1.QUEST

QUEST is a multi-task data mining system developed by IBM's Almaden Research Center. It aims to provide an efficient data mining basis for the application development of a new generation of decision support systems. member. The system has the following characteristics:

Provides functions specifically for various mining on large databases: association rule discovery, sequence pattern discovery, time series clustering, decision tree classification, incremental active mining, etc.

Various mining algorithms have approximately linear (O(n)) computational complexity and can be applied to databases of any size.

The algorithm is comprehensive, that is, it can find all patterns that satisfy the specified type.

Corresponding parallel algorithms are designed for various discovery functions.

2.MineSet

MineSet is a multi-task data mining system jointly developed by SGI and Stanford University in the United States. MineSet integrates a variety of data mining algorithms and visualization tools to help users discover and understand the knowledge behind large amounts of data intuitively and in real time. MineSet has the following characteristics:

MineSet is famous for its advanced visual display method.

Provides a variety of extraction methods: 0 ǚ throws howl 錌 ⒒ Yu Bin J Jian ⒐ Qi Huan Jie ⒕ impeachment beak ⑴ unloading narrow brother plaque take?br>

Supports a variety of relational database. Data can be read directly from Oracle, Informix, and Sybase tables, or queries can be executed through SQL commands.

A variety of data conversion functions. Before mining, MineSet can remove unnecessary data items, collect statistics, group data, convert data types, construct expressions to generate new data items from existing data items, sample data, etc.

It is easy to operate, supports international characters, and can be published directly to the Web.

3.DBMiner

DBMiner is a multi-task data mining system developed by Simon Fraser University in Canada. Its predecessor is DBLearn. The system is designed to integrate relational databases and data mining to discover various knowledge based on attribute-oriented multi-level concepts. The DBMiner system has the following characteristics:

It can discover a variety of knowledge: generalization rules, characteristic rules, association rules, classification rules, evolutionary knowledge, deviation knowledge, etc.

Integrate a variety of data mining technologies: attribute-oriented induction, statistical analysis, step-by-step in-depth discovery of multi-level rules, meta-rule guided discovery and other methods.

An interactive SQL-like language - Data Mining Query Language DMQL is proposed.

Can be smoothly integrated with relational databases.

Implemented Unix and PC (Windows/NT) versions of the system based on client/server architecture.