Data mining system

Abstract

The invention discloses a data mining system. The data mining system comprises a data mining terminal which is used for mining data, meeting conditions preset by a user, in a network and is connected into the network through a bus. The data mining terminal further comprises a configuration module, a recognition module and a storage, the configuration module is used for configuring data mining conditions, a data structure and a data type, the recognition module is used for recognizing physical layers, link layers and application protocol layers of the network, and the storage is used for storing the configured data mining conditions and data obtained through mining according to the configured data structure and the data type. The data mining system has the advantages that data mining methods such as a statistic analysis method, a decision-making tree method, a rough set method, a fuzzy set method, a reasoning method based on models and a neural network method are used for analyzing, processing, reasoning and predicting the mass data in an information data set, and finally the optimal scheme is pushed automatically according to the conditions set by the user.

Description

technical field [0001] The invention relates to a data mining system. Background technique [0002] Data mining is to extract potentially useful information and knowledge hidden in it from a large number of incomplete, noisy, fuzzy and random practical application data. process. [0003] Data mining is applied in a wide range of fields, such as banking, telecommunications, insurance, transportation, retail and other commercial fields. Typical business problems that data mining can solve include: Database Marketing, Customer Segmentation & Classification, Profile Analysis, Crossselling and other market analysis behaviors, as well as Churn Analysis. ), customer credit scoring (Credit Scoring), fraud detection (Fraud Detection) and so on. [0004] The data mining process usually includes three main steps: data preprocessing (ETL), data mining algorithm implementation, and result display. Through the ETL step, the source data can be preprocessed to obtain the data to be mine...

AI Technical Summary

Problems solved by technology

[0005] The operation of the existing technology is relatively complicated, the cost is high, the software and hardware are relatively closed, and the dependence on the manufacturer is strong, and the method still uses a serial data mining mechanism, so its performance is still difficult to greatly improve