Oveview

In this project we attempt to study technological progress and the diffusion of technology across space over several millennia by assembling data from publicly available collections of museum records. The leading examples include the British Museum, the Metropolitan Museum of Art, the Victoria & Albert Museum, and many and a growing number of other museums across the whole world that have allowed us electronic access to their collection records. The data often include metadata on holdings such as the location where the object has been found, the time when the object was made, the provenance, material, techniques, and other keywords. Together, we hope that this data will provide one of the most comprehensive pictures available to-date of the humanity’s knowledge of materials and techniques used in the past. Using this data, we aim to test some of the leading theories of the drivers of technological diffusion across space.

A large part of the project has been the assembly and harmonization of data from different sources. In particular, a key problem has been the extraction of the information regarding the production techniques and materials from the original meta data provided to us by the museum collection data. In solving this problem, we have built on recent advances in Natural Language Processing (NLP). The following working paper presents the results of the tools that we have developed for this purpose:

Sebastian Cadavid-Sanchez, Khalil Kacem, Rafael Aparecido Martins Frade, Johannes Boehm, Thomas Chaney, and Danial Lashkari. Working Paper. 2022 "Evaluating End-to-End Entity Linking on Domain-specific Knowledge Bases: Learning about Ancient Technologies from Museum Collections".

Download

• NLP Working Paper