The pipeline activities will make use of the inputs and outputs that are contained in the dataset, which contains those activities. A connected data store can be any kind of file, folder, document, or anything else imaginable; datasets frequently represent the organization of information within such a store. An Azure blob dataset, for example, details the blob storage folder and container from which a particular pipeline activity must read data to continue processing. This information is used to determine where the data will be read from.
