Itemset Mining Method and Device

Abstract

The invention discloses a method and a device for item-sets mining, and belongs to the field of data mining. The method comprises: obtaining a self-defined minimum expected support [mu] and a lowest utility proportion [epsilon]; calculating actual expected support expSup and an actual utility value u of item-sets in an uncertainty database D, the item-sets including at least an item; when expSup >= |D| x [mu] and u>= total utility TU x [epsilon], determining the item-sets to be a high-probability and high-utility item-sets, wherein TU represents the sum of utility of all items in the uncertainty database D, and |D| represents transaction total included in the uncertainty database D. The method achieves effects of high utility values of mined item-sets, and high occurrence probability, so that the mined item-sets is ensured to have relatively high practical use values.

Description

technical field [0001] Embodiments of the present invention relate to the technical field of data mining, and in particular to an itemset mining method and device. Background technique [0002] The database usually includes at least one transaction (English: Transaction), and each transaction includes at least one data item (English: item), for example, a transaction about a person record includes name, date of birth, gender, blood type, etc. data item. [0003] In order to discover the association rules between different data items, it is necessary to mine the target data item set. Itemset (English: Itemsets) is a collection of at least one data item, which is used to represent an inherent association rule in the database. HUIM (High-Utility Itemsets Mining, High-Utility Itemsets Mining), as a common data mining method, is used to mine itemsets with high utility values ​​composed of different data items from the database. In the existing HUIM-based algorithm, by calculat...

AI Technical Summary

Problems solved by technology

[0004] In the process of implementing the embodiments of the present invention, the inventors found that the above-mentioned technology has at least the following problems: In actual situations, the data stored in the database is often uncertain data, that is, there is a probability of occurrence of transactions in the database, and each There is a large difference in the probability of occurrence corresponding to the transaction

However, the existing HUIM-based algorithm does not consider the occurrence probability of transactions, and it is easy to mine itemsets with high utility value but low occurrence probability, resulting in low actual use value of the mined itemsets.

Images