چكيده لاتين
The presence and sustained success of businesses in competitive markets require access to precise information, a deep understanding of prevailing market dynamics, and accurate assessments of supply and demand. Several factors, such as inflation rates, market size, the availability of substitute or complementary products, and especially customer preferences and opinions, influence product demand. Among these, the present study places specific emphasis on analyzing customer opinions. In recent years, the rapid expansion of digital platforms and the growing tendency of customers to share reviews on online marketplaces have created unprecedented opportunities to analyze consumer perspectives, needs, and preferences. This study aims to identify the key factors influencing customer satisfaction and demand by analyzing reviews posted on the Digikala and Emalls websites within the home appliance sector. Five key product categories were examined: washing machines, dishwashers, refrigerators, televisions, and gas stoves. Conducted under the framework of a contract with the Entekhab Industrial Group, this research places particular focus on the Snowa brand, benchmarking it against major competitor brands, with the aim of offering a comprehensive understanding of customer expectations in this competitive landscape. The study employs an aspect-based sentiment analysis approach. Following data collection and preprocessing, a manual labeling process was carried out on portions of each product’s dataset to identify the specific aspects mentioned and the corresponding sentiments expressed. Subsequently, a range of traditional machine learning models, deep learning architectures, and the ParsBERT language model were implemented and evaluated. The results indicate that for datasets related to washing machines, refrigerators, and gas stoves, the Bi-LSTM model demonstrated superior performance, achieving accuracies of 80%, 72%, and 73%, respectively. For dishwashers and televisions, the CNN model outperformed others, reaching accuracies of 82% and 72%, respectively. An analysis of 5,470 washing machine reviews identified the most influential factors on customer satisfaction and demand, ranked as follows: product performance, noise and vibration levels, Digikala services, price, after-sales service quality, physical specifications, overall product quality, available washing programs, and energy consumption. These aspects dominated the themes discussed by customers in their feedback. The findings reveal that excessive noise and vibration, product performance, and after-sales services accounted for the largest share of negative comments — underscoring the need for manufacturers to prioritize improvements in these areas to enhance their competitive positioning. Among the ten brands examined, Snowa washing machines ranked fourth overall. The most positively perceived aspects were energy consumption and physical specifications, while the productʹs performance received the most negative evaluation. Similar trends were identified across the other selected home appliance products. By delivering a data-driven, precise, and scalable framework for analyzing Persian-language customer reviews, this study provides actionable insights to inform strategic decision-making in product design, after-sales service enhancement, and the development of targeted marketing strategies. The primary innovation of this study lies in its utilization of real market data, the application of advanced natural language processing techniques, and the in-depth sentiment analysis of customer feedback.
