A new study evaluating the capabilities of artificial intelligence in identifying vocabulary transferred from one language to another.


Researchers at the University of Pontificia in Catalana in Peru and at the Max Planck Institute for Human History in Germany conducted a study to examine the ability of artificial intelligence to identify vocabulary transferred from one language to another based on a specific algorithm. . Equation.

The phenomenon of “language borrowing” is the use of vocabulary from one particular language to another language or languages, which helps researchers in the field of language to follow and monitor the development of modern languages. Cultural relationships between different language groups, but researchers generally find it difficult to maintain. Practical because it requires difficult comparisons between different languages.

In a statement issued by the technology website “Viz.org”, researcher Johan Mattis said, “Automated monitoring of language computing terminology is still a difficult task in the context of language computing.”


As part of the experiment, researchers attempted to train several systems of artificial intelligence to transfer languages ​​used by linguists from one language to another in phonology-based monitoring terminology. That is, by overseeing words that are pronounced in many ways. Languages, indicating that this is the subject of study. During the stages of language development, they actually moved from one language to another.

The study team claims to have tried an artificial intelligence system on a modified version of the “World Color Word” database, which includes word patterns or vocabulary that move between forty languages ​​around the world to explore their potential. A system that recognizes words borrowed from one language to another.

The researchers are confident that the results obtained by the artificial intelligence systems will in most cases not be satisfactory in their perspective, reflecting the inability of the artificial intelligence methods to monitor such terminology according to the currently applied methods.


John Miller, a researcher at the University of Cotelca, Pontificia, says: “After conducting basic experiments on the phenomenon of language debt, it is possible to move forward after solving some of the problems faced by artificial intelligence systems.” However, other researchers, such as study team member Thiago Trisoldi, say that “new computer-aided curricula already demonstrate the importance of using computational methods in the fields of language comparison and historical linguistics.”


Leave a Reply