeline for annotating written transcripts of spoken data.

1 January 2014
Verified
Conference Paper
Closed access
ŽAŽeljko Agić
Željko Agić
JTJörg Tiedemann
Jörg Tiedemann
DMDanijela Merkler
Danijela Merkler
SKSimon Krek
Simon Krek
KDKaja Dobrovoljc
Kaja Dobrovoljc
SMSara Moze
Sara Moze

1 October 2018 in Language Resources and Evaluation
Verified
Journal Article
Closed access
PDF file
ŠHŠpela Arhar Holdt
Špela Arhar Holdt
KDKaja Dobrovoljc
Kaja Dobrovoljc
NLNataša Logar
Nataša Logar

13 April 2020 in International Journal of Lexicography
Verified
Journal Article
Closed access
KDKaja Dobrovoljc
Kaja Dobrovoljc
In view of the pervasiveness of formulaic language in human communication and the growing awareness of its relevance to modern lexicography, this study presents a corpus-driven identification, analysis and comparison of dictionary-relevant formulaic sequences in reference corpora of written and spoken Slovenian. The sequences were identified using a semi-automatic approach, whereby the most frequently recurring word combinations in each corpus were ranked according to their statistical salience and manually inspected for formulaic expressions with lexicographic relevance. Despite its semantic heterogeneity, the resulting list illustrates the distinct characteristics of formulaic multi-word expressions, such as high frequency of usage, prevalent inclusion of grammatical words and common non-propositional meaning, especially in speech, where research revealed numerous understudied formulaic expressions related to interaction management and mitigation. The final evaluation of measures used in the identification process demonstrates their relative suitability for corpus-driven identification of dictionary-relevant formulaic expressions, with their precision varying in relation to corpus size and length of sequences under investigation.