Perancangan Aplikasi Chatbot FAQ Berbasiskan Aplikasi Android

Abstract

Abstract

Many users in Indonesia experience difficulties in obtaining information related to the applications they use; for instance, Traveloka customers face challenges in interacting with customer support through email and telephone. Chatbots emerge as a technological solution that replaces human interaction in seeking information, as implemented on other platforms like Shopee. The application development was carried out using Android Studio with Kotlin, along with libraries such as Retrofit for fetching data in JSON format through REST API, for the dialogue interface as well as Text to Speech functionality. Machine Learning was employed using a question-answering model like BERT, and the Stanford Question Answering Dataset (SQuAD) version 2.0 translated into Indonesian language was utilized. The backend is built using Python and RESTful API to handle HTTP Requests such as Post and Get, supported by Terraform for automatic deployment on the Google Cloud Platform. The result is an Android application named 'Loka,' enabling users to easily inquire through a chatbot regarding Traveloka's FAQs, providing a solution for those who prefer not to read the FAQs directly. This solution is expected to effectively address communication issues regarding Traveloka application information through the designed application.