The goal of this project was to develop an innovative software capable of handling zero-shot tasks in a specific domain. The focus is on semantic information extraction from globally distributed data.
The challenges lie not only in processing heterogeneous data but also in managing distributed information from various parts of the world. To overcome these challenges, a combination of advanced technologies is employed, including Transformer models, the XGBoost network, empirical methods, and classical statistics.
IMPLEMENTATION OF THE SOLUTION
A key component of the project is the construction of an extensive language model (Large Language Model, LLM) designed to classify real-world events and transform them into directed acyclic graphs. These graphs visually represent organizational structures and allow for in-depth analysis.