User Requirement Identification and Dynamic Analysis Based on Improved Kano Model
Ruijie Zhang,Yanan Qiao,Keqin Wang,Junle Zhou,Kailong Zhang,Weili Xia
DOI: https://doi.org/10.1142/s0218126624503134
2024-08-21
Journal of Circuits Systems and Computers
Abstract:Journal of Circuits, Systems and Computers, Ahead of Print. The widespread adoption of e-commerce has rendered online reviews an indispensable source for acquiring knowledge about product features. The Kano model serves as a classic framework for categorizing customer requirements. It was proposed by the Japanese scholar Noriaki Kano in 1984 as a tool primarily used to help businesses better understand and categorize customer needs, guiding product development and service quality improvement. However, existing research on mining online review information often relies predominantly on static analysis, lacking comprehensive categorization of user demands. To address this gap, this paper introduces an enhanced Kano model for the identification and dynamic analysis of user demands. This model classifies review data based on emotional polarity and user attention, leveraging the characteristics inherent in online reviews. Furthermore, it constructs a user demand evolution model using time series analysis and utilizes SnowNLP for emotional calculations. By conducting word frequency statistics and employing LDA models to analyze product features, this paper identifies the evolving trends in user demands. Additionally, the exponential smoothing method is used to forecast the trend in users' emotional values towards the product. The findings demonstrate that this model effectively extracts valuable insights regarding product feature information from user reviews, thereby offering novel perspectives and methodologies for related fields. Consequently, this contributes significantly to advancing the identification and dynamic analysis of user demands.
engineering, electrical & electronic,computer science, hardware & architecture