SynthBio: A Case Study in Human-AI Collaborative Curation of Text Datasets
arXiv:2111.06467
Abstract
NLP researchers need more, higher-quality text datasets. Human-labeled datasets are expensive to collect, while datasets collected via automatic retrieval from the web such as WikiBio are noisy and can include undesired biases. Moreover, data sourced from the web is often included in datasets used to pretrain models, leading to inadvertent cross-contamination of training and test sets. In this work we introduce a novel method for efficient dataset curation: we use a large language model to provide seed generations to human raters, thereby changing dataset authoring from a writing task to an editing task. We use our method to curate SynthBio - a new evaluation set for WikiBio - composed of structured attribute lists describing fictional individuals, mapped to natural language biographies. We show that our dataset of fictional biographies is less noisy than WikiBio, and also more balanced with respect to gender and nationality.
10 pages, 2 figures, accepted to NeurIPS 2021 Datasets and Benchmarks Track
References in corpus (9)
- Snorkel: Rapid Training Data Creation with Weak Supervision
- The Impact of Multiple Parallel Phrase Suggestions on Email Input and Composition Behaviour of Native and Non-Native English Writers
- Assessing Social and Intersectional Biases in Contextualized Word Representations
- Human-in-the-Loop for Data Collection: a Multi-Target Counter Narrative Dataset to Fight Online Hate Speech
- Process for Adapting Language Models to Society (PALMS) with Values-Targeted Datasets
- Data, Power and Bias in Artificial Intelligence
- Nine Potential Pitfalls when Designing Human-AI Co-Creative Systems
- Topic-Preserving Synthetic News Generation: An Adversarial Deep Reinforcement Learning Approach
- Generating Synthetic Text Data to Evaluate Causal Inference Methods