Abstract
Consideration of the user's preferences when generating a presentation of a simulation results in a concise and expressive explanation tailored to the user. The approach consists of (1) generating a simulation, (2) ascertaining the user's presentation utility, (3) applying linear programming to optimize the utility function, and (4) generating a presentation based on the optimized presentation utility function. This research investigates the application of multiattribute utilities to presentations of simulation results. Presentations of simulations divide naturally into two types: presentations of the model (structural descriptions) and presentations of the results (behavioral descriptions). In particular this research illustrates the use of user's preferences for completeness, complexity, and effectiveness when presenting the simulation results of a four compartment glucose model. Considering user's preferences of three classes, novices, researchers, and clinicians, illustrates a possible range of simulation presentations.
Full text
PDF
Selected References
These references are in PubMed. This may not be the complete list of references from this article.
- Cobelli C., Mari A. Validation of mathematical models of complex endocrine-metabolic systems. A case study on a model of glucose regulation. Med Biol Eng Comput. 1983 Jul;21(4):390–399. doi: 10.1007/BF02442625. [DOI] [PubMed] [Google Scholar]
- Langlotz C. P., Fagan L. M., Tu S. W., Sikic B. I., Shortliffe E. H. A therapy planning architecture that combines decision theory and artificial intelligence techniques. Comput Biomed Res. 1987 Jun;20(3):279–303. doi: 10.1016/0010-4809(87)90059-0. [DOI] [PubMed] [Google Scholar]
- Wallis J. W., Shortliffe E. H. Explanatory power for medical expert systems: studies in the representation of causal relationships for clinical consultations. Methods Inf Med. 1982 Jul;21(3):127–136. [PubMed] [Google Scholar]