Heang Seavleu1, Author, and Suraph Sigh2, Supervisor
1Woosong University, AI & Big Data, Daejeon, South Korea
2Woosong University, AI & Big Data, Daejeon, South Korea
ABSTRACT This study aims to improve the accessibility, affordability, and quality of care in the healthcare sector via artificial intelligence (AI) chatbots, natural language processing (NLP), and blockchain technology. Chatbots can be used to provide 24/7 support to all patients despite their background and their privilege of accessibility, answer medical questions, and even provide personalized medical advice. With the help of natural language processing (NLP), NLP could be used to train chatbots to understand and respond to complex medical queries, and to develop new tools for disease diagnosis and treatment. Blockchain technology can help to improve security, privacy, and transparency. This paper proposes a new chatbot architecture that uses NLP, and ML that are capable of learning from data and improving its performance over time, prompt engineering for fine-tuning, and blockchain technology to secure the data. The study aims to obtain a flexible and scalable architecture that can be used to develop chatbots for a variety of healthcare applications. This paper also discusses the potential benefits and challenges of using blockchain technology in chatbot-based healthcare systems.
INDEX TERMS Chatbot, Generative AI, Blockchain, Smart Contract
A “bot” is defined as software capable of performing an automated task. Chatbots perform automated functions through a user interface for human interaction. Some chatbots follow predefined scripts and can handle only specific queries, such as asking about store opening times. On the other hand, more sophisticated chatbots leverage artificial intelligence, enabling them to learn and improve with each human interaction. As they engage more with users, these advanced chatbots become increasingly intelligent.
AI chatbots use a combination of machine learning and what’s known as natural-language processing. Machine learning involves employing specific algorithms to analyze input data and recognize patterns within that data—the more input data the system receives, the better it becomes at identifying patterns, leading to increased intelligence. Natural language processing, on the other hand, transforms human speech into a format comprehensible to computers. It then analyzes the input and provides a response in a manner understandable to humans. The synergy between these two fields enables platforms like Google to analyze a partial phrase entered in the search bar and suggest likely search queries (Indorse, 2001)[2].
Blockchain technology has shown a great potential to revolutionize various industries beyond finance. One notable area where blockchain development services certainly could make a huge impact is the invention of blockchain integrated chatbot (Dasila, 2023) [1]. Traditional chatbot systems frequently strongly on a single point centralized server architecture and infrastructure, meaning all the all are process and store in the single location or nodes, which potentially pose as a single point failure. Noticeably, same as other advance technology this particular infrastructure contain several limitations and risks exposure, including scalability constraints, and high security vulnerabilities to malicious attack (Dasila, 2023)[1]. A decentralized AI built-in chatbot can potentially solve it by distributing the ethical chatbot’s logic and data across multiple nodes in a network. To provide redundancy and fault tolerance, each node maintains a copy of the blockchain ledgers—a paradigm shifts in ensuring the reliability and robustness of AI-driven chatbot systems. This paper explores the synergies between AI, blockchain, and chatbot technologies, presenting a novel architecture poised to redefine accessibility, affordability, and quality of care within the healthcare sector.
In recent years, the healthcare industry has received significant popularity when it comes to new technologies such as in the field of NLP. NLP is known for its effective ability to manage massive amounts of data from diverse clinical documents and patent records, thereby assisting healthcare professionals and patients.
In this paper, we will focus on the process of how to build a generative AI chatbot for healthcare by integrating blockchain technology.
Large language models (LLMs) over the past few years, LLMs has been substantial press recently regarding its impressive performance on natural language processing (NLP) tasks [4]. They achieved their success by expanding the training of models based on transformers. Model performance and data efficiency have been found to increase with both model and dataset size. Large language models (LLMs) are commonly trained through self-supervision on extensive datasets, such as Wikipedia and BooksCorpus. They have shown promising outcomes across various tasks, including those requiring specialized scientific knowledge and reasoning. Particularly intriguing is the in-context few-shot capability of these LLMs, enabling them to adapt to diverse tasks without relying on gradient-based parameter updates. This rapid adaptation allows them to generalize quickly to unseen tasks and even demonstrate apparent reasoning abilities when prompted appropriately. Several studies have shown that LLMs have the capacity to act as implicit knowledge bases. Despite this capacity, there is a notable risk associated with these models, including the potential for generating hallucinations, amplifying social biases ingrained in their training data, and exhibiting shortcomings in their reasoning abi
Define abbreviations and acronyms the first time they are used in the text, even after they have already been defined in the abstract. Abbreviations such as IEEE, SI, ac, and dc do not have to be defined. Abbreviations that incorporate periods should not have spaces: write “C.N.R.S.,” not “C. N. R. S.” Do not use abbreviations in the title unless they are unavoidable (for example, “IEEE” in the title of this article).