Relationships data—information—knowledge

Let us think abont how we can make a first hypothetical approach to what an expert system wonld constitnte. A standard model in the science of knowledge management is the knowledge pyramid, which basically describes the quantitative and logical relationship among data, information, knowledge, and action (please also refer to Fignre 2.1). 

FIGURE 2.1 The relationship among data, information, knowledge, and action is described by the knowledge pyramid. The transition from data to information is performed by calculation, analysis, and interpretation performed by computer software, whereas in the transition from information to knowledge, human interpretation, experience, and intuition play a major role. The final step of reasoning leads to actions to be taken. Amount, context, and patterns can be considered as theoretical factors in contrast to that, real data have to be described by concepts of certainty, probability, and fuzziness. 

Figure 1. The relationship between Data, Information and Knowledge...

The information on rare earth alloys is vast and nearly incomprehensible, and yet we estimated that perhaps only one half of the possible binary alloy systems with the rare earth metals have been studied in any detail at all. And if one considers in addition all possible ternary and higher component alloys our present day knowledge of rare earth alloys is minuscule. To give the reader an idea of the vastness of our knowledge it would take one person working full time twelve years to compile, critically evaluate and write up his/her evaluation of all of the known crystallographic and phase relationship data of the rare earth binary alloys. If one were to do the same with the thermodynamic, magnetic and electrical, and miscellaneous property data of these same alloys one would need another 12 person years to complete this aspect. 

The quality and the relationship of the data to the information and knowledge are crucial points in the learning process... 

A variety of methods have been developed by mathematicians and computer scientists to address this task, which has become known as data mining (see Chapter 9, Section 9.8). Fayyad defined and described the term data mining as the nontrivial extraction of impHcit, previously unknown and potentially useful information from data, or the search for relationships and global patterns that exist in databases [16]. In order to extract information from huge quantities of data and to gain knowledge from this information, the analysis and exploration have to be performed by automatic or semi-automatic methods. Methods applicable for data analysis are presented in Chapter 9. 

Mechanistic Approaches. Adequate and appropriate river-quality assessment must provide predictive information on the possible consequences of water and land development. This requires an understanding of the relevant cause and effect relationships and suitable data to develop predictive models for basin management. This understanding may be achieved through qualitative, semi-quantitative or quantitative approaches. When quantitative or semi-quantitative methods are not available the qualitative approach must be applied. Qualitative assessments involve knowledge of how basin activities may affect river quality. This requires the use of various descriptive methods. An example of this kind of assessment is laboratory evaluation of the extent to which increases in plant nutrients, temperature or flow may lead to accelerated eutrophication with consequent reduction of water quality. 

The information contained in karma s knowledge bases is based upon quantitative structure-activity relationships (QSAR), kinetic data, and structural chemistry. The combination of QSAR and kinetic data allows for the study of enzyme-ligand interactions. The Hansch approach to QSAR, based on a set of congeners, states ... 

Primary model selection should be based on the experimental data observations but other Information may also be useful, such as knowledge of the drug s mechanism of action, results from earlier studies, or concentration-effect relationships of related compounds. The performance of different models can be systematically tested by using a nonlinear regression program, which has readymade routines for common models and also allows the user to formulate his own models. Model selection and validation Is an Important Issue, which Is however beyond the scope of this chapter. 

---