Research On The Consumer Preference Prediction System In Neuromarketing

It is a hope and desperate task for all companies to understand exactly what consumers really want and what consumers are going to do.So many companies invest heavily in consumer surveys to understand how consumers response to products or advertisements.Traditional research techniques such as focus group interviews and surveys,however,often fail to deliver satisfactory performance relative to the cost and time the firm has invested.Consumers tend to not tell the truth about surveys or interview surveys,and these surveys have limitations because they are designed to answer only what consumers perceive themselves.In fact,in the consumer survey,not only the content expressed by the consumer is a small part,but also the content hidden or not expressed by the consumer is only a small part.Most of them are unconscious that even consumers do not know.Recently,neuromarketing has attracted attention as a more scientific research technique to overcome the limit of traditional consumer research.Neuromarketing,as a compound word of a neuro meaning neuroscience and a marketing,explores consumer psychology and behavioral mechanisms by using neuroscience technology to measure the brain response of consumers,and applies it to marketing.With the emergence of brain measurement devices that do not damage the human body,neuromarketing research is being activated and research results are being introduced into the marketing activities of companies.The purpose of this research is to predict and analyze consumer preference by constructing a consumer preference prediction system in neuromarketing based on electroencephalograph(EEG).Consumer preference researches allow companies to analyze consumer responses to marketing activities such as products,advertisements and services.From point of view of a company,consumer preference research can analyze the consumer response to products,advertising,and services in advance to reduce the risk of market uncertainty,and use the analysis results to drive corporate strategy.From the importance of consumer preference research in marketing,this thesis has designed a consumer preference prediction system that can answer "What do consumers prefer unconsciously?".The details of this thesis are as follows.First,based on the theoretical analysis of the neuromarketing research trend and the EEG based neuromarketing,the consumer preference prediction system in EEG based neuromarketing was proposed.Based on the analysis of neuromarketing techniques,recent neuromarketing researches on consumer purchase decision,advertising,consumer brand attitude,consumer promotion strategy and consumer characteristics were analyzed comprehensively.In addition,neuromarketing based on EEG technique which has the characteristics of low cost,easy operation and wireless portability was selected in the neuromarketing techniques,and a more detailed review was given.From this,a new consumer preference prediction system consisting of consumer EEG preprocessing and consumer preference classifier or predictor was proposed.Next,as consumer EEG preprocessing methods,the method of removing ocular artifacts in EEG signals and the EEG topographic video method were proposed.In order to remove ocular artifacts which distort consumer EEG signals,the new method combining independent component analysis and adaptive filter was proposed.In addition,the consumer EEG topographic video which can reflect the time-frequency-space characteristics of consumer EEG signals was newly defined,and the method of calculating it using the short time Fourier transform and the biharmonic spline interpolation was proposed.Then,based on the analysis of consumer EEG topographic video,a classification prediction model was designed,which is more suitable and more robust for consumer preference prediction.Based on the analysis of the non-stationary and nonlinear characteristics of the EEG topographic video,a new consumer preference prediction model combining the convolutional neural networks and the long short term memory neural networks,was proposed.In the prediction model,in particular,the mask layer for consumer EEG topographic video processing was newly defined and inserted between the convolution layer and the pooling layer of the convolutional neural network.It was also theoretically discussed that the proposed model,consisting of the feature learning stage and the prediction stage,is suitable and robust to the processing of EEG topographic video which reflects time-frequency-spatial information of consumer brain activity.Finally,performance evaluation and application emulation of the proposed consumer preference prediction system were carried out.The system performance was evaluated in the consumer dependent mode and the consumer independent mode by determining the structure and parameters of the prediction model.The performance evaluation results showed that the proposed consumer preference prediction system is stable,robust and applicable to practice,in both consumer dependent mode and consumer independent mode.In addition,the application of proposed system for marketing such as product,advertising,brand and product design was emulated.In the application emulation process,consumer preference for product,advertising,brand and product design was analyzed,and results useful for marketing activities were derived.This research focuses on constructing a system for approaching practical use of neuromarketing.This research uses EEG as a neuromarketing technique.This is the cheaper technique among the neuromarketing techniques to date,and requires little additional cost for the neuromarketing operations.Besides,it is possible to measure the EEG of consumers in actual marketing environment.This research suggests methods of preprocessing of EEG signals and predicting of consumer preference.The consumer EEG preprocessing method proposed by considering the non-linear and non-stationary of the consumer EEG signals can improve the performance of the consumer preference prediction system.Also,the consumer preference prediction method is an adaptive method more suitable for the time-frequency-space characteristics of EEG signals of consumers.Consumer preference prediction system to be proposed in this thesis can be effectively applied to neuromarketing research as well as actual marketing environment.As neuromarketing research requires a lot of cost,application studies of neuromarketing have been mainly led by large companies to date.The system to be proposed in this thesis does not require much cost,so it can be easily introduced to small and medium companies.In conclusion,the results of this research can be applied more effectively,and used for the research and commercialization of neuromarketing at lower cost.Today,business is growing by consumers.Therefore,consumer preference analysis is a very important issue in marketing strategy research and marketing management.The proposed consumer preference prediction system and its application method can be used efficiently in the consumer preference analysis for the marketing strategy research and development and the marketing management.And the proposed system is based on low cost and low maintenance EEG technique,so it can be applied to large company as well as small and medium company,at low cost.In addition,consumer EEG signals can be measured wirelessly so that the preference of consumers can be analyzed by approaching to the marketing environment rather than experimental environment.Furthermore,it is also applicable to neuromarketing research in which consumer preference analysis is a fundamental problem.