Network analysis in the humanities differs from the use of traditional neural networks (ANNs) in artificial intelligence and cognitive science research. The former approach allows data to determine the structure and weighting of a network rather than adjusting weights in a predefined network structure as is done in the latter. However, using additional network analysis techniques we will show that the same networks used in humanities research can be transformed into predictive, goal-pursuing mechanisms, thereby crossing the gap from social and textual analysis to artificial intelligence. The rationale for taking such an approach is to show that content-driven artificial intelligence can, in time, be used to find its own patterns in text-based data which will drive automated analysis for the Noesis project (http://noesis.evansville.edu) and other related projects. Though our presentation will be very preliminary, we hope to provide an existence proof that such an approach is viable. The target application for the project of which this is a part is automated semantic analysis of the Stanford Encyclopedia of Philosophy in order to isolate "semantic signatures" that can help to map the profession of philosophy insofar as it is represented online.
Chris Harrison (University of Evansville), Co-Author
Copyright. All Rights Reserved.