[Paper Review] An ontology-based chatbot for crises management: use case coronavirus

Khouloud Hwerbi|arXiv (Cornell University)|Nov 2, 2020

Advanced Text Analysis Techniques49 references4 citations

TL;DR

This paper presents an ontology-based chatbot named 'COVID Assistant' for crisis management during the COVID-19 pandemic, integrating web scraping, a finite state machine, keyword extraction, and a trained NLP model to deliver real-time, accurate information. The system achieves 24/7 availability and was formally verified using Uppaal for correctness, ensuring reachability, safety, and liveness in dialogue states.

ABSTRACT

Today is the era of intelligence in machines. With the advances in Artificial Intelligence, machines have started to impersonate different human traits, a chatbot is the next big thing in the domain of conversational services. A chatbot is a virtual person who is capable to carry out a natural conversation with people. They can include skills that enable them to converse with the humans in audio, visual, or textual formats. Artificial intelligence conversational entities, also called chatbots, conversational agents, or dialogue system, are an excellent example of such machines. Obtaining the right information at the right time and place is the key to effective disaster management. The term "disaster management" encompasses both natural and human-caused disasters. To assist citizens, our project is to create a COVID Assistant to provide the need of up to date information to be available 24 hours. With the growth in the World Wide Web, it is quite intelligible that users are interested in the swift and relatedly correct information for their hunt. A chatbot can be seen as a question-and-answer system in which experts provide knowledge to solicit users. This master thesis is dedicated to discuss COVID Assistant chatbot and explain each component in detail. The design of the proposed chatbot is introduced by its seven components: Ontology, Web Scraping module, DB, State Machine, keyword Extractor, Trained chatbot, and User Interface.

Motivation & Objective

• To address the urgent need for timely, accurate, and accessible information during the COVID-19 pandemic.
• To design a chatbot that operates 24/7 with high reliability and contextual awareness in crisis scenarios.
• To integrate structured knowledge via ontology with dynamic learning from web sources and trained NLP models.
• To formally verify the chatbot’s dialogue logic for correctness, safety, and deadlock-free behavior.
• To provide a scalable, extensible architecture applicable to broader disaster management contexts.

Proposed method

• Developed a domain-specific ontology (CVIO) in OWL to represent medical, epidemiological, and administrative knowledge about COVID-19.
• Implemented a web scraping module to dynamically extract and populate up-to-date data from official sources into the ontology.
• Designed a finite state machine (FSM) to model the chatbot’s dialogue flow, managing user input states such as symptom inquiry, country status, and general queries.
• Integrated a trained chatbot using ChatterBot with AIML-based intent recognition and response generation.
• Employed keyword extraction to map user queries to appropriate ontology concepts and FSM states.
• Validated the system’s behavioral correctness using Uppaal, verifying reachability, safety, and liveness properties through model checking.

Experimental results

Research questions

• RQ1 How can an ontology-based chatbot be architected to ensure real-time, accurate, and context-aware responses during a public health crisis?
• RQ2 What combination of knowledge representation, dynamic data ingestion, and dialogue management is most effective for crisis chatbot systems?
• RQ3 Can formal verification techniques like model checking ensure the reliability and correctness of a chatbot’s dialogue logic?
• RQ4 How can a hybrid architecture combining rule-based, trained, and ontology-driven components improve response quality and system robustness?
• RQ5 What are the key components and design patterns that enable 24/7 availability and scalability in disaster response chatbots?

Key findings

• The chatbot successfully integrated an ontology (CVIO) with real-time data from web scraping, ensuring up-to-date information delivery.
• The finite state machine enabled structured, context-aware dialogue management across diverse user queries, including symptoms, country status, and general information.
• Formal verification using Uppaal confirmed that the system is deadlock-free and guarantees response generation under all valid state transitions.
• Reachability, safety, and liveness properties were verified using temporal logic formulas, with all properties confirmed as valid (marked in green in Uppaal).
• Parallel execution simulation demonstrated that multiple chatbot instances could operate concurrently without conflict, supporting scalability.
• The system achieved full 24/7 operational availability, with no reported failures in the tested dialogue scenarios.