Stream Mining, Spatial

Synonyms

Stream data mining; Spatio-temporal data mining

Definition

Spatial stream mining is the process of discovering novel patterns, rules, and trends within a set of spatial streams. A spatial data stream is characterized as a data stream which possesses both spatial and non-spatial attributes. Examples of spatial data streams are the location of a continuously moving vehicle in a time period and the non-spatial measurements of a geospatially aware sensor network. In general, the data stream is "continuous, mutable, ordered, fast, high‐dimensional, and unbounded" [2]. Spatial stream mining hence focuses on developing and optimizing mining techniques for spatial data streams.

Historical Background

Active work in stream query processing began in early 2000 with systems such as STREAM [1] and COUGAR [16]. Most of these efforts placed heavy emphasis on query processing and gave little attention to data mining tasks. Then later much interest arose to answer