In rule-based expert systems, what can the explanation facility be used for in practice when supporting users and developers?

Introduction / Context:
One distinguishing feature of symbolic expert systems is their transparency. The explanation facility traces which rules fired and why, allowing users to understand recommendations and enabling knowledge engineers to refine rule sets. This capability is central to trust, adoption, and maintenance of expert systems in domains like healthcare and finance.

Given Data / Assumptions:

• The facility can show how conclusions were reached.
• It can assist developers in finding errors or gaps in rules.
• ‘‘Construct a diagnostic model’’ is broader than explanation and typically precedes rule authoring.

Concept / Approach:

Explanation facilities generally support two key activities: they help end users by articulating the reasoning chain behind a conclusion (why and how), and they aid developers in debugging by exposing the rule firings and data dependencies. Building a diagnostic model, however, is a knowledge engineering task outside the explanation feature's remit; it is performed during system design, not as a function of explanation output.

Step-by-Step Solution:

Verification / Alternative check:

Documentation of classic systems (e.g., MYCIN) describes ‘‘why’’ and ‘‘how’’ explanations and their use during rule refinement, confirming (b) and (c) as core functions.

Why Other Options Are Wrong:

Construct a diagnostic model: That is a design-time activity, not what the explanation facility does.

All of the above: Overbroad because it includes (a), which is not a typical explanation feature.

None: Incorrect because (b) and (c) are valid uses.

Common Pitfalls:

Assuming explanation automatically creates new diagnostic structures; it clarifies reasoning but does not author knowledge.

Final Answer:

both (b) and (c)