Design of Computational Identification Model and Optimization Strategy for Consumer Demand Based on Multimodal Data Fusion

This paper draws a framework for constructing user demand modal information, uses crawler technology to obtain online review text information, processes the text information, and mines relevant consumer demand information.The LDA topic model is used to extract the topics of consumer concern from the online comments, identify the topics of consumer demand and clarify the concern degree of each demand.The KANO model is proposed to establish a consumer demand classification method based on the KANO model by combining product characteristic attributes and consumer demand information.Examine the theme discrimination performance of the LDA model on the hotel category, footwear category, and food category datasets.Combine the preprocessed user demand data to statistically quantify user demand for quantitative Kano transformation.Classify user demands into Kano categories and calculate the priority order of user demands to get the product optimization strategy.The weighted order of consumers' demands for automobiles is footrest, cigarette lighter, antenna, window, low beam, key, etc. in order.It can be found that automobile consumers pay more attention to the needs of antenna, cigarette lighter, pedals, and enhancement of accessory functions.As a result, automobile manufacturers should increase the seat comfort, improve power, enhance the flexibility of shifting such aspects of the whole vehicle handling experience, in addition to improving the lights, keys and other car quality related needs.