Geological profile-text information association model of mineral exploration reports for fast analysis of geological content

Abstract

Mineral exploration reports include not only a large number of geological profiles but also geological text by offering valuable information and knowledge about the geological environments in which mineral deposits form. Extracting and understanding historical data can assist in the fast analysis of geological content and support 3D model construction. However, geological texts are written in unstructured form and have many correlations with geological profiles. It is a challenging task to derive meaningful geological information without manually reading through a large collection of reports, which is a formidable task for geologists. This paper proposes a geological profile-text association framework for constructing a knowledge graph, and it aims to understand the contents of the geological profile, transform a larger amount of textual data into structure form, and link the geological profile and text to a graph-based knowledge representation that assists further analysis of knowledge discovery. The concept of constructing vector geological profile rock layer objectification is proposed to make each rock layer with geometric features and attribute information for the geological profile, and the geological entity relationship is extracted to form a triple by deep learning and stored and expressed in the form of a graph structure for geological text. Finally, a geological profile and text association model is established by word vector similarity. The proposed approach is capable of rapidly and robustly understanding geological profiles, extracting geological texts, establishing correlations between them, and performing geological knowledge mining.