International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 11 Issue: 10 | Oct 2024
p-ISSN: 2395-0072
www.irjet.net
A Survey on Various Chatbots Dr. G.Jessy Sujana1, V. Sai Sankar Raju2, P. Sri Sai Rishith3, P. Revanth Balaji4 Asst. Professor1, Student2 , Student3 , Student4 Department of Computer Science and Engineering1, SRM Institute of Science and Technology, Chennai, India1 ----------------------------------------------------------------------------***-------------------------------------------------------------------------marketing, sales, and education. They offer numerous Abstract— In today's fast-paced world, people
constantly seek new information and more efficient learning methods. Traditional approaches, such as books, search engines, and online encyclopedias, can be both time-consuming and challenging. Chatbots offer a solution by providing instant answers across various sectors, including banking, retail, travel, healthcare, and education. These computer applications are designed to mimic real-world conversations, powered by artificial intelligence (AI) and natural language processing (NLP) technologies, allowing them to comprehend and respond to human language. Chatbots can perform tasks, provide information, and address queries with ease.
This paper surveys several methodologies for chatbot implementation, including cutting-edge techniques like Long Short-Term Memory (LSTM), Natural Language Generation (NLG), Supervised Machine Learning (SVM), Model Driven Engineering (MDE), Dialogue Management, Human-in-the-Loop (HITL), Audio-Frame Mean Expression (AFME), Random Forest, Support Vector Machine (SVM), Recurrent Neural Network (RNN), and Convolutional Neural Network (CNN). These methodologies enable the development of chatbots with advanced natural language understanding and response capabilities. The objective of this paper is to review and compare various chatbot development approaches to identify the most suitable for specific applications. The study concludes that while multiple methodologies and algorithms can be used to implement a chatbot, each method has its strengths depending on the task requirements. The paper also presents evaluations of the methodologies' accuracies and provides suggestions for their application.
Keywords — CHATBOT, ARTIFICIAL INTELLIGENCE, NATURAL LANGUAGE PROCESSING. I. INTRODUCTION Chatbots are computer programs that use conversational AI technology to mimic human user chats. They are widely used in customer service and healthcare settings, where they can answer inquiries, provide assistance, and solve problems. Beyond these areas, chatbots are also utilized in
© 2024, IRJET
|
Impact Factor value: 8.315
|
benefits, such as making our lives easier, more convenient, and enjoyable. Chatbots can respond to questions 24/7, deliver personalized recommendations, automate various customer service tasks, engage with customers across multiple channels, and tailor the customer experience by learning individual preferences. While still evolving, chatbots have the potential to revolutionize how we interact with machines. They are a powerful tool for improving customer service, reducing costs, and boosting efficiency. In education, chatbots can support students struggling with specific concepts or those needing more personalized attention. By making learning more engaging and interactive, chatbots can enhance the overall educational experience and increase students' interest in the subject matter. This paper explores the key technologies available for building educational chatbots. It addresses the vanishing gradient problem by using memory cells to store long-term information. It also discusses Natural Language Generation (NLG), which creates natural language text from structured data. Supervised Machine Learning (SVM) is highlighted for its ability to learn from labeled data, where each data point is associated with a known outcome or label. Model Driven Engineering (MDE) is presented as a powerful chatbot development platform that helps businesses create highquality chatbots more efficiently. Dialogue management, which combines statistical models and rule-based systems, ensures smooth conversations and accurate recognition of user intent. The paper also covers Human-in-the-Loop (HITL) tasks, such as fraud detection, object detection, image classification, and medical diagnosis. AFME (Audio-Frame Mean Expression), a feature extraction technique used in speech processing and music information retrieval, is discussed as well. Random Forest, which constructs multiple decision trees and bases predictions on their consensus, and Support Vector Machine (SVM), which classifies data points by finding a hyperplane, are also examined. Convolutional Neural Networks (CNNs), known for learning spatial hierarchies in data, are particularly well-suited for tasks like image classification, object detection, and semantic segmentation. The paper concludes that there is no one-sizefits-all solution; the best approach depends on the specific context and requirements.
ISO 9001:2008 Certified Journal
|
Page 384