Novel data structures and algorithms are needed to scale up code searches on systems of systems. In information retrieval, repositories consist of documents, which are treated as bags of words. In database management systems, databases are organized into tables of rows, with keys and attributes. But source code also has a structure, as given by its syntax. Characteristics of all three must be take into consideration to provide fast and efficient searching.
Oliver Hummel, Colin Atkinson & Marcus Schumacher