Controlled Crowdsourcing for High-Quality QA-SRL Annotation
This work addresses the need for reliable semantic annotations in NLP research, though it is incremental as it builds on existing QA-SRL methods.
The authors tackled the problem of low-quality crowdsourced QA-SRL annotations by developing an improved protocol involving worker selection, training, and data consolidation, resulting in a new high-quality gold evaluation dataset with drastically higher coverage.
Question-answer driven Semantic Role Labeling (QA-SRL) was proposed as an attractive open and natural flavour of SRL, potentially attainable from laymen. Recently, a large-scale crowdsourced QA-SRL corpus and a trained parser were released. Trying to replicate the QA-SRL annotation for new texts, we found that the resulting annotations were lacking in quality, particularly in coverage, making them insufficient for further research and evaluation. In this paper, we present an improved crowdsourcing protocol for complex semantic annotation, involving worker selection and training, and a data consolidation phase. Applying this protocol to QA-SRL yielded high-quality annotation with drastically higher coverage, producing a new gold evaluation dataset. We believe that our annotation protocol and gold standard will facilitate future replicable research of natural semantic annotations.