DeltaNI: an efficient labeling scheme for versioned hierarchical data

Main-memory database systems are emerging as the new backbone of business applications. Besides flat relational data representations also hierarchical ones are essential for these modern applications; therefore we devise a new indexing and versioning approach for hierarchies that is deeply integrated into the relational kernel. We propose the DeltaNI index as a versioned pendant of the nested intervals (NI) labeling scheme. The index is space- and time-efficient and yields a gapless, fixed-size integer NI labeling for each version while also supporting branching histories. In contrast to a naive NI labeling, it facilitates even complex updates of the tree structure. As many query processing techniques that work on top of the NI labeling have already been proposed, our index can be used as a building block for processing various kinds of queries. We evaluate the performance of the index on large inputs consisting of millions of nodes and thousands of versions. Thereby we show that DeltaNI scales well and can deliver satisfying performance for large business scenarios.