Commodity sales volume prediction algorithm applied to electronic commerce

Abstract

The invention discloses a commodity sales volume prediction algorithm applied to electronic commerce. After original transaction data is obtained, two social influences, i.e., an influence generated through mutual actions of consumers in the same kind of commodites and an influence generated through mutual actions of consumers between different commodities, are defined and calculated. After the two social influences are constructed, firstly, a commodity sales volume is divided into two parts, i.e., a main part of the sales volume and a noise part of the sales volume; secondary, the two social influences are used to respectively predict the main part and noise part of the sales volume; and, finally, two sales volume predictions are integrated into one unified prediction model, and optimization solving is performed. The algorithm devotes the influences generated through the mutual actions of the consumers to interpret a commodity sales volume change, and adapts a commodity sales volume prediction need in real electronic commerce environment well.

Description

technical field [0001] The invention belongs to the technical field of electronic commerce, and in particular relates to a commodity sales prediction algorithm applied in electronic commerce based on social influence discovery. Background technique [0002] With the development of e-commerce, the transaction form of commodities has undergone earth-shaking changes. On this new type of virtual trading platform, consumers quickly and widely browse, purchase, and evaluate various cheap and high-quality commodities, and commodities are also undergoing product innovation, marketing expansion, and logistics improvement. Therefore, the relationship between consumers and commodities is no longer a simple one-to-one direct buying and selling relationship, but a social network between consumers and consumers. The transaction network between consumers and commodities constitutes the most important material in e-commerce. carrier. [0003] Therefore, the traditional economic problem of...

AI Technical Summary

Problems solved by technology

The currently developed forecasting models involving consumers’ social influence have the following problems: (1) The current commodity sales forecasting model assumes that there is only one type of relationship between consumers, but in reality, the relationship between consumers (2) The current commodity sales forecasting model directly takes the commodity sales as the forecast target, ignoring that the change of commodity sales is affected by many factors, and its internal components can be diversified

Images