<![CDATA[The advancement of Artificial Intelligence (AI) technology has significantly contributed to the development of interactive systems, including chatbots. This study aims to implement an AI-based chatbot capable of providing skincare product recommendations by understanding the intent behind users' questions or statements. The methods used in this research include Natural Language Processing (NLP) for natural language understanding, and the Long Short-Term Memory (LSTM) model as a machine learning algorithm for intent classification. Conversational data was collected and processed through text preprocessing stages such as tokenization, stemming, and padding. The LSTM model was trained using a labeled dataset to recognize various user intents, such as product recommendations, product information, and general skincare-related inquiries. Evaluation results showed that the model achieved an accuracy of 80%, with a precision and recall of 85.7% each. Furthermore, the Mean Average Precision (MAP) score, which measures the relevance between user queries and the system’s responses, reached 100%, indicating that all 20 test questions were answered as expected. The implementation of this chatbot is expected to assist users in efficiently selecting skincare products that match their needs, eliminating the need for manual information searching.]]>
Copyrights © 2025
