Data mining method for big data processing

Abstract

The invention relates to the technical field of internet, in particular to a data mining method for big data processing, which is comprehensive in mining result and high in data processing speed. The data mining method comprises the following steps of acquiring a retrieval demand of a user, processing the retrieval demand to be consistent data, then performing correlation matching on the consistent data and pre-stored purchased words, acquiring at least one correlation data source between the pre-stored purchased words and the retrieval demand data, building a topological graph of network, and mining in a database based on the topological graph of network. Compared with the prior art, the simple word matching or semantic mining way is abandoned, and the mining for potential key information is carried out on the data to be analyzed through starting with the correlation topological network, so that the data mining method for big data processing has the remarkable advantages that the acquired results are more comprehensive and more accurate, and the like.

Description

Technical field: [0001] The invention relates to the technical field of the Internet, in particular to a data mining method for big data processing with comprehensive mining results and high data processing speed. Background technique: [0002] Big data technology, or huge amount of data, refers to the amount of data involved is so large that it cannot be obtained, managed, processed and organized within a reasonable period of time through current mainstream software tools to help enterprises make more positive business decisions. The strategic significance of big data counting is not to master huge data information, but to professionally process these meaningful data. In other words, if big data is compared to an industry, then the key to achieving profitability in this industry is to The "processing capability" of data is to realize value-added by processing data. [0003] The process of how to extract hidden, unknown, but sometimes potentially useful information from a l...

AI Technical Summary

Problems solved by technology

At present, the common data mining methods can be roughly divided into the following types: one is to establish a webpage classification system in a semi-automatic way, and introduce attributes such as data classification, query word classification, or purchase classification, and combine webpage search results for correlation feedback, so as to obtain The desired information; the other is based on literal text matching; the other is to use semantics-based analysis of latent semantic correlation models to obtain retrieval data. The above analysis systems tend to miss potential features in information services, resulting in data analysis results Incomplete

Images