Research On Perceptual Demand Analysis And Its Application For Online Consumers

In order to deal with the problem of "information overloaded", E-commerce recommender system has been the most valuable technique following traditional information retrieval technique. But with the development of social productivity and economy, consumers’ consumption concept has changed, transferring from material consumption to emotional consumption, and new products are emerging constantly, resulting in that traditional recommender systems cannot meet consumers’ increasing personalized demands. Based on the theory of consumer psychology, consumer behavior is determined by consumer psychology while consumer psychology can reflect consumer behavior. In accordance with the causality, as a concept of consumers’deep-level psychological need, perceptual demand is introduced into our study. By analyzing historical records, consumers’personalized preference can be obtained in the form of perceptual demand and be utilized to predict consumer’s preference for new products.Focusing on garments in E-commerce, main content of our study includes the following two aspects.On one hand, the consumer’s personalized perceptual demand model is built to deeply describe the preference, solving the problem that preference information is limited in behavior rules or product attributes in traditional recommender systems. Firstly, expert knowledge is used to establish the mapping relations between attributes and perceptual features in the example of garments. Then personalized perceptual demand is analyzed with indexes of product attribute importance, perceptual feature preference and perceptual feature weight. Finally, results of examples are verified from the perspective of stability, accuracy and universality.One the other hand, to address the issues of low accuracy caused by ignorance of psychological factor in traditional recommender systems and consumers’ stronger perceptual demand for products, a method to recommend new products is proposed based on perceptual demand model. By calculating the similarity between products and perceptual demand, the consumer’s preference can be predicted and homogenized products can be distinguished and recommended. Lastly from the perspective of accuracy, experiment results confirm the rationality of our proposed methodThis paper describes consumer’s preference deeply into the psychological level. From the point of perceptual demand, personalized preference behind purchasing behavior is analyzed and used to produce recommendation results. It offers guidance for the analysis of online consumer psychology and provides a new idea to recommend on basis of psychological factor.