Çağrı Çöltekin
Çağrı Çöltekin works in computational linguistics, with papers here that span multilingual syntactic analysis and the construction of annotated language resources. One strand concerns Universal Dependencies and parsing across many languages within a common annotation framework. Another addresses languages that have received little attention in natural language processing, Turkish in particular, by building new corpora with careful annotation schemes. Across both strands the work is grounded in data creation and shared evaluation rather than in a single model or method.
The most cited of these papers is CoNLL 2017 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies. It set participating systems the job of going from unprocessed text, without gold tokens, sentence boundaries or tags, to full dependency trees for dozens of languages. Test sets were kept hidden until the end and included "surprise" languages with almost no training material. Scoring relied on a labelled attachment measure computed over tokens aligned to the gold standard. The second most cited work, A Corpus of Turkish Offensive Language on Social Media, filled the absence of any labelled offensive-language resource for Turkish. It sampled tweets at random from the stream rather than filtering by keywords, which avoids topic bias. The tweets were labelled with a layered scheme modelled on target-oriented efforts such as OffensEval, separating targeted from untargeted offense and then distinguishing types of target.
The newer work turns from parsing toward social media text and the area of sentiment and emotion, while keeping a broader interest in multilingual and cross-lingual NLP. Çöltekin is based at the University of Tübingen, after earlier papers from the University of Groningen. The shared task brought collaboration with a large group of researchers, including Daniel Zeman, Christopher D. Manning and Filip Ginter.
Çağrı Çöltekin has 3 papers in our library, in 2017–2025, most of them on Universal Dependencies, Treebanks, Sentiment & emotion; often with Ali Elkahky, Aljoscha Burchardt, Anna Nedoluzhko.
OpenAlex · ORCID · Wikidata · h-index 19 · 91 works · 1,872 citations OpenAlex, 2026-10-14
📄 3 works in our library · 🔗 330 citations in our library · 2 more in our library without a page yet · ⭐ Paper Star rank #2281
Affiliations
Topics
Universal Dependencies (2) Treebanks (1) Sentiment & emotion (1) Multilingual & cross-lingual NLP (1) Morphology (1)
Works by year
2020s
2010s
Papers
- CoNLL 2017 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies (2017 · 604 citations)
The shared task asks systems to parse raw text, with no gold tokenization, sentence splitting or tags, into UD trees for 49 languages. - A Corpus of Turkish Offensive Language on Social Media (2020 · 200 citations)
This paper addresses the lack of any annotated resource for offensive language in Turkish, a language with a very active Twitter user base but almost no attention in this research area. - Developing a Universal Dependencies Treebank for Alaskan Gwich'in (2025)
The authors created the first UD treebank for Gwich'in. It is an Athabaskan language of Alaska and northwestern Canada that is endangered: it is spoken by at most 300 people, most of them elderly. Indigenous languages of the USA and Canada are barely represented in UD, and the syntax of Athabaskan languages has been…
Also in our library
Co-authors here
- Ali Elkahky
- Aljoscha Burchardt
- Anna Nedoluzhko
- Christopher D. Manning
- Daniel Zeman
- Dima Taji
- Emily Pitler
- Filip Ginter
- Francis Tyers
- Georg Rehm
- Hans Uszkoreit
- Héctor Martínez Alonso
Cited by
This page in Markdown → https://md.research.elamur.ai/free/en/authors/cagri-coltekin/