Gabriel Cifuentes

New methodologies applied to the study of Pleistocene bone surface modifications in the fossil record of Olduvai Gorge: a union between Taphonomy and Experimental Archaeology.

Acknowledging the problems that traditional approaches have shown in Taphonomic research, where structurally similar marks and highly subjective classification methodologies have propitiated a situation of uncertainty about the archaeological record, this Thesis is aimed to develop new methodologies to overcome this equifinality by joining Artificial Intelligence (AI) and Experimental Archaeology.

Experimental archaeology will be used both to replicate the different possible marks found in the African Pleistocene archaeological record, and to test modern methodologies regarding the study of past remains. The AI approach will be based on the use of Convolutional Neural Networks and Computer Vision algorithms that will be able to analyse archaeological images to an extent that is far beyond human capabilities, while maintaining human input to the lowest possible.

With this, it is expected that this methodology will be able to overcome the problems posed by subjective approaches and highly similar marks, and it will be used to study the bone record of some of the Bed II archaeological sites of Olduvai Gorge (specifically FLK West, SHK, TK and BK).