The Prakerja Card Program is a job skills development and entrepreneurship program for job seekers. The Prakerja Card Program has a list of frequently asked questions or called Frequently Asked Questions (FAQ), because FAQ is still a list of questions that approach what the user typed, and users should look to find out which questions match the questions. Therefore, it takes a way to respond directly to users, namely with a chatbot. Chatbot is a conversational modeling program that uses human natural language to simulate interactive conversations between machines and humans. The chatbot interprets messages from the user, processes the user's words, determines what the chatbot needs to do based on the user's instructions, executes them, and finally informs the user of the conclusions. This research used Neural Network Multilayer Perceptron algorithm and Nazief & Adriani stemming in creating a chatbot with the data used is FAQ data on the Prakerja website. The purpose of this study is to find out the great performance accuracy of chatbot responses produced by Neural Network Multilayer Perceptron and Nazief & Adriani on Prakerja FAQ chatbots. The results showed that the chatbot could answer 72 questions with a difference of 100 questions asked by respondents
Copyrights © 2022