Introducing the Data Measurements Tool: an Interactive Tool for Looking at Datasets

tl;dr: We made a tool you can use online to build, measure, and compare datasets.

Click to access the 🤗 Data Measurements Tool here.

As developers of a fast-growing unified repository for Machine Learning datasets (Lhoest et al. 2021), the 🤗 Hugging Face team has been working on supporting good practices for dataset documentation (McMillan-Major et al., 2021). While static (if evolving) documentation represents a necessary first step in this direction, getting a good sense of what is actually in a dataset requires well-motivated measurements and the ability to interact with it, dynamically visualizing different aspects of interest.

To this end, we introduce an open-source Python library and no-code interface called the 🤗 Data Measurements Tool, using our Dataset and Spaces Hubs paired with the great Streamlit tool. This can be used to help understand, build, curate, and compare datasets.

What is the 🤗 Data Measurements Tool?

The Data Measurements Tool (DMT) is an interactive interface and open-source library that lets dataset creators and users automatically calculate metrics that are meaningful and useful for responsible data development.

Why have we created this tool?

Thoughtful curation and analysis of Machine Learning datasets is often overlooked in AI development. Current norms for “big data” in AI (Luccioni et al., 2021, Dodge et al., 2021) include using data scraped from various websites, with little or no attention paid to concrete measurements of what the different data sources represent, nor the nitty-gritty details of how they may influence what a model learns. Although dataset annotation approaches can help to curate datasets that are more in line with a developer’s goals, the methods for “measuring” different aspects of these datasets are fairly limited (Sambasivan et al., 2021).

When can I use the 🤗 Data Measurements Tool?

What can I learn using the 🤗 Data Measurements Tool?

Dataset Basics

For a high-level overview of the dataset