Abstract:
"The research initiative aims to create a cosmetics-selling website that provides personalized cosmetic recommendations to customers based on their browsing preferences. The goal is to simplify the often challenging and time-consuming process of choosing the right cosmetic products. To achieve this, a recommendation engine utilizes a combination of content-based and collaborative filtering strategies, employing the cosine similarity algorithm. It analyzes customers' browsing histories and identifies products viewed by users. Then, it compares these products with others in the same category to pinpoint the most similar items, generating tailored recommendations for each customer.
To evaluate the system's performance, data science measures including accuracy, recall, and F1-score were employed. The results showed that the recommendation engine achieved high levels of precision and recall, affirming its success in suggesting appropriate cosmetic products based on user browsing history. Furthermore, usability testing was conducted to assess the website's user interface, and the findings indicated that the design was user-friendly and straightforward to navigate.
In conclusion, this research endeavor offers a practical solution to the cosmetics selection dilemma by creating a user-friendly website that harnesses advanced algorithms to provide customers with personalized product recommendations, ultimately enhancing their shopping experience and product satisfaction."
