Modeling Education Studies Indexed in Web of Science Using Natural Language Processing

Easier access to information and resources allowed researchers conduct more studies publish most of them electronically. They are indexed in scholarly citation databases such as Web Science Scopus. These index huge volumes research reports. Even though they offer search engine filtering options, it is still hard locate the publications which their contents closely related. Artificial intelligence technologies, Natural Language Processing, allow documents be categorized based on content. Top2Vec an unsupervised topic modeling algorithm that enables users categorize semantically. The purpose current study twofold: (1) provide with ability group applying Processing techniques, (2) reveal topics highest number articles ‘education scientific disciplines’ category within Core Collection database 2021. Colab notebook used type Python codes for executing algorithm. This yielded 68 distinct among 8125 published 2021 under Education Scientific Disciplines category. After modeled were ranked from having largest (i.e., N=549) least N=29), first eight topics' findings presented discussed. listed follows: Physics (N=549), online education covid (N=438), Chemistry (N=381), Math Reasoning (N=377), Psychology Emotions (N=257), Educational Diversity (N=228), Health Life (N=223), Mentoring Leadership (N=204).