Dynamic Analysis of User Requirements Based on Natural Language Processing
Mengting Song, Ke Zhang, Xuan Wu · 2024
Tracing the dynamic evolution of user requirements based on time dimension and accurately identifying user requirements is crucial for enterprises to innovate their products in a fiercely competitive market in order to pursue sustainable development. Firstly, the Latent Dirichlet Allocation (LDA) topic model is used to extract user requirement themes for target product from online reviews; Secondly, based on natural language processing, conduct importance and sentiment analysis on user requirement themes; Again, conduct dynamic analysis of user requirements based on the IPA-Kano model within the target product time window; Finally, predict the evolution trend of user requirements for the target product based on the Long Short Term Memory (LSTM) model and identify key user requirements. Taking the Redmi K60 smartphone as an example, the research results demonstrate the dynamic evolution panorama of user requirements for the target product, verifying the effectiveness of the method proposed in this paper.