Virtual Facilitation of Human Group Interaction employing a Learning Classifier System with Crowdsourced Feedback
Department
Computer Science
Major
Computer Science
Research Advisor
Tauritz, Daniel R.
Luechtefeld, Ray
Advisor's Department
Computer Science
Second Advisor's Department
Engineering Management and Systems Engineering
Abstract
Natural human group dynamics sometimes can lead a group down unproductive pathways. An expert group facilitator may need to intervene to return the group to a productive workflow. However, human expert group facilitators are scarce and prohibitively expensive. We can codify the circumstances that lead the group astray into a set of matching rules, with an appropriate intervention for each situation. This proposal is concerned with developing a Virtual Facilitator software system which employs a Learning Classifier System to evolve increasingly higher quality matching rules based on crowdsourced feedback. Such a Virtual Facilitator can replace human expert group facilitators at a fraction of the cost and be ubiquitously available.
Biography
Matthew Nuckolls is scheduled to graduate from Missouri S&T in May 2010 with a B.S. in Computer Science and a minor in Cognitive Neuropsychology. He is an undergraduate researcher in the Natural Computation Laboratory, studying uses of Learning Classifier Systems to solve real-world problems. In addition to his research work, Matthew tutors and grades for the course CmpSc 253 - Algorithms. Prior to his academic career, Matthew served in the US Air Force for 10 years as an Explosive Ordnance Disposal Technician.
Research Category
Research Proposals
Presentation Type
Poster Presentation
Document Type
Poster
Location
Upper Atrium/Hallway
Presentation Date
07 Apr 2010, 9:00 am - 11:45 am
Virtual Facilitation of Human Group Interaction employing a Learning Classifier System with Crowdsourced Feedback
Upper Atrium/Hallway
Natural human group dynamics sometimes can lead a group down unproductive pathways. An expert group facilitator may need to intervene to return the group to a productive workflow. However, human expert group facilitators are scarce and prohibitively expensive. We can codify the circumstances that lead the group astray into a set of matching rules, with an appropriate intervention for each situation. This proposal is concerned with developing a Virtual Facilitator software system which employs a Learning Classifier System to evolve increasingly higher quality matching rules based on crowdsourced feedback. Such a Virtual Facilitator can replace human expert group facilitators at a fraction of the cost and be ubiquitously available.
