Research work · 2026

Tendencias temáticas de la inteligencia artificial: el caso de la producción científica de la revista Artificial Intelligence Journal

Cristina Restrepo-Arango

Revista Ibero-Americana de Ciência da Informação

Abstract

Objective: Perform a co-word analysis to identify the research topics that have been used in artificial intelligence. Method: Keywords from 3,050 academic articles published in the Artificial Intelligence Journal (AIJ) from 1970 to 2024 were analyzed. Co-word analysis was used, and the R programming language was used in the RStudio integrated development environment (EDI). Therefore, with the refined keywords, a co-word matrix was generated using BibExcel software. This matrix was manipulated with the igraph package to obtain co-word networks for five time periods. Results: It was found that in the 1970s and 1980s, the predominant themes were systems science, cybernetics, and computer science. In the 2000s, the most recurrent themes were algorithms, semantics, heuristic methods, and game theory, among others. These themes continued into the 2010s and 2020s; although other keywords appeared in the latter decade, such as computer circuits, learning algorithms, polynomial approximation, and reinforcement learning, among others. Conclusions: The evolution of AI topics is evident in the keywords with the highest number of links in each of the five periods analyzed. The first period (1970 to 1980) was systems science and cybernetics. The second period (1981 to 1990) was artificial intelligence and systems science and cybernetics. The third period (1991 to 2000) was artificial intelligence and constraint theory. The fourth period (2001 to 2010) was artificial intelligence and computational complexity. The fifth period (2011-2024) was semantics and data description.

DOI: 10.26512/rici.v19.n2.2026.57764