Techniques and experience in mining remotely sensed satellite data

The paper presents a set of requirements for a data mining system for minin g remotely sensed satellite data based on a number of taxonomies that chara cterize mining of such data. The first of these taxonomies is based on know ledge of the mining objectives and mining algorithms. The second is based o n various relationships that are found in data, including those between dif ferent types of data, different spatial locations of the data and different times of data capture. The paper then describes the ADaM data mining syste m, which was developed to address these requirements. The paper describes s everal data mining techniques that have been applied to remotely sensed dat a. The first type is target independent mining, which mines data for transi ents and trends, with mined results representing a highly concentrated form of the original data. The second type is the milling of vectors (represent ing multi-spectral or fused data) for association rules representing relati onships between the various types of data represented by the elements of th e vector. The third type mines data for association rules that characterize the texture of the data.