Expert system development

The testing phase is important in expert system development. The practical applicability of the expert system will largely depend on this phase. Testing expert systems is different from normal software engineering in a number of ways. First, it is difficult to test exhaustively the full code and all possible paths the reasoning process may follow. Secondly, the nature of expert systems poses some typical problems. Due to their heuristic nature the correctness of the results cannot be easily verified. A certain degree of errors may be acceptable and, moreover, an... 

Concerted Organic Analysis of Materials and Expert-System Development... 

General. We have studied the characterization of multicomponent materials by combining modem analytical instrumentation with a commercially available AI expert system development tool. Information generated from selected analytical databases may be accessed using TIMM, ( The Intelligent Machine Model, ) available from General Research Corp., McLean, VA. This Fortran expert system shell has enabled development of EXMAT, a heuristically-1inked network of expert systems for materials analysis. 

Additionally, use of a commercial AI shell for expert system development has been demonstrated without the need to learn computer programming languages (C, Pascal, LISP or any of its variations), nor to have an intermediary knowledge engineer. Although this development effort of 4-5 man months was on a minicomputer, adaptation of EXMAT to the microcomputer version of TIMM is anticipated. The completed implementation of EXMAT will support the belief that AI combined with intelligent instrumentation can have a major impact on future analytical problem-solving. 

Fig. 6 Structure of the expert system developed by Rowe for the formulation of tablets. (From Ref...

HazardExpert was the first computer-based toxicity prediction expert system developed by CompuDrug Chemistry Ltd. [25] in 1985. HazardExpert predicts a range of toxicity endpoints including irritation, neurotoxicity, immunotoxic-ity, teratogenicity, mutagenicity, and carcinogenicity. The KB of HazardEx-... 

Many expert systems contain a knowledge base in the form of a decision tree that is constructed from a series of decision nodes connected by branches. For instance, in expert systems developed for the interpretation of vibrational spectra, decision trees are typically used in a sequential manner. Similar to the interpretation of a spectrum... 

Expert systems developed with CLIPS may be executed as follows ... 

Computer Associates has a series of products in the market that support expert system development and knowledge management. A remainder of the original Aion development is CleverPath Aion Business Rules Expert (BRE), which is an advanced... 

XpertRule is a series of Windows-based expert system development tools, which use genetic algorithms for optimization. 

SHAMAN is an expert system developed for qualitative and quantitative radionuclide identification in gamma spectrometry. 

Chemical Equilibrium with Applications (CEA) is an expert system developed by NASA for determining compositions in chemical equilibria for deriving thermodynamic properties of a chemical system in propulsion jet engines. 

Green Chemistry Expert System (GCES) is an expert system developed within the scope of the EPA Green Chemistry program. It includes modules for SMART, reaction databases, chemicals design, solvents database, and literature references for green chemistry. 

MOLGEN is an expert system developed to model the experimental design activity of scientists in molecnlar genetics. 

TEXSYS is an expert system developed by NASA to monitor conditions and determine malfnnctions in thermal bus components of propnlsion jet engines. 

Features real-world examples of recent applications, such as expert systems developed for NASA... 

The first chapter presents an overview of the state of the art of environmental expert system development. The chapter describes the system platforms, languages, and trends in system development It is fol-... 

Together, these chapters present an important summary of the majority of the current work in environmental expert system development and represent most of the efforts actually being commercialized. This is a rapidly evolving field, but one with significant paybacks in terms of providing environmental problem solutions to end users faster and with less hassle and expense. 

These differences must be considered in choosing an approach for developing a new expert system and in selecting an expert system development tool. Forward chaining is preferred for identifying options while backward chaining is preferred for identifying whether specific options are viable. 

Table IV shows the most commonly used shells for environmental expert system development, and the number of systems developed in each.

Second, expert system development projects must be reasonably scoped from initiation stage. Failure to do so can lead to developing a solution to the wrong problem, tackling overly complex problems, or attempting to solve nebulous problems. 

Seventh, the investment in an expert system application must produce a payoff either in terms of improvement productivity or a measurable profit. Since expert system development requires a significant investment in terms of people and money, the expected return on that investment must be well understood, along with the means of measuring the return. 

This chapter provides an overview of expert-system verification and validation (V V) techniques. Several methods are presented. First, many of the conventional software V V techniques such as requirements analysis and unit testing can be applied to expert-system development. Second, an expert-system developer can use automated tools to test rule consistency and structure. A more viable alternative, however, is for the developer to create his own set of consistency and completeness tests. Finally, a developer should rely on qualitative judgment to determine the validity of a knowledge base. This judgment could include expert opinion as well as specialized tests designed to determine knowledge-base certification. The chapter suggests that methods should be combined into an optimal mix in order to best undertake V V. 

Validation involves determining that the system performs with a reasonable level of accuracy. Validation is accomplished through test and evaluation of ES software and integrated hardware. Validation thus ensures that the capabilities that have been specified in the ES requirements have been exercised and meet levels acceptable to the user. Thus, without a true V V methodology, much time is lost in the evolutionary expert system development process. 

---