Key points are not available for this paper at this time.
In a highly competitive and frequently changing market environment, businesses need to precisely capture and dynamically track user demands in order to optimize products and maintain a competitive advantage. However, as product homogenization increases and replacement rates rise, the focus of user demands has extended from the overall experience to more fine-grained product features, and indicating a dynamic evolutionary trend. This research proposes a hybrid framework for identifying product improvement opportunities and optimization directions. First, we develop Multi-Level User Demands Extraction Model (MDEM) to automatically mine multi-level user demand features. Second, we introduce the Sentiment-Intensity-based Satisfaction Evaluation (SI-SE) and the Attention Measurement (AM) to model and assess user demands from satisfaction and attention perspectives. Finally, we propose Dynamic Opportunity Algorithm (DOA), which incorporates temporal trends of satisfaction and attention to identify high-value product improvement opportunities. We implement and evaluate the proposed approach on vehicle products as a case study. The findings demonstrate that this method can effectively reveal the multi-level structure of user demands, accurately capture the dynamic change characteristics of demands, and thus effectively identify key product improvement opportunities, assisting businesses in developing specific and feasible product optimization strategies. This study provides an important methodology and tool for analyzing user demands and optimizing products. • A novel framework is proposed to identify dynamic product improvement opportunities. • The multi-level demands extraction model (MDEM) extracts multi-level user demands. • Sentiment-intensity-based satisfaction evaluation (SI-SE) measures user satisfaction. • The attention measurement (AM) quantifies user attention based on core demands. • The dynamic opportunity algorithm (DOA) identifies future improvement opportunities.
Li et al. (Mon,) studied this question.