Crowdsourcing Universal Part-Of-Speech Tags for Code-Switching
This addresses the problem of limited labeled data for code-switching, which affects NLP applications for large bilingual populations, but it is incremental as it focuses on annotation methods rather than new models or broad breakthroughs.
The paper tackles the lack of high-quality annotated data for code-switching in NLP by crowdsourcing universal part-of-speech tags for the Miami Bangor Corpus of Spanish-English speech, achieving overall agreement between 0.95 and 0.96 and recall between 0.87 and 0.99 across tasks.
Code-switching is the phenomenon by which bilingual speakers switch between multiple languages during communication. The importance of developing language technologies for codeswitching data is immense, given the large populations that routinely code-switch. High-quality linguistic annotations are extremely valuable for any NLP task, and performance is often limited by the amount of high-quality labeled data. However, little such data exists for code-switching. In this paper, we describe crowd-sourcing universal part-of-speech tags for the Miami Bangor Corpus of Spanish-English code-switched speech. We split the annotation task into three subtasks: one in which a subset of tokens are labeled automatically, one in which questions are specifically designed to disambiguate a subset of high frequency words, and a more general cascaded approach for the remaining data in which questions are displayed to the worker following a decision tree structure. Each subtask is extended and adapted for a multilingual setting and the universal tagset. The quality of the annotation process is measured using hidden check questions annotated with gold labels. The overall agreement between gold standard labels and the majority vote is between 0.95 and 0.96 for just three labels and the average recall across part-of-speech tags is between 0.87 and 0.99, depending on the task.