What is Spatiotemporal data mining?

Spatiotemporal data mining define the process of finding patterns and knowledge from spatiotemporal data. An instances of spatiotemporal data mining contains finding the developmental history of cities and lands, uncovering weather designs, forecasting earthquakes and hurricanes, and deciding global warming trends.

Spatiotemporal data mining has become important and has far-extending implications, given the recognition of mobile phones, GPS devices, Internet-based map services, weather services, and digital Earth, and satellite, RFID, sensor, wireless, and video technologies.

There are several types of spatiotemporal data, moving-object data are important. For instance, animal scientists connect telemetry machinery on wildlife to explore ecological behavior, mobility managers embed GPS in cars to superior monitor and guide vehicles, and meteorologists need weather satellites and radars to find hurricanes.

An instances of moving-object data mining such as mining movement designs of multiple moving objects (i.e., the finding of relationships between multiple moving objects including moving clusters, leaders and followers, merge, convoy, swarm, and pincer, and different collective movement patterns).

A cyber-physical system (CPS) generally includes a huge number of interacting physical and data components. CPS systems can be interconnected so as to form huge heterogeneous cyber-physical networks.

An instance of cyber-physical networks contain a patient care system that connect a patient monitoring system with a web of patient data and an emergency managing system.

A transportation system that connect a transportation monitoring web, including multiple sensors and video cameras, with a traffic data and control system; and a battlefield commander system that connect a sensor network with a battlefield data analysis system.

Data generated in cyber-physical systems are powerful, explosive, noisy, inconsistent, and interdependent, including rich spatiotemporal data, and they are essential for real-time decision making.

Multimedia data mining is the finding of interesting designs from multimedia databases that save and manage huge set of multimedia objects, such as image data, video data, audio data, and sequence data and hypertext data including text, text markups, and linkages.

Multimedia data mining is an interdisciplinary area that merge image processing and perceptive, computer vision, data mining, and pattern identification. There are several problems in multimedia data mining involve content-based retrieval and similarity search, and generalization and multidimensional analysis. Multimedia data cubes include more dimensions and measures for multimedia data.

Text mining is an interdisciplinary application that draws on data retrieval, data mining, machine learning, statistics, and calculation linguistics. A big portion of data is saved as text including news articles, high-tech papers, books, digital libraries, email messages, blogs, and web pages.