QueryPlot: Mineral Prospectivity Mapping with NLP

A new framework, called QueryPlot, promises to simplify and accelerate mineral prospectivity mapping, a traditionally manual and knowledge-intensive process. QueryPlot integrates large-scale geological text corpora with geologic map data, leveraging modern Natural Language Processing (NLP) techniques.

The system is able to synthesize heterogeneous deposit models and geospatial datasets to identify regions with a high probability of hosting specific types of mineral deposits. QueryPlot transforms state geological maps into structured textual representations and, through natural language queries, encodes both queries and region descriptions using a pre-trained embedding model.

Semantic similarity scores are calculated to rank and spatially visualize regions as continuous evidence layers. QueryPlot supports compositional querying over deposit characteristics, enabling the aggregation of multiple similarity-derived layers for multi-criteria prospectivity analysis. A study on tungsten skarn deposits demonstrated that the system achieves high accuracy in identifying known occurrences.

The source code and datasets used in this study are publicly available to support future research.