Exam · Knowledge graphs and structured retrieval

Take the exam and earn your certificate.

A real, server-scored exam for this course. Score strictly above 80% to pass and get a genuine, verifiable certificate.

Before you start

We ask for your name and email first, the same as a PDF download, so we know you are here and your certificate can carry the right name.

We privately record your name, email and request IP for this exam request. Read the privacy notice and retention policy.

Visit Mickai on LinkedIn

Complete the security check before submitting.

Cloudflare Turnstile helps prevent automated submissions. How this check uses data.

You can retake the exam any time. Each attempt draws a fresh set of questions.

Written practice, ungraded

Write your response in your own notes, then compare it with the sample answer. This practice is separate from the scored exam and does not affect your certificate.

Practice 1

Write an adjacency table of outgoing edges for the five Atlas fixture nodes, mark the dependency cycle and the isolated node, and state the expected one-step and two-step answers from Atlas. Say which parts are recorded edges and which are computed paths.

Compare with a sample answer

Atlas: e1 to Parser. Parser: e2 to Index and e4 to North. Index: e3 to Parser and e5 to North. North has no outgoing edges, and Archive has no recorded edges at all. The cycle is e2 and e3 between Parser and Index. One step from Atlas reaches Parser; two steps reach Parser, Index and North, and Archive is absent. Edges e1 to e5 are recorded facts; the two-step reachability is a computed path, and the cycle does not create a recorded fact that Atlas directly depends on Index.

A good answer reads every edge in its recorded direction and separates recorded edges from computed reachability.

Practice 2

Using a copy of the fixture, explain what validation does when e1's object is changed to north, and when e1's source is changed to an unknown identifier. Then give one false statement that would still pass the schema and explain what that shows.

Compare with a sample answer

Changing e1's object to north fails the endpoint-kind rule, because uses must go from a project to a package and North is a team. Changing e1's source to an unknown identifier fails the known-source rule. A uses edge from atlas to archive citing s1 would pass the structural schema even if the manifest never listed Archive. Structural validation therefore cannot replace review of the document, version and passage.

A good answer names the endpoint-kind and known-source rules and distinguishes structural validity from factual support.

Practice 3

Describe what the Atlas query with all three relations returns at max_hops 0, 1 and 2, what happens at two hops with max_edges 2 and 3, and why a query starting at north returns nothing.

Compare with a sample answer

Zero hops returns no rows, one hop returns Parser and two hops return Parser, Index and North. At two hops the traversal must examine e1, e2 and e4, so a budget of two raises an error and a budget of three succeeds. North has no outgoing edges, so a query starting there returns no rows; following an incoming maintained_by edge would be a different query. A budget error should be reported as an incomplete query, not as evidence that no relationship exists.

A good answer derives each result from the fixture's outgoing edges and treats a budget error as incomplete rather than empty.

Practice 4

Rewrite the claim 'North owns Atlas, and Atlas has no connection with Archive' so that every clause is supported by the fixture records and the query limits.

Compare with a sample answer

Owns is not a permitted relation, and there is no direct edge between North and Atlas. A supported version: in the fictional snapshot, the maintenance register names North as maintainer of Parser (e4, s3), and the Atlas manifest lists Parser as a package Atlas uses (e1, s1). For Archive: this query found no outgoing path from Atlas to Archive in the supplied graph within two hops, which is not proof that no real-world connection exists.

A good answer keeps both recorded relationships, cites both sources and states the graph and hop limits for the missing path.

Practice 5

In a disposable copy, replace sources + [edge["source"]] with sources + ["s1"] and run the learner suite. Explain which expectation fails, why checking node names alone would miss the defect, and what you should record.

Compare with a sample answer

The exact-witness test fails because the Index and North paths need s2 and s3 respectively for their second edge, while the mutated code cites s1 on every hop. The returned node names still look correct, so a test that checks only node names would miss the defect. Restoring the original line restores the pass. The record should include the changed line, the failing result and the restored result, without claiming that one mutation proves every case.

A good answer ties the failure to the second-edge sources and records the mutation, the failure and the restored pass.

Practice 6

Explain how you would combine graph retrieval and text retrieval to answer the Atlas question, including what to do if a fetched passage no longer supports its edge.

Compare with a sample answer

First resolve Atlas to its exact project identifier rather than matching a similar name. Run a bounded graph query, keep its witness edges and source IDs, then fetch only the corresponding reviewed passages from the same snapshot. The graph enforces the relation pattern, while the passages show what each source actually says and can supply detail never stored as an edge. If a passage no longer supports its edge, stop or mark the answer unresolved instead of substituting a similar passage from another revision.

A good answer uses the graph for the relation pattern, the passages for supporting text from the same snapshot, and abstains when they disagree.

About this exam

Is this exam graded?

Yes. Every question is scored on our server, never in your browser, and you need a score strictly above 80% to pass. The questions are a real, reviewed bank for this course, drawn fresh for each attempt.

Is it only multiple choice?

Yes. This is a multiple-choice knowledge assessment, with at least one question from each chapter in every attempt. Written exercises are separate practice and are not automatically scored. Passing does not establish practical professional competence or independently verify identity.

Is there a time limit?

Yes, 45 minutes, shown as a countdown once you start. If time runs out, whatever you have answered is submitted automatically.

Do I need to download the PDF first?

No. This course's full text is on its own page, so you can read it online and come straight here when you feel ready.

What do you do with my name and email?

We privately record your name, email, request IP, course and privacy acknowledgement for this request. Newsletter consent is separate and optional. See /privacy.html for storage and retention. A passed exam creates a server-stored named certificate. An email confirmation is not identity assurance.

Can I retake it if I fail?

Yes, straight away. Each attempt draws a fresh set of questions and shuffles the answer order, so it will not be identical to your last attempt, and an attempt can only be submitted once.

Is the certificate real?

It is a genuine, server-issued certificate with its own verification page and a unique ID, not a self-declared one. It records that you passed this course's exam on this date. It is not an accredited qualification, a professional licence or proof of employment.