Using transformer architectures to analyze text data by automating the search for common comments

The article discusses machine learning and natural language processing methods for automating the analysis of text data. Using the example of project documentation review, it shows how the use of the BERT model allows for the identification of typical comments, the acceleration of information processing, and the reduction of the human factor. The developed software package automates the process of finding repetitive errors, facilitates the work of experts, and provides a basis for optimizing business processes in the context of increasing data volumes.

Authors: O. V. Afanasieva, I. M. Novozhilov, D. A. Pervukhin, P. A. Skukovskii

Direction: Informatics, Computer Technologies And Control

Keywords: natural language processing (NLP), BERT, text data analysis, text processing automation, transformer architectures

---

View full article