A source is the raw data you want to use to create a model, where each row represents an instance of your data and each column represents a field, also called feature; usually the first row of the file contains the field names. Whereas a dataset is a structured version of your source, ready to be used for your Machine Learning tasks. BigML has a large number of options and visualizations that allow you to analyze your data and get the best out of it to build your Machine Learning models. Please take a look at BigML's Sources and Datasets documentation and find out what’s in it for you!