Search ISU

 

Plains Typology Conference
Conference Overview

This project will create a projectile point typology for the Northwestern Plains of North America by eliciting knowledge from expert archaeologists through a computer-based anonymous collaboration and decision tool, as those experts interact with each other and with an artificial intelligence (AI) that is an automated expert classification system.

Archaeologists who study the prehistory of the Northwestern Plains will attend a 3-day conference at the Simplot Decision Support Center (SDSC) at Idaho State University. During six discrete sessions, the archaeologists will interact with each and with the automated expert system through a knowledge elicitation protocol. The goal of the sessions is to use the decision tool of the SDSC to support the knowledge elicitation exercise and to create the first definitive and validated classification of projectile points from the Northwestern Plains. Twelve archaeologists will be invited to attend, representing senior and junior expert knowledge.

The project has several significant outcomes:

  • The project will create a new typology for projectile points from the Northwestern Plains.
  • This typology will result from the consensus of human experts, from the automatic classification of projectile points by the automated expert system (known as “SIGGI”), and from the interaction between the human experts and automated expert system.
  • The project will pilot the use of knowledge elicitation protocols in both expert-expert and expert-SIGGI interactions.
  • The project will pilot the use of knowledge elicitation protocols in an archaeological context.
  • The project will use the Simplot Decision Support Center (SDSC) for the knowledge elicitation exercises.

Principal Researchers

  • Skip Lohse
    • Professor, Department of Anthropology
    • Phone: 208-282-5189
    • Email: lohserne@isu.edu
  • Kandi Turley-Ames
  • Corey Schou
    • Director, Informatics Research Institute
  • Jim Frost
    • Research Assistant Professor of Informatics
  • Al Strickland
    • Professor, Department of Instructional Methods & Technology
  • D. Sammons
    • Associate Professor, Department of Instructional Methods & Technology
    • Phone: 208-282-2569
    • Email: sammdott@isu.edu

Plains Typology Conference Staff

Review of Literature

Three areas of research provide the background for the Plains Typology Conference. These include autoclassification of projectile points through use of a neural network, study of how expert knowledge is structured, and models of expert decision making and consensus.

  • The autoclassification system central to the Plains Typology Conference is an expert classification system that we call “SIGGI.” SIGGI is so-named because, as a neural network, it employs a Sigmoid function for backpropagation of classification protocols. SIGGI has been previously described in a number of conference presentations and publications, which can be found here .
  • The study of expert knowledge is a relatively underdeveloped field of psychology. The Plains Typology Conference will provide an unprecedented data set of expert knowledge in a tightly defined field. To observe twelve experts in the singular field of Plains projectile point classification provides a unique opportunity to make judgements about how expert knowledge is structured, shared, and articulated. A set of publications on expert knowledge systems can be found here.
  • The Plains projectile experts involved in this project will review and discuss projectile point classification in the Simplot Decision Support Center . This computer-based system allows for anonymous discussion, evaluation, and rating of proposed projectile point types. Research on both the Delphi and nominal group evaluation processes are used to create a new synthetic model of expert evaluation; previous research on the Delphi and nominal group models may be found here .