Evolution of Natural Language Processing

Abstract

Written language has not been immune to cultural and technological changes. For example, the invention of the printing press in the 15th century, or the development of personal computers and smartphones in recent decades, were milestones that dramatically marked its evolution. In parallel with the development of language, the evolution of computing and the emergence of intelligent algorithms capable of making decisions have made impressive advances in recent years, forming a new area of study known as artificial intelligence (AI), considered by many as the next great revolution. The integration of AI techniques to interpret, manipulate and understand human language results in the branch of natural language processing (NLP), which, taking into account the latest advances in language generation models, can become the next great milestone in written language. The first NLP algorithms were rule-based, but later supervised classification schemes based on models such as logistic regression, support vector machines, hidden Markov models, or conditional random trees began to be used. The problem with traditional classification models is that they are oriented to evaluate each word, or the relationships between each word and the previous word, but they do not capture the context of the words in a complete sentence.