Using structured changes for elucidating data evolution

In this paper we argue that changes should be treated as first class citizens in data management systems. In our approach, changes are not just transformation operations but complex objects retaining structural, semantic, and temporal characteristics. We believe that accommodating structured changes in information modeling and querying can provide users with new insights into data lifecycle. In previous work we proposed a graph model called evo-graph for capturing in a coherent way the relationships between evolving data and changes applied on them. We also presented evo-path, a path expression language for evo-graphs based on XPath. In the present paper we define an XML representation of evo-graph, and discuss the use of XQuery for expressing a number of interesting query categories. We demonstrate the feasibility and usefulness of our approach through an example inspired by bioscientific databanks.