In a context characterized by increasing digitalization and the abundance of textual data, the ability to identify people through linguistic descriptions represents a significant advance in fields such as security, forensics, and human talent management. This paper presents a comprehensive proposal that combines natural language processing (NLP) techniques and semantic analysis to address the variability and complexity of human language in the identification of individuals. A detailed review of the literature is carried out, a modular architecture is proposed and validated through the development of a functional prototype. The experimental results show notable improvements in the accuracy of identification, which demonstrates the potential of this methodology for applications in real environments with high requirements for accuracy and beyond robustness.
Cossio et al. (Mon,) studied this question.