Natural language and mathematical reasoning
Daniela López De Luise · Institution of Engineering and Technology eBooks · 2023
Every natural language presents certain regularities that have been studied for years. From statistical approaches to machine learning, many authors have found that automatic processing is far more complex than any other brain production. Despite the current status in the field, many applications like chatter-bots, speech recognition, and sentiment analysis, show that there is still an interesting gap between analysis and production of sentences. Despite the solutions used nowadays, the deep essence of linguistic reasoning dynamics keeps mostly not revealed, and many of the current approaches involve the conception of restricted patterns for human linguistic reactions, usage, and interpretation. This chapter presents a perspective of this problem centered in the idea followed by many authors that considers the brain as a complex device working under some kind of fractal rules, and deeply related to entropy. As part of the scope, there is a short review of some of the most paradigmatic proposals of language mathematical descriptions related to entropy and fractals, the relationship between them in this context and a summary of previous publications of the author that aim to reveal the deep nature of natural language, modeling linguistics-related brain activity.