Research On The Method Of Product Positioning Driven By Big Data

With the advent of the era of big data,the analysis and utilization of big data has become a source of new economic value,which also brings huge benefits to enterprises.In recent years,with the rapid development of e-commerce,online shopping has become the main way for people to consume,so a large amount of online shopping data has been generated;However,how to use these data resources to make product positioning more precise has become an important issue for everyone.This paper studies a product positioning method based on big data driving,and applies potential useful information mined from massive customer shopping data to product positioning.The specific operation is as follows: The first step is to use the web crawler Scrapy to perform data crawling on the product reviews and product related information on the Jingdong Mall platform,and perform noise removal processing to obtain preprocessed data.In the second step,Jieba participle is extracted from the denoised data,and Apriori algorithm is used to mine the product attribute feature words to get the product attributes and attribute weights that need to be studied.According to the product attributes obtained,the product attributes level is crawled and analyzed,and the product attributes level with the next research value is selected.According to the analysis results,the questionnaire was designed to conduct surveys,and the customer's preference data for each attribute level of the product was obtained.In order to solve the ambiguity problem of preference data,this paper uses cloud model to transform the qualitative language evaluation of customer's product attribute level into quantitative data.Then,using the discrete selection model Multinomial Logit Model(MNL),the utility value of the product attribute is calculated;In order to make the customer preference function more in line with the actual product market,this paper introduces the attribute weight and price factor to improve the original customer preference function.Finally,the product positioning optimization model is established with the goal of maximizing product profit.Since the solution process of the product positioning model can be regarded as a product attribute combination optimization problem,the problem is an NP-hard problem.Therefore,this paper uses the artificial bee colony algorithm to solve the optimal combination of product attribute levels.The feasibility and validity of the product positioning method are verified by an example of mobile phone.Among them,the improved method is also given for the shortcomings of the artificial bee colony algorithm,such as easy to fall into local optimum and slow convergencerate in later stage.Four classical test functions were selected for experimental comparison.The experimental results show that the improved artificial bee colony algorithm has certain improvement in convergence speed and accuracy.It can also be verified by solving the product positioning model.