Schemas, Graphs, Ontologies - I

Abstract

Since Google declared, in a 2012 blog post, that they would begin focusing on “things, not strings” there has been increasing talk in the media about the value of semantically enhancing data for reuse and sharing. As vast amounts of unstructured data grow thanks to the resources of media technology companies and governments, there is a push to structure those data to produce semantic interoperability and thus improve understanding. Various information infrastructures like metadata schemas, knowledge graphs, and applied ontologies are experiencing a resurgence as the limitations of statistical algorithms and relational databases become clearer and researchers argue for training data that must be made understandable and structured in such a way as to provide value. Yet, along with increasing our ability to make sense of heterogeneous data through semanticization, there are identifiable problems related to data interoperability that may negatively impact people, practices, and places. This panel will explore the idea of schemas, graphs, and ontologies through their use by media technology companies and scientific researchers by asking if these methods are appropriate for modeling all kinds of knowledge, how such modeling affects access and control, what risks of misrepresentation and error exit, and what potential there is of transforming or appropriating knowledge. This panel includes empirically grounded papers and novel theoretical approaches to understanding how semantically constraining technologies are operationalized in practices across domains, including scientific, governmental, and business contexts. Papers that focus on single case studies, introduce new methods, or propose ethical and policy frameworks are included.

Chair

• Amelia Acker, The University of Texas at Austin

Discussant

• Alexander Monea, George Mason University

Individual Presentations

• Discovering Traces of an Archived Component of Google’s Knowledge Graph from the Freebase Data Dumps - Niel Chah, University of Toronto

• Anyone Can Say Anything About Any Organism: Exploring the Implications of Taxonomy's Move to the Semantic Web - Andrea Thomer, University of Michigan School of Information

• The Image of Domains - David Ribes, Universty of Washington

• Semantic Media: Critical Questions about Metadata and Social Ontology - Andrew Iliadis, Temple University

• From the Social Sciences to Data Mining: Data Journalism’s Sense-Making of Expanding Data - Bernat Ivancsics, Columbia University

Session Organizer

• Andrew Iliadis, Temple University