ChatBot AI Economics Tutor
Project Description
This project is an educational chatbot specifically designed for an economics university. The primary goal of the chatbot is to support the learning process by providing an intelligent conversational tool that enables students to learn, test their knowledge, and get answers to questions related to economic topics.
The chatbot was built using the Voiceflow platform, which allows for smooth creation and management of conversational flows. Additionally, the chatbot integrates with OpenAI GPT-4 to provide answers when the required information cannot be found in the local knowledge base, which consists of economic books.
Features and Modes of Operation The chatbot supports three main interaction modes, tailored to different educational needs:
Step-by-Step Teaching Mode: This mode is designed to guide students towards answers by asking additional questions and suggesting directions for their thinking. The goal is to develop the student's ability to solve problems independently and acquire knowledge through an active learning process.
Quiz Mode: The quiz mode is designed to test the student's knowledge. The chatbot asks questions and evaluates the correctness of the student's answers, enabling them to check their understanding in the form of interactive quizzes.
Full Answer Mode: In this mode, the chatbot provides complete answers to students' questions based on the local knowledge base. If the information is not found in the base, the chatbot automatically switches to OpenAI GPT-4 to deliver a comprehensive answer. Chatbot Workflow The chatbot's workflow is designed to be simple and intuitive, optimizing the user's interaction:
Conversation Start: The user starts the conversation by choosing one of three modes: Step-by-Step Teaching, Quiz, or Full Answer mode. Mode Selection: After selecting a mode, the chatbot informs the user of the chosen mode and begins the appropriate interaction. Answering Questions: The chatbot first checks the local knowledge base, which is built from economic books. If no answer is found, it switches to the OpenAI GPT-4 model to provide an answer. Interaction Loop: After providing an answer, the chatbot allows the user to ask further questions. The user can return to the main menu at any time by typing the command "Menu." Knowledge Base Integration The chatbot is integrated with a knowledge base composed of economic books, which has been optimized to allow the AI to quickly and effectively find relevant information. If the knowledge base does not contain the required answer, the chatbot switches to OpenAI models to generate a response.
Knowledge Management: The chatbot first uses the local knowledge base before passing the query to GPT-4. Fallback to GPT-4: If the knowledge base does not provide a sufficient answer, the chatbot automatically switches to GPT-4 to ensure that the user receives a complete answer. Voiceflow as the Main Platform The chatbot was built on the Voiceflow platform, which allows for rapid design of conversational flows and easy integration with external tools like OpenAI GPT-4. Voiceflow offers flexibility and user-friendliness, allowing for dynamic chatbot management and quick implementation of updates to its operation.