Frequent pattern mining method for semi-structured data based on sfp-link

Abstract

The invention discloses a SFp-Link-based semi-structured data frequent pattern mining method. The method comprises the following steps of: establishing a semi-structured data frequent pattern linked list SFp-Link for semi-structured data; carrying out frequent pattern mining on the basis of the semi-structured data frequent pattern linked list SFp-Link, so as to effectively extract a frequent itemset in the semi-structured data according to a mining aim. According to the method, when the semi-structured data frequent pattern linked list SFp-Link is established, scanning is only carried out onthe mined sample data for one time, only sample item sets, project combinations in which are scanned for the first time, are stored, and for the sample item sets, the project combinations in which are scanned again, only corresponding sample frequencies are accumulated for one time, so that the method has the advantages of being small in consumed storage space, short in consumed mining time and high in mining efficiency.

Description

technical field [0001] The invention relates to a method for mining frequent patterns of semi-structured data based on SFp-Link, belonging to the technical field of data mining. Background technique [0002] In the medical field, a large amount of diagnostic data is unstructured or semi-structured data, and there are many types of attributes, often as many as hundreds or even thousands of items (such as genes), but the sample size is very small. If such semi-structured data is converted into structured data, it will inevitably generate a very sparse matrix. However, since some rare diseases need more research, the internal correlation between these rare information cannot be ignored, and some effective technology is needed to extract relevant information. Existing algorithms such as Apriori, FP-tree and other algorithms are mostly aimed at structured data, and cannot effectively solve the problem of extracting frequent patterns, associations and correlations of unstructured...

AI Technical Summary

Problems solved by technology

However, since some rare diseases require more research, the inherent correlation between these rare information cannot be ignored, and some effective technology is needed to extract relevant information

Existing algorithms such as Apriori, FP-tree and other algorithms are mostly aimed at structured data, and cannot effectively solve the problem of extracting frequent patterns, associations and correlations of unstructured or semi-structured data.

In addition, this type of algorithm also needs to read the sample database multiple times, often resulting in high space and time complexity.

Images