Good news for researchers and academics around the world as Google’s Dataset Search is out of beta mode and will now roll out in full version serving up, up to 25 million datasets.
Google launched Dataset Search in 2018 as part of the company’s goal to put an end to the fragmentation of open-access data.
Many universities and academics publish their work online but it’s difficult to search that niche of information on the normal search, they often don’t show up but by adding open source metadata tags to their web pages, these groups can have their data indexed by Google’s Dataset Search.
No data or specifics were revealed but the company did admit that “hundreds of thousands of users” of academics have responded positively to the tool since its launch.
The Verge spoke with a research scientist at Google AI who helped create the tool named Natasha Noy who said that “most [data] repositories have been very responsive” and that Dataset Search has even encouraged older scientific institutions to take “publishing metadata more seriously”.
A few upgrades to the beta version includes the ability to filter data by type (tables, images, text, etc). It also details whether it’s free to use and in what geographic area. The tool is now available on mobile with expanded dataset descriptions.
To reiterate, Google has confirmed that the tool’s search engine covers almost 25m datasets, though this is only a “fraction of datasets on the web”. The largest topics indexed by Dataset Search include geosciences, biology, and agriculture with education, weather, cancer, crime, soccer, and dogs being the most common queries.
The company plans to continue to expand the scope and volume of datasets available for search queries.