Handbook for Evaluating Knowledge-Based Systems: Conceptual by Leonard Adelman

By Leonard Adelman

Knowledge-based structures are more and more present in a large number of settings and this instruction manual has been written to fulfill a particular want of their widening use. whereas there were many profitable functions of knowledge-based platforms, a few functions have failed simply because they by no means acquired the corrective suggestions that assessment presents for protecting improvement enthusiastic about the clients' wishes of their genuine operating surroundings. This instruction manual offers a conceptual framework and compendium of tools for appearing reviews of knowledge-based platforms in the course of their improvement. Its concentration is at the clients' and subject material specialists' overview of the usefulness of the approach, and never at the builders' checking out of the adequacy of the programming code. The instruction manual allows evaluators to systematically solution the next varieties of questions: Does the knowledge-based procedure meet the clients' job necessities? Is the approach effortless to take advantage of? Is the data base logically constant? Does it meet the mandatory point of workmanship? Does the method enhance functionality?
The authors have produced a instruction manual that may serve audiences: a software that may be used to create knowledge-based structures (practitioners, builders, and evaluators) and a framework that may stimulate extra examine within the region (academic researchers and students). to complete this, the guide is equipped round a conceptual framework that integrates the differing kinds of reviews into the approach of improvement strategy. the categories of questions that may be spoke back, and the equipment to be had for answering them, will swap in the course of the method improvement existence cycle. And all through this method, one must recognize what could be performed, and what cannot. it truly is this dichotomy that addresses wishes in either the practitioner and educational learn audiences.

Show description

Read Online or Download Handbook for Evaluating Knowledge-Based Systems: Conceptual Framework and Compendium of Methods PDF

Similar nonfiction_7 books

Functional Adaptive Control: An Intelligent Systems Approach

The sector of clever keep an eye on has lately emerged as a reaction to the problem of controlling hugely advanced and unsure nonlinear platforms. It makes an attempt to endow the controller with the foremost houses of version, research­ ing and autonomy. the sector continues to be immature and there exists a large scope for the improvement of recent equipment that increase the most important homes of in­ telligent platforms and increase the functionality within the face of more and more advanced or doubtful stipulations.

Computational Intelligence Systems and Applications: Neuro-Fuzzy and Fuzzy Neural Synergisms (Studies in Fuzziness and Soft Computing)

Conventional synthetic Intelligence (AI) platforms followed symbolic processing as their major paradigm. Symbolic AI structures have proved powerful in dealing with difficulties characterised by way of specified and whole wisdom illustration. regrettably, those structures have little or no energy in facing vague, doubtful and incomplete information and data which considerably give a contribution to the outline of many genuine­ global difficulties, either actual platforms and procedures in addition to mechanisms of determination making.

Additional resources for Handbook for Evaluating Knowledge-Based Systems: Conceptual Framework and Compendium of Methods

Example text

34, p. 3 Operational Prototype Figure 2-4 presents a schematic of the operational prototyping life cycle. Operational prototypes attempt to combine the best of throwaway and evolutionary prototyping approaches. A suitable portion of the system is constructed and incorporates only those features that are well known or understood. This initial version is then deployed to users at various locations. As users operate the system they will discover problems and otTer suggestions for solving them. The prototyper can develop "quick and dirty" features for implementing these improvements.

We will try to identify these research issues throughout the handbook, for one needs to know what can be done and what can't. Our vision throughout is that the handbook be a tool for practioners (developers and evaluators) and a stimulus for researchers and their students. INTRODUCTION 15 Chapter 2 reviews the literature with respect to different (a) development models, (b) evaluation dimensions, and (c) evaluation models. Chapter 3 presents our conceptual framework for evaluating knowledge-based system, and describes each of the five types of evaluation in more detail.

Consequently, we could have easily included a discussion of relevant evaluation dimensions and criteria within the review of each evaluation model. We have separated them to assist readers in (1) formulating evaluation questions and broader evaluation issues and, in turn, (2) selecting evaluation dimensions and criteria for defining them operationally. By grouping the review of dimensions and criteria into one section, we hope that readers will be better able to identiJY those that are most relevant to their evaluation context and needs.

Download PDF sample

Rated 4.40 of 5 – based on 28 votes