Big Data is partly the problem of large, uniform data sets. For example, matrices with millions of rows. But it is also a problem of non-uniform data, e.g. graph-based data, with sparse content. Is the linking or larger data sets with each other to create "big data" also within the scope of this StackExchange?