Abstract: 【Purpose/significance】Currently, more and more consumers use online reviews for information interaction and emotional
expression. Identifying and analyzing user needs from a wealth of online product reviews can help companies improve the quality of products and services in a targeted manner, thereby promoting the sustainable development of the company.【Method/process】This pa⁃per uses the LDA model to mine online mobile phone reviews for review topics and product features to effectively identify user needs elements. Based on Kano model, the user demand questionnaire is set up, and the influence of each demand on the user satisfaction is analyzed by combining with the user satisfaction index, and the importance of various user demand and the priority of supply are deter⁃ mined.【Result/conclusion】In this study, 24 user demand elements are divided into 6 high charismatic needs, 8 low charismatic de⁃mands, 3 high expectation demands, 3 high essential demands, 2 low essential requirements, and 2 no difference requirements, and further puts forward the optimization strategy of enterprise product management.【Innovation/limitation】This study uses text mining methods to analyze user needs for real online reviews, and effectively overcomes the problems of lagging demand sources and insuffi⁃
cient reliability in traditional user demand survey methods. In addition, the brands of the products selected in this paper are the same,and follow-up research can be improved and deepened to the analysis of product demand on multiple platforms and brands.