Knowledge management procedures

Validation of expert systems still poses a problem that remains to be solved. This is mainly due to the dynamic nature of the reasoning process and the multiplicity of cases that can be solved by an expert system. Whereas the functionality of a linear software system provides the basis for test cases, an expert system can barely be covered comprehensively by a few hundred test cases. Expert systems, as well as most AI software, have to be transitioned carefully from the implementation to the production phase and require a continuous quality control through experimentation as part of the standard maintenance procedure. However, the same applies to a series of other systems, such as knowledge management systems, data mining systems, and, finally, human experts and is no reason for a general denial of expert systems in productive environments. 

The Hierarchical Approach developed in this chapter incorporates a knowledge-based procedure for the synthesis of separations. This consists of dividing the separation section in three subsystems gas vapour, liquid and solid separations. Each subsystem is further managed by selectors, which makes use of unit operations. Split sequencing is based mainly on heuristics, although may include algorithmic or optimisation techniques. This chapter describe in more detail the synthesis of distillation trains for zeotropic distillations, the non-ideal case being left for the Chapter 9. 

Another way of defining the word change in the context of MOC is to determine if the proposed change affects any of the other elements of process safety. For example, if a new operating procedure has to be written then the MOC process must be followed. Similarly, if the proposed change means that updated P IDs are needed, then the topic of knowledge management has... 

Knowledge management (KM) can be defined as the approaches (methods, procedures, tools, etc.) for handling the registrations (writings) in order... 

Gottschalk, P. (2002). Empirical validation procedure for the knowledge management technology stage model. In Informing Science, Volume 5. Informing Science Institute. 

Other causes of stmctural failure are improper assembly and maintenance. Some failures may result from human error or may be due to designs that are difficult, impractical or impossible to implement. Some may simply result from careless work and poor decisions on the part of workers and management. Sometimes several causes combine and lead to failure. Lack of communication, skill, knowledge, training, procedures, and management commitment can all contribute to faulty workmanship. 

Curation methods are dependent on personal preference, organizational protocols, and storage capability. A wide range of methods and concepts are available and beyond the scope this book. In addition, the organization may have a knowledge manager and/or developed procedures for storage and retrieval of information. If not, use the system you find most efficient for your purposes. 

The initial incident report is very important. This document captures the initiator s firsthand knowledge of what occurred in the moments after the specific event. Example 5-2, Sample Toller Initial Incident Report, is an example of the data that should be documented as soon as possible. Note that it should be modified using the company s management system procedure and incident investigation procedure, which should describe the type of data needed, and level of detail desired. 

---