Implementasi dan Analisis Kinerja Chatbot Telegram Rekomendasi Kuliner di Kabupaten Semarang Menggunakan Framework Rasa

The advancement of information technology has driven innovation in various sectors, including the culinary industry. Semarang Regency, as a culinary tourism destination, offers a wide range of dining options that often make it difficult for tourists to decide where to eat. This study aims to implement and analyze the performance of a Telegram-based chatbot using the Rasa framework as a culinary recommendation medium in Semarang Regency. This chatbot is designed to provide quick and relevant culinary recommendations according to user preferences through the utilization of Natural Language Processing (NLP). The system development was carried out through several stages, starting from user needs identification, system design, chatbot implementation, to testing using the System Usability Scale (SUS) method. The test results showed that the developed chatbot achieved an average SUS score of 79.16, indicating that the system meets feasibility standards and provides a satisfying user experience. Therefore, this chatbot is effective in helping people quickly, flexibly, and efficiently find culinary recommendations in Semarang Regency.