Framework for measuring adaptive knowledge-rich systems performance

Stud Health Technol Inform. 2005:118:313-24.

Abstract

The universe is non repeatable in nature--most of events cannot be prestated and do not repeat themselves. The only way to create systems that are truly useful is to make them adaptive (able to reason by analogy and learn) and rich in knowledge (including common sense knowledge). Adaptive and knowledge-rich health management could get us closer to errorless health care where small incremental adjustments happen all the time preventing occurrence of an error. In the era of adaptive systems we need to have a way to evaluate their performance. Are they truly adaptive? How adaptive are they? Are they accurate enough? Are they fast enough? Are they cost effective? This chapter presents general framework for measuring adaptive knowledge-rich systems' performance and includes among others definitions of adaptiveness factor, britt (a unit of brittleness) and uso-quant (unit of usefulness of a piece of knowledge). Measuring adaptive knowledge-rich systems performance is one of the most important research areas that can have a big pay-off in healthcare now and in the future.

MeSH terms

  • Artificial Intelligence*
  • Automation
  • Biomedical Technology
  • Delivery of Health Care, Integrated / trends*
  • Forecasting
  • Guidelines as Topic
  • Humans
  • Information Management / trends*
  • Medical Informatics / trends*
  • Problem Solving
  • Process Assessment, Health Care*
  • Research