Identification of AI-Related Topic Evolution Based on Extraction of International Conference Proceedings Titles

International conference proceedings are a forum for researchers in various parts of the world to communicate their research findings to the scientific community and to the public in general. International conference proceedings published by globally recognized institutions are of unquestionable quality. Every conference must have a title that describes the theme or topic being discussed. This study wants to reveal the topic evolution by extracting conference titles. This study collects conference titles that contain terms related to artificial intelligence (AI) in them. Then the titles that have been collected are extracted to obtain a base noun phrase and filtered to eliminate uninformative terms. Of the 139,495 conference titles, it was obtained 8,133 titles containing topics related to AI. The results of the title extraction are 4,601 terms which are dominated by the exact term ‘artificial intelligence’, followed by the terms ‘machine learning’, ‘applications’, ‘intelligent systems’, ‘semantic’, ‘pattern recognition’ and other important related terms. Based on the glossary, frequency of occurrence, and year of publication, further identification of the evolution of a more specific topic is presented in this study. This study identified the early emergence and evolutionary phases of topics for specific terms such as natural language processing, cloud computing, and cybernetics or cyber security. Using the model proposed in this study, the process of identifying the evolution of specific topic can be extracted to get new insights.