Modeling Event-Based Historical Narratives: A Conversation Between Digital Humanists, Information Scientists and Computer Scientists

Modeling Event-Based Historical Narratives: A Conversation Between Digital Humanists, Information Scientists and Computer Scientists

Meeks, Elijah, Stanford University, emeeks@stanford.edu
Mostern, Ruth , University of California, Merced, rmostern@ucmerced.edu
Grossner, Karl, University of California, Santa Barbara, karlg@geog.ucsb.edu
Shaw, Ryan, University of North Carolina at Chapel Hill, ryanshaw@unc.edu
Jain, Ramesh, University of California, Irvine, jain@ics.uci.edu
Kantabutra, Vitit, Idaho State University, vkantabu@coe.isu.edu

As digital humanities scholars increase the technological sophistication of their methodologies and tool use to represent historical change, their work becomes more theoretically complex from an information and computer science perspective. While much DH scholarship can be performed using off-the-shelf tools and simple data models, attempts to model highly nuanced historical knowledge, especially in event-based representation, requires a greater theoretical investment. This panel will present humanities scholars whose work focuses on using graph and semi-graph data models that track historical change as events in different parts of the digital scholarly media ecosystem—from the theoretical layout of the data and the representation of historical fact to the creation and population of a data structure to the analysis, representation and interaction with the product—in a dialectic with computer scientists and information scientists who have grappled with the problems of designing robust models and implementing performant technology in this domain. Through analysis of the digital scholarly media process from theory to creation to review, we hope to reveal sophisticated practical and theoretical solutions for representing, creating and reviewing historical knowledge.

The panel will be chaired by Elijah Meeks and allow for a short presentation by each of the discussants on the topics described below. This will be followed by a roundtable-style discussion with media examples from each discussant's work cued up so that, if the conversation focuses on one of these works, it can then be demonstrated and examined in detail. The guiding themes of the panel are theoretical challenges of event-based historical models, practical issues in creating sophisticated digital scholarly media using such models, and methods to review the models and the media created from them.

The panel presents a continuum of scholars from the humanities through information science to computer science, with Elijah Meeks (Digital Humanities Specialist, Stanford University) and Ruth Mostern (Associate Professor of History, University of California, Merced) representing the traditional humanities scholars. Elijah Meeks will speak about his work transitioning the Mapping the Republic of Letters database into a pure graph model and the challenges and benefits that this representation of historical knowledge presents in contrast to traditional relational databases as well as the suitability of graph data for event modeling. Ruth Mostern will focus on the role of event modeling in her current Yellow River research, tracking the linkages between historical environmental change and socio-political change, with an emphasis on the temporal and spatial reconciliation of differing scales and emphases of data, specifically comparing historical data collected for environmental science and traditional textual sources for tracking political and social change over time and space.

Karl Grossner (Geography, University of California, Santa Barbara) and Ryan Shaw (Assistant Professor of Information and Library Science, University of North Carolina at Chapel Hill) are both information scientists whose work has focused on issues critical to historical narrative modeling: ontologies and periodization. Ryan Shaw will speak to the complexities of event representation, focusing on periods and events as not only existing in the past, but also produced by discourse about the past, and the need for an approach to representation in which we can move back and forth between treating periods and events as, on the one hand, shared intersubjective points of reference and, on the other, unique perspectives on the past articulated through narrative techniques. Karl Grossner will discuss his work developing a spatial history ontology in which historical processes are modeled as theories of event relations, as well as its implementation in a PostgreSQL spatial database able to support mapping, analytical applications and representations of narrative.

Finally, Ramesh Jain (University of California, Irvine) and Vitit Kantabutra (Idaho State University) will discuss the practical issues of implementing such theoretical constructs in software. Vitit Kantabutra will spend some time explaining his work with Jack Owens fleshing out the Intentionally Linked Entity model, which is a generalized graph data model with a wide range of applications, and is especially suitable for the representation of historical and spatiotemporal events. Ramesh Jain will engage with the processing of event-based data semantically, both through typical searches and by exploring the emergent attributes of the database as an object itself.

By presenting a continuum of academic specialties that are unified in their focus on sophisticated digital representation of historical knowledge, we feel this panel will provide useful practical and theoretical value to the Digital Humanities community. We also hope to add to the growing discussion of how to judge the quality of the media produced in these endeavors, not only from the perspective of humanist scholars but from scholars in the field of information science and computer science.