Indexing and Querying A Large Database of Typed Intervals

Assume that a database stores a set of intervals associated with types and weights. Typed intervals enrich the data representation and support applications involving different kinds of intervals. Given a query time and type, the system reports k intervals that intersect the time, contain the type and have the largest weight. We develop a new structure to manage typed intervals based on the standard interval tree and propose efficient query algorithms. Experiments with synthetic datasets are conducted to verify the performance advantage of our solution over alternative methods.