Researchers have created artificial intelligence systems and advanced software to deal with different forms of language, whether spoken or written, and this software has attracted a great deal of interest, especially in the field of communicating with robots and voice assistance systems And other interactive programs.
In this context, researchers at the University of Colorado and Drury in the United States have developed an artificial intelligence system that can system hair.
The origin of this idea is due to researcher Brendan Pena of the University of Colorado who originally wanted to devise a system to imitate the feelings that are expressed through the lyrics of songs, but since most of the lyrics of the songs are subject to laws protecting intellectual property, Pina decided to develop a system for deep learning of the poetry system.
Rather than focusing on the literary standards that control the writing of verses, the researchers focused on the creative and emotional aspects that can be expressed in poetry.
The TechExplore website, which specializes in technology, reported that the study team fed the system with large quantities of verses from different sources, and then classified them under different sections according to the type of feelings expressed by these poems, such as a section of poems expressing happiness, sadness or anger Etc.
The new system works by creating verses based on the original verses in the basic database.
The quality of the verses that were produced by literary critics and specialized programs were tested in this purpose, and they concluded that the verses that were organized succeeded in expressing sadness at 87.5 percent and joy at 85 percent.
"The new system has proven that texts that are produced electronically can stimulate the same emotions in the hearts of readers as the literary texts written by human authors, and we hope that this experience will open the door wide to the use of artificial intelligence applications in other literary fields in the future," Pina says. .
