Source-grounded · four indexes · one book
Publication Intelligence Index Generator scored 9.62 points higher in this four-way study.
The Oxford History of the French Revolution (2002). Four indexes evaluated against the same 425 body-text pages, within their documented scope. Each was assessed without using the findings to repair it.
Publication Intelligence’s 98.05/100 result is recorded in report v13 under the V10 evaluation rubric. It is a weighted quality score for this book, not an accuracy rate. The supporting evidence is open to inspection, including two unsupported locators flagged for editorial review.
Products used in this study
Index Generator · Subject Index
We used only Index Generator’s Subject Index to produce our candidate.
Explore Subject Index generationReview index entries against source evidence
- Index Check
- Index Audit
- Index Repair
Evaluate Subject Index
We evaluated all four indexes with Evaluate Subject Index, our free, open-source precursor to Index Score.
Evaluate Subject Index SkillProducts NOT used in this study
Index Score
We did not use Index Score in this study. It is not included in Generator. It produces reports equivalent in structure to the evaluations shown in our full studies.
Index Repair
Index Repair also addresses actionable findings from Index Score.
To keep the comparison fair, we evaluated our candidate without using the evaluation findings to improve it.
Why we evaluated this book
We chose The Oxford History of the French Revolution because both IndexerLabs and Indexia had already published indexes for it. We retrieved their candidate indexes from their own public sites; Publication Intelligence Index Generator did not create or edit them.
How we evaluated these indexes
We evaluated each candidate within its documented scope against the same source edition, benchmark, scoring policy, and publication gates. We did not use the findings to repair the indexes.
- Index → source
- Are headings and locators supported by the cited pages, factually accurate, and faithful to the book’s meaning and stance?
- Source → index
- Can readers answer the expected research questions? Are the book’s important concepts, arguments, people, and treatments covered and findable?
- Whole index
- Does the index work as a coherent whole, with useful structure, selective coverage, valid cross-references, consistent mechanics, and efficient navigation?
What the scores show
Under this study’s rubric, Publication Intelligence Index Generator scored above the other three exact candidates.
For these candidate versions and this source edition, Publication Intelligence Index Generator combined a higher retrieval score with a smaller measured repair burden.
+9.62
Vs. professional benchmark
Ahead of the professional index
Publication Intelligence Index Generator leads the professional index in four dimensions and ties it in two:
- Coverage: +9.4 points
- Selectivity: +16.4 points
- Page-reference reliability: +19.4 points
- Navigation: +2.2 points
- Fidelity and mechanics: tied at 100%
+19.71
Vs. closest automated rival
Further ahead of IndexerLabs
IndexerLabs reaches 1.2 points more benchmark coverage. Publication Intelligence Index Generator leads in four other dimensions:
- Selectivity: +28.9 points
- Fidelity: +60.0 points
- Page-reference reliability: +19.4 points
- Navigation: +8.9 points
2 vs. 150+
Unsupported locators
A much smaller repair
The unsupported-locator gate hides very different repair loads:
- Publication Intelligence Index Generator: 2 unsupported locators
- Professional index: 150
- IndexerLabs: 151
- Indexia: 701
- IndexerLabs and Indexia also triggered three additional gates each.
Inspect the work
Don’t take our word for it. Dig into the evidence.
Every percentage opens into the judgments and calculations behind it. Inspect every heading, entry, locator, research question, concept, mechanics finding, and density calculation—not just the final score.
Inside the evaluation
Move across the stack or use the arrows →
- 01Excellent
98.05/ 100
Publication Intelligence Index Generator
Inspect the strongest overall result, its two unsupported locators, and the gate that still failed.
Open the full study - 02Strong
88.43/ 100
Published index
Use the professional index as the human baseline, including every locator and repair finding.
Open the full study - 03Mixed
78.34/ 100
IndexerLabs
Trace high coverage alongside lower selectivity, fidelity, and page-reference reliability.
Open the full study - 04Weak
64.33/ 100
Indexia
Review the widest repair load, including 701 unsupported locators and three omitted subjects.
Open the full study
Method, rules, and practical guidance
- Evaluation methodologyThe source-first protocol, rationale, controls, and worked application.
- Subject Index Evaluation StandardThe normative six-dimension rubric, calculations, evidence requirements, and gates.
- Subject Indexing RulesThe editorial rules used to define useful subject access.
- Can AI Evaluate a Book Index?A plain-language introduction to the evaluation problem and approach.
- Should a Book Index Include Every Mention?A practical guide to substantive passages, passing mentions, and useful locators.
Scope: this result applies to the revised 2002 second edition of this book and the exact four candidate versions evaluated. It does not by itself establish performance across all books or indexes.
Disclosure: Publication Intelligence LLC develops and markets Publication Intelligence Index Generator and is responsible for this comparison and its evaluation workflow. Candidate-blind source review and separate pipelines are methodological controls; they do not make the study independent.
Cite this comparison
Publication Intelligence LLC. “Oxford History Subject Index Comparison.” Comparison of four exact candidate indexes for William Doyle’s The Oxford History of the French Revolution, second edition (2002). Consult the linked studies for the evaluated versions, methodology, evidence, and limitations.
https://publicationintelligence.com/comparisons/oxford-history-of-the-french-revolution