A Data Mining Method

Abstract

The invention discloses a data mining method and solves the problems that network information is high in repetition degree and much in redundant information. The data mining method is high in data mining speed, good in recall ratio and ideal in effect. The data mining method comprises the steps of fixed point collection, automatic discovery and data analysis and storage. In the fixed point collection, websites in industries are preset and utilized as data sources, reliability weight values are set for the data sources, a data collecting mode is set according to the data sources, and data are mined from the data sources regularly or irregularly. In automatic discovery, a network probe is set and can automatically find out websites with high similarity, the websites with the high similarity are utilized as collecting point websites, the collecting point websites are added into a collecting point website base, reliability weight values are set for the collecting point websites, a data extracting mode is set according to the collecting point websites, and data are mined from the data sources regularly or irregularly. In data analysis and storage, the mined data are encoded in a unified mode, repetition information is removed, data are screened, clustering analysis is carried out on the screened data, the information amount of the same topic can be calculated out, the topic attention weight is labeled, data are stored, and indexes are established.

Description

technical field [0001] The invention relates to the technical field of data mining, in particular to a data mining method. Background technique [0002] Today's society has entered the era of high-speed information dissemination. While this brings convenience to people, more and more problems have emerged. For example, the search results of existing search engines are too repetitive and there are many redundant information that do not meet expectations. , long search time, low efficiency, etc. [0003] Due to the high rate of information reprinting on the Internet at present, search engines such as Baidu and Google take a long time to search for the recall rate of search, and the search results are very repetitive, which is not conducive to users to quickly find valuable content. In addition, some industry search engines only aim at industry websites, which improves the search efficiency, but the recall rate is low and it is easy to cause omissions. [0004] The current co...

AI Technical Summary

Problems solved by technology

[0002] Today's society has entered the era of high-speed information dissemination. While this brings convenience to people, more and more problems have emerged. For example, the search results of existing search engines are too repetitive and there are many redundant information that do not meet expectations. , long search time, low efficiency, etc.

[0003] Due to the high rate of information reprinting on the Internet, Baidu, Google and other search engines take a long time to search for the recall rate of search, and the search results are very repetitive, which is not conducive to users to quickly find valuable content.

In addition, some industry search engines only aim at industry websites, which improves the search efficiency, but the recall rate is low and it is easy to cause omissions

[0004] The current commercial competition largely determines the extent to which enterprises have mastered the latest information. In other words, the update and analysis of industry information by enterprises determines the potential of enterprises. Enterprises are often unable to afford independent information search consumption. On the other hand, enterprise-customized search engines often only search industry websites, and do not catalog the entire Internet, which is likely to cause information omission

Images