Identification of Question Types Based on Theories of Reasoning: Informing the Expert Knowledge Capture Process, 07-R9863
Printer Friendly VersionPrincipal Investigators
Sean Milton Duncan
Doretta E. Gordon
Jenifer Wheeler
Inclusive Dates: 10/01/08 – 09/30/09
Background - Because of a range of factors, from an aging workforce to a need for specialized knowledge, corporations are losing their expertise at a growing pace. Because of this loss, organizations must find efficient, but accurate, methods for capturing their expert knowledge before it is too late. While traditional knowledge collection methods can provide valuable information, they involve one-on-one, live interviews and are very time-consuming.
SwRI has been systematically capturing experts' tacit knowledge since 2003. The process currently used to capture expert knowledge involves conducting a series of one-to-one interviews. These interviews yield highly valuable information, but are very time consuming. Because experts in an organization are often the exemplary performers, their time on task is very valuable.
Approach - In this study, SwRI sought to lay the groundwork for creating a semi-automated questioning tool, which "asks the next best questions" for gathering an expert's most valuable knowledge in the most expeditious manner.
This study addressed the challenges of expert knowledge collection by developing a more efficient, yet accurate, semi-automated method for gathering expert knowledge from subject matter experts. Specifically, the researchers in this study developed two methodologies to analyze and replicate expert interviewers' questioning behavior. First, a methodology was developed to systematically examine interviewers' question generation and question selection behavior. This resulted in a coding schema for transcripts and generic questions based on the coding schema. Second, a methodology was developed to use the coding schema and generic question framework in a semi-live evaluation. The semi-live evaluation served as a successful proof-of-concept for the coding schema, generic question framework, and underlying methodologies.
Accomplishments - The research team was able to create the foundation for a semi-automated expert knowledge capture technique that, with further development, could significantly ease the resource intensive nature of traditional techniques.
