OREANDA-NEWS  Specialists from the Rosneft Tyumen Scientific Institute have developed and implemented a service based on large language models (LLM, an artificial intelligence program capable of recognizing and generating text) to automate the analysis of lithological descriptions of the core. This is stated in the message of Rosneft.

The digital tool "LithOtext" automatically recognizes 16 physical parameters of the core in the texts of geological reports, including rock, color, saturation, texture, grain size, fracturing, and type of void space. The parameters are determined using a large language model, an artificial intelligence capable of reading and understanding text.

Lithology experts train the model: they suggest how to accurately determine the characteristics of the core, explain the meaning and correct interpretation of geological terms. The company emphasizes that the use of innovative development minimizes the human factor, reduces the likelihood of errors, and most importantly, speeds up information analysis by 5-7 times.

The tool is used both to process new information and to check the quality of descriptions in databases. All lithological descriptions and primary data from laboratory core studies are consolidated in the RN-LAB information system developed and implemented at the scientific institute. Currently, RN-LAB contains digital data from core research, which has a length of more than 52 kilometers.