<![CDATA[This research develops a chatbot based on the Telegram platform that integrates Natural Language Processing (NLP) technology to provide information about mobile phones quickly, accurately, and efficiently. With the increasing need for users to access data regarding specifications, prices, reviews, and device damage diagnosis interactively, this chatbot becomes a relevant solution in supporting digital literacy and improving user experience. The system was developed using Agile methodology through stages of needs analysis, interface design, Telegram API implementation, and NLP integration with BERT architecture for intent recognition and named entity recognition. Data was collected through web scraping from e-commerce platforms, technology review websites, and community forums, then structured in a MongoDB database. Main features include product specification searches, damage identification, latest news, and interactive guides that support device problem-solving. White Box Testing showed satisfactory results with Statement Coverage 92%, Branch Coverage 88%, Intent Recognition accuracy 87%, Response Time 2.1 seconds, and Query Success Rate 97.5%. Evaluation results confirm that the chatbot is able to perform its functions responsively and practically, ready for deployment, and has potential to be expanded to other digital platforms to increase information technology competitiveness in the digital era.]]>
Copyrights © 2025
