Application of the massive data precision classification in e-commerce based on big data

Abstract

Accurate classification of massive data in the field of electronic commerce, can service consumers to the greatest degree. The variety and quantity of goods fast growth make a large number of useful products information would be submerged in massive data. The traditional electronic commerce data classification system applied in the user data validation, is affected by massive information, to make data validation process time-consuming and low efficiency. The massive data precision classification system of electronic commerce based on big data is proposed. Large amounts of data is made big data relational decision-making calculation, to get the relevance between all data. Based on the above data relevance, the relational decision-making tree is established, to obtain user’s interest data recommended goal. The experimental results show that, by using the improved algorithm for the user massive interest data classification in electronic commerce, can effectively reduce the time to confirm the user interest commodity information data, ensure data quality to meet the requirement of electronic commerce. Introduction With the rapid increase of data classification technology, the technology has been widely used in database management in various industries, playing a more and more important role [1]. In the field of electronic commerce, with the rapid increase of the number of and types of goods and commodities, lots of useful information will be submerged in the vast amount of data, therefore, how to make efficient commodity recommendation based on user interest [2], has become the core problem of classification recommendation system needed to study in electronic commerce [3]. In the condition of massive commodity information data, user interest data classification recommendation system, become the [4] key subject to study in the field of electronic commerce. Currently, the precision massive data classification system in electronic commerce mainly includes the system based on the improved support vector machine algorithm [5], the Gauss model and fuzzy clustering algorithm. One of the most commonly used is the accurate massive data classification system in electronic commerce field based on improved support vector machine algorithm. Due to the accurate massive data classification system in electronic commerce field plays an irreplaceable role in the commercial areas, it has a broad prospect for development, and has become the focus of many experts research topic. Principle analysis of accurate massive data classification in electronic commerce field The accurate classification of massive data in the field of electronic commerce, can service consumers to the greatest degree. Usually, the distribution of commodity information data in electronic commerce field follows normal distribution, which can be used 1 2 { , , , } q V v v v =  to describe. Using the following formula can calculate the transformation parameters: 2 (0, ) Z Y N δ = + (1) The above commodity data are made wavelet transform processing, the data are still subject to the normal distribution, which can use formula (2) to describe: 2 2 (0, ) (0, ) Z Xz Xy N Y N σ σ = = + = + (2) The probability of the commodity information data subjected to normal distribution in the interval 1 2 { , , , } p K k k k =  is 0.9991, therefore, it can be according to the wavelet transform parameters International Conference on Automation, Mechanical Control and Computational Engineering (AMCCE 2015) © 2015. The authors Published by Atlantis Press 1227 to divide the massive commodity information data in electronic commerce into two different parts. If 2 1 1 ( , ) ( , ) q e p p l kl k l k l L Y B y f z b = = = ∑∑ , the degree of importance of the data is relatively high, on the contrary, its importance is relatively low. Wherein, , , , 1 e q p d = is the wavelet transform parameter which includes data. Use the following formula can deal with wavelet transform: