Dataset of Exploring Consumer Perspective on Bittersweet by Najla: A Data-Driven Analysis Using Naive Bayes
Description
This dataset contains consumer reviews of Bittersweet by Najla, a popular dessert brand in Indonesia, collected from three major platforms: Tokopedia, PergiKuliner, and GoFood. The reviews are in Bahasa Indonesia and reflect customer opinions on various aspects of the brand, including product quality, taste, pricing, and overall satisfaction. The research hypothesis suggests that consumer sentiment is mostly positive, with strengths in taste, packaging, and variety, while concerns related to pricing and texture consistency may contribute to negative feedback. By analyzing these reviews, this study seeks to uncover recurring themes and factors that influence purchasing decisions. Data collection was conducted through web scraping using Octoparse, covering a period from January 2022 to October 2024. Initially, 1,400 reviews were gathered, but after filtering out duplicates, irrelevant entries, and incomplete data, the final dataset consists of 1,314 reviews. The dataset is structured with three main attributes: a unique Review ID, the Platform (Tokopedia, PergiKuliner, or GoFood), and the Review Text in its original form. The sentiment classification (positive or negative) has not yet been assigned, allowing for further analysis using machine learning or natural language processing techniques. Preliminary observations indicate a mix of positive and negative feedback. Many consumers praise the taste, packaging, and product quality, while concerns over pricing and texture consistency appear frequently in negative reviews. Differences in feedback across platforms also suggest variations in customer expectations and experiences, particularly regarding delivery and service quality on food-ordering platforms. This dataset holds significant value for various applications. Researchers can use it for sentiment analysis studies, particularly in Indonesia’s F&B industry, to understand consumer preferences and market trends. Businesses can analyze the data to identify key customer concerns, refine pricing strategies, and improve overall product consistency. Additionally, the dataset can serve as a training set for machine learning models in sentiment classification. Companies looking to enhance their digital marketing strategies can leverage these insights to optimize brand positioning and better understand consumer behavior. By exploring customer feedback, this dataset provides a foundation for improving customer satisfaction, strengthening brand loyalty, and guiding data-driven decision-making in the dessert industry.