A Meta-Model for Predicting the Quality of Knowledge Elicitation Sessions
64th ISI World Statistics Congress - Ottawa, Canada
Format: CPS Abstract
Keywords: bayesian, elicitation, modeling
Session: CPS 18 - Statistical modelling I
Monday 17 July 4 p.m. - 5:25 p.m. (Canada/Eastern)
Capitalizing on expert knowledge can be useful for a company. It can be for transmitting all the know-how on a given field, incorporating technical aspects for decision making, or building causal models for doing predictions. This knowledge can be represented through a Bayesian Network  to introduce uncertainty on the phenomenon, and, combined with Data, its performance can be improved. Elicitation is done thanks to sessions where experts works together to build models with a facilitator and a modeler. It is asked the experts to be available for a given amount of time, which can be large (several days) and with a risk that at the end of the sessions, they will not be able to have a satisfying tool. In the context of multi-project management, we propose a tool to assess the probability of success of Elicitation sessions on a given problem. This tool is obtained thanks to the Elicitation of a Bayesian Network  (meta-model), quantified with prior distributions.
After a brief description of the context (elicitation modeling prioritization), we explain how we were been able to elicit the meta-model.
First, a theoretical justification of the Bayesian network structure & nodes combination functions is done. Then, we detail the different dimensions and the output, followed by a demo of the app that has been generated with the model.
 Probabilistic Graphical Models: Principles and Techniques. D. Koller, and N. Friedman. Adaptive computation and machine learning MIT Press, (2009 )