The Extraordinary Failure of Complement Coercion Crowdsourcing

Yanai Elazar, Victoria Basmova, Shauli Ravfogel, Yoav Goldberg, Reut Tsarfaty

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review


Crowdsourcing has eased and scaled up the collection of linguistic annotation in recent years. In this work, we follow known methodologies of collecting labeled data for the complement coercion phenomenon. These are constructions with an implied action -- e.g., "I started a new book I bought last week", where the implied action is reading. We aim to collect annotated data for this phenomenon by reducing it to either of two known tasks: Explicit Completion and Natural Language Inference. However, in both cases, crowdsourcing resulted in low agreement scores, even though we followed the same methodologies as in previous work. Why does the same process fail to yield high agreement scores? We specify our modeling schemes, highlight the differences with previous work and provide some insights about the task and possible explanations for the failure. We conclude that specific phenomena require tailored solutions, not only in specialized algorithms, but also in data collection methods.
Original languageEnglish
Title of host publicationProceedings of the First Workshop on Insights from Negative Results in NLP, Insights 2020, Online, November 19, 2020
EditorsAnna Rogers, João Sedoc, Anna Rumshisky
PublisherAssociation for Computational Linguistics
Number of pages11
StatePublished - 2020


Dive into the research topics of 'The Extraordinary Failure of Complement Coercion Crowdsourcing'. Together they form a unique fingerprint.

Cite this