Google's Dataset Search Exits Beta With New Mobile Features
Google officially launches Dataset Search out of beta, giving researchers access to 25 million publicly available data sets. The updated tool introduces a mobile-friendly interface and new filtering options to help users find specific data types.
Google announces that Dataset Search is officially out of beta, providing researchers with access to close to 25 million publicly available data sets. The tool, which first launched in September 2018, allows users to find everything from small demographic tables to large annotated image and audio collections used for training machine learning models. Currently, the search engine indexes about 6 million tables and includes approximately 2 million data sets from U.S. government agencies.
With this official release, Dataset Search introduces a mobile version to help researchers find information on the go. Google also adds a new filtering feature that enables users to sort results by specific data types, such as tables, images, or text. Additionally, the updated service displays more detailed information about the data sets and the organizations that publish them.
The search index pulls data from a variety of public and private organizations, including government agencies and Google's own Kaggle platform. Anybody who owns an interesting data set can ensure it appears in Dataset Search by using standard schema.org markup to describe their data in detail. This open approach helps Google continuously expand the variety of data available to the research community.