Changing the World by Changing the Data

By Anna Rogers. 2021

81 citations recorded by the source corpus.

NLP community is currently investing a lot more research and resources into development of deep learning models than training data. While we have made a lot of progress, it is now clear that our models learn all kinds of spurious patterns, social biases, and annotation artifacts. Algorithmic solutions have so far had limited success. An alternative that is being actively discussed is more careful design of datasets so as to deliver specific signals. This position paper maps out the arguments for and against data curation, and argues that fundamentally the point is moot: curation already is and will be happening, and it is changing the world. The question is only how much thought we want to invest into that process.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Evaluation & Benchmarks

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Evaluation & Benchmarks

Citations in the corpus

Repairing the Cracked Foundation: A Survey of Obstacles in Evaluation Practices for Generated Text (2022, background): “Rogers (2021) summarize discussions around data curation, the act of manufacturing distributions that differ from naturally occurring ones, pointing out that dataset creators should maybe not be the ones deciding what distribution should represent the world, and that studying the world as it is…”
The ROOTS Search Tool: Data Transparency for LLMs (2023, background): “In the Computer Vision community, related efforts 5 target large text and image datasets such as LAION (Schuhmann et al., 2022, 2021).”
The BigScience ROOTS Corpus: A 1.6TB Composite Multilingual Dataset (2023, background): “While this kind of documentation is far from standard practice, it is becoming increasingly common given recent calls for better documentation (Rogers, 2021; Bender et al., 2021) as well as empirical studies on data memorization in language models (Carlini et al., 2019, 2022).”
Data Governance in the Age of Large-Scale Data-Driven Language Technology (2022, background): “Indeed, in order to support such ambitions of generality, this collection would need to include language data from a great diversity of carefully curated sources to minimize harms in downstream applications [12, 120], with international rights holders spanning multiple jurisdictions, and extend to m”
Just What do You Think You're Doing, Dave?' A Checklist for Responsible Data Use in NLP (2021, unknown): “NLP is in dire need of research into the cost–beneﬁt of the use of its systems (Rogers, 2021). of these (and many other) issues see Ashurst et al. (2020), Hecht et al. (2018), and the NAACL 2021 Ethics FAQ • responsible data use/reuse : whether the collection methodology is sufﬁciently described,…”
Is GPT-3 a Good Data Annotator? (2022, methodology): “Data annotation is the process of labeling data to help machine learning models better understand the purpose of any given task (Liu et al., 2016; Rogers, 2021), recognize patterns and make predictions (Tan et al.”
When Combating Hype, Proceed with Caution (2021, background): “Jin et al. (2021) and Rogers (2021) offer broader discussion of how to ensure that the net impact of near-future NLP deployments on the world is positive.”
SynthBio: A Case Study in Human-AI Collaborative Curation of Text Datasets (2021, unknown): “Rogers [45] summarize the arguments for and against curation that followed after the publication of work by Bender et al.”