Julia Hockenmaier
Julia Hockenmaier works in natural language processing, where the papers here deal with two linked concerns: how to build large annotated resources for language, and how to make those resources reliable. One strand addresses grammar formalisms, especially Combinatory Categorial Grammar (CCG), and the conversion of existing treebank annotation into richer syntactic representations. The other turns to collecting language data from crowd workers, with attention to how quality can be controlled when there is no fixed answer to check against.
The more cited of the two works is CCGbank (2007). It describes an algorithm that turns Penn Treebank phrase-structure trees into CCG derivations. Along the way it makes binarization, head and complement/adjunct decisions, and it recovers word-to-word dependencies, including long-range ones arising from extraction, control, raising and coordination. The original annotation is first cleaned of noise and inconsistencies, and the resulting corpus covers 99.44% of the Penn sentences. The paper also reports that rule instantiations follow a Zipfian distribution. The later paper, Collecting Image Annotations Using Amazon's Mechanical Turk (2010), moves toward joint work on vision and language. It gathers full-sentence descriptions of photographs and compares ways of keeping free-text contributions accurate. Among its findings, screening workers with a short qualification test before they wrote improved quality more than any correction applied afterwards. Workers who scored perfectly on control items did not agree more closely with experts.
Hockenmaier's current institution is the University of Illinois Urbana-Champaign, with bylines there from 2008 onward, and a 2026 paper is also linked to the Korea Advanced Institute of Science and Technology. Earlier papers came from the University of Pennsylvania and the University of Edinburgh in the 2000s. Co-authors on the site include Mark Steedman, and the papers here fall under Treebanks and Syntax & grammar formalisms.
Julia Hockenmaier has 2 papers in our library, in 2007–2010, most of them on Treebanks, Syntax & grammar formalisms; often with Mark Steedman.
current affiliationKorea Advanced Institute of Science and Technology, Daejeon, South Korea; University of Illinois Urbana-Champaign, Urbana, United States
OpenAlex · ORCID · h-index 32 · 115 works · 10 624 citations OpenAlex, 2026-10-14
2 works in our library · 512 citations in our library · 6 more works in our library · ⭐ Paper Star rank #1770
Affiliations
- University of Illinois Urbana-Champaign · Urbana, United States — current · 2008–2026 on 81 papers
- Korea Advanced Institute of Science and Technology · Daejeon, South Korea — current · 2026 on 1 paper
- University of Pennsylvania · Philadelphia, United States — 2003–2007 on 11 papers
- University of Edinburgh · Edinburgh, United Kingdom — 2000–2007 on 10 papers
Topics
Works by year
2010s
2000s
Papers
- CCGbank: A Corpus of CCG Derivations and Dependency Structures Extracted from the Penn Treebank (2007 · 611 citations)
The paper gives an algorithm that translates Penn Treebank phrase-structure trees into Combinatory Categorial Grammar derivations, with binarization, head and complement/adjunct decisions, and word-word dependencies, including long-range ones from extraction, control, raising, right-node raising and coordination. - Collecting Image Annotations Using Amazon's Mechanical Turk (2010 · 562 citations)
The authors wanted a corpus that pairs photographs with plain full-sentence descriptions of what they show, such as objects, their attributes, relations and actions.
Also in our library
Co-authors here
Cited by
This page in Markdown → https://md.research.elamur.ai/free/en/authors/julia-hockenmaier/