Consumer intention recognition method based on user forgetting curve

Abstract

The invention discloses a consumer intention recognition method based on a user forgetting curve, which is characterized in that each user has an ID to store corresponding user data, wherein the ID stores intention network information of the corresponding user, a user input a statement, keywords are extracted based on an LDA technology, semantic analysis is performed through a Word2vec tool, the content is mapped to the network so as to form description for the user; the correlation between a keyword set K extracted at an initial moment t0 and a set P of commodities talked about by a consumer is calculated through the Word2vec so as to acquire an initial intention intensity value S0 of the consumer, then the initial intention intensity value is substituted into a user forgetting curve based intention intensity model to calculate an intention intensity value Sn at a moment tn, a customer service robot accurately recognize commodity information which the consumer wants to acquire according to the intention intensity value Sn of the consumer for different commodities. The consumer intention recognition method effectively avoids a circumstance of irrelevant answers of the online customer service robot, and is conducive to popularization of online shopping.

Description

technical field [0001] The invention relates to a consumer intention recognition method, in particular to a consumer intention recognition method based on a user forgetting curve. Background technique [0002] With the development of the Internet, online shopping has quietly risen, and it has become more and more fashionable for people to buy various items through the Internet, which directly leads to a large demand for online customer service personnel. However, traditional online customer service robots often do not have a clear understanding of consumers' intentions, and answer questions that are not relevant, which will hinder consumers' acquisition of product information and is not conducive to the development of online shopping. The main reason why the existing customer service robots do not understand the consumer's intention is because the existing system does not refer to the time parameter and the forgetting parameter for the recognition of the consumer's intention...

AI Technical Summary

Problems solved by technology

[0002] With the development of the Internet, online shopping has quietly risen, and it has become more and more fashionable for people to buy various items through the Internet, which directly leads to a large demand for online customer service personnel. However, traditional online customer service robots often do not understand consumers' intentions clearly, and answer questions that are not relevant, which will hinder consumers from obtaining product information and is not conducive to the development of online shopping

The main reason why the existing customer service robots do not understand the consumer's intention is because the existing system does not refer to the time parameter and the forgetting parameter for the recognition of the consumer's intention, because when the consumer inquires about the product, he often asks for multiple products, and Often after asking a question, the next question will be asked every few hours, or even a few days or longer. At this time, the robot's recognition of the consumer's purchase intention will be confusing, and the answer will not be what was asked.

Images