ResearchDataWiki is a free and open database about the different types of data produced by researchers in the sciences and social sciences.
It presents current best practice for sharing each type of research data; this best practice includes both file format and recommended repositories.
Have a look at the different Data Types for which we've created guidance.
Our philosophy for deciding what constitutes a data type is pragmatic: it’s a separate type if it needs its own 'best practice' advice. We applied this heuristic to datasets in 2000 articles from PLOS ONE, Scientific Reports, Nature Communications and some BMC journals, and decided that c. 70% were a single type, 'Tabular data'. We identified about 100 data types in total; most of these relate to the output from sophisticated machines, such as high throughput DNA sequencers or gas chromatographs.
This wiki is also the source for the advice presented in DataSeer, a AI-powered tool that helps researchers identify the datasets associated with their article and guides them through the data sharing process.
N.B. The DataSeer algorithm will be trained to recognize all of the Data Types in this wiki, so users should avoid adding extra Data Types unless absolutely necessary.