Integrating Categorical Legacy Data in Spatial Models: A Unique Dataset from Southeast Asia

Summary

This is an abstract from the "SAA 2024: Individual Abstracts" session, at the 89th annual meeting of the Society for American Archaeology.

Despite advances made in open access publishing, significant archaeological information remains confined to the grey literature or to unpublished, internal reports in the possession of institutions. For at least twenty years, archaeologists have realized that digital archiving could make this material more accessible at a larger scale, but the tools to access these documents and convert them to be a legible part of a comparative dataset are still developing. This poster presents the results of an archival research project using natural language processing to make 159 legacy reports digitally accessible for incorporation in spatial models using generalized linear modeling and machine learning approaches. These data were gathered from archaeological reports generated by the Defense POW/MIA Accounting Agency (DPAA) and its antecedents, in partnership with the center for Digital Antiquity at Arizona State University. Though challenging to implement, we found that context-specific categorical data, such as aircraft type, could be extracted from reports and used to refine generalized linear and machine learning spatial models. The project shows how creative digital archaeology approaches can work with diverse datasets.

This Resource is Part of the Following Collections

• SAA 2024: Individual Abstracts   •
• Society for American Archaeology 89th Annual Meeting, New Orleans, LA (2024)

Keywords

Spatial Coverage

Record Identifiers

Abstract Id(s): 40119.0

File Information