Abstract  Recognizing distribution patterns of ore-related pathfinder elements requires an   understanding of the corresponding ore-forming systems and may lead to the definition   of more reliable exploration targets. Given that complex ore-forming processes in   complex geological terrains can result in either enrichment or depletion of ore elements,   simultaneous consideration of both processes in geochemical anomaly modelling   would be advantageous for defining exploration targets. Here, we adapted the “cosine   similarity measure” method for pattern recognition of ore-related geochemical   anomalism to facilitate simultaneous modelling of element enrichment and depletion.   We tested the performance of this new method on porphyry copper exploration data   from the Manzhouli belt, China, and demonstrated that our method is highly efficient   with respect to recognizing ore-related geochemical signatures and, thus, useful for   “vectoring” towards porphyry copper deposits and defining more reliable   geochemical exploration targets.