Scope

We began mapping out the territory of this trade in 2016. We collected thousands of posts and studied the language of the posts - how the collectors and enthusiasts described their engagement with the remains. At that time, we studied only one platform. Our methods were primarily textual.¹

In this project, we intend to explore the leads suggested in that first study by developing and adapting approaches from machine learning, computer vision, and artificial intelligence (various neural network models) to scale up our ability to study this trade. We are looking at a number of social media platforms and marketplaces.

Building on our previous research, can we marry these insights from machine learning and computer vision, to those generated from text analysis of the posts, and social network analysis of followers? How do particular patterns of display move over the network of participants - are there fads, trends, key players? Finally, what are the ethical, moral, and legal implications of using machine learning in this way?

Our objectives are therefore:

1. To develop and share a trained neural network that can be employed by other researchers interested in this trade in particular;
2. To develop the computational and theoretical tools to allow others to adapt our approach to their own area of interest in humanities’ research;
3. To determine the patterns in the visual rhetorics of the trade in human remains online so that this trade can be tracked across social media, monitored, and disrupted;
4. To enable the possibility of sourcing these materials so that they may be repatriated to descendent communities;
5. To build ethical frameworks into our computational approaches;
6. To develop a cohort of highly trained personnel who will take this research forward into other domains.