Buildings of Texas

Exploring Linked Data by Mapping Places, Events and People over Time

Abstract

This note describes an ongoing project at the University of Texas Libraries (UTL) that is transforming the way we think about place, people, and events in managing archival collections. This project was developed around a dataset donated by architectural historians, Gerald Moorhead and Mario Sánchez to the Alexander Architectural Archives. This dataset was collected by a team of researchers studying architecturally significant buildings for the two-volume publication, Buildings of Texas. Our team at UT Libraries has used it as a test-bed for geolocating built works in Texas and mapping our Architectural collections. This dataset presented a clear opportunity to develop map-based digital exhibitions and finding aids for archival material, but also posed several challenges due to naming ambiguities, vague building location descriptions, and repeated references to people and architectural firms that were difficult to disambiguate and interconnect. In an attempt to overcome these challenges and develop a set of methods for dealing with similar and related collections, our cross-disciplinary team is evaluating ways to develop a flat spreadsheet into a collection of inter-related datasets. Our goal was a system that could be managed more flexibly, be easily represented through spatial and non-spatial visualizations, and—crucially—contain references to concepts and typologies defined in widely-used ontologies. We have also sought ways to contribute our data as a local authority to both the Getty vocabularies and Wikidata, as a means to broaden representation and allow for multivocality and multiplicity. We have found unique advantages to a team-approach to the cleaning and data normalization process that transformed a single spreadsheet into a graph database. Our experimentation in this process has facilitated our own visualization of the connections between places, people, and events and, in turn, has informed the way we will present this material in our online exhibitions. In addition to the graph database, we explored traditional relational database technologies, which turned out to be much easier to use for mapping the data and managing it with GIS software. Throughout the project, we have been mindful of breaking down traditional modes of archival description, of contributing to but also looking beyond perceived authorities, and about the value of the work we do for architectural artifacts and landscapes, but also any cultural heritage community of practice seeking to meaningfully describe artifacts, events, people, and places over time.