Temporal Data Mining

Spatiotemporal Data Mining     ◾     325

region and the temporal interval of the query can be of any size. In the pro-

posed methods, at the nest aggregation level, there are regions associated

with measures, such as the number of vehicles or the number of visitors

per day. e goal of the authors is to answer the spatiotemporal window

aggregate query, which returns an aggregate measure, such as sum, over

a specied rectangle and a specied time interval. e authors propose a

number of multitree indices that support the following:

Ad hoc groupings•

Arbitrary query windows•

Historical time intervals•

e authors propose two indexes: (1) a host