Entity Coverage Analyzer

Free, no signup. Find likely entities in your page text, then reconcile selected names against Google’s Knowledge Graph and Wikidata. Local extraction works when the AI path is unavailable.

Built by Patrick Stox

Runs entirely in your browser — nothing you paste is uploaded or stored. The optional AI extraction sends your pasted text to the bounded model endpoint. Reconciliation sends only a selected entity name to Google KG and Wikidata; local extraction stays in your browser. Anonymous run-level outcome counters may be used for aggregate research; URLs, domains, IPs, and identifiers are never included, and no statistic is released below 100 runs.

Feedback

Sample report Deterministic local example

For “Cloudflare operates Workers AI. Cloudflare provides services to Acme Analytics.”, local extraction returns:

Cloudflare · 2 mentions · heuristic

Workers AI · 1 mention · heuristic

Acme Analytics · 1 mention · heuristic

No sameAs links appear in the JSON-LD until a user performs a lookup and accepts an identifier. The example intentionally stops before reconciliation because a real match depends on the live service and human identity review.

How to use it

Paste the visible text or HTML from one page.
Choose Extract entities for the local capitalized-name heuristic, or the optional model-assisted extraction.
Review the proposed names and select Reconcile IDs only for entities you intend to check against Google KG and Wikidata.
Inspect ambiguous results before using the generated about and mentions JSON-LD.

Local data

Saved targets, named lists, and recent check summaries remain only in this browser.

Entity JSON-LD

What you get

heuristic — proposed by the local capitalized-name pattern.
observed — parsed from the point-in-time model-assisted response, not independently verified.
linked — a lookup returned a stable identifier; confirm the identity manually.
ambiguous — more than one candidate was returned and the first must not be assumed correct.

How it works

The local path counts capitalized sequences of up to five terms, removes a small stop list, and sorts by mentions. The optional AI path asks the bounded endpoint for candidate names. Google KG and Wikidata lookups are separate and user-triggered. Generated WebPage JSON-LD includes only linked entities: the first under about, subsequent matches under mentions, and the reconciled KG, Wikidata, Wikipedia, and claimed social URLs under sameAs.

Features

Local extraction that works without an API or model.
Optional model-assisted candidate extraction with visible provenance.
Per-entity, on-demand Google KG and Wikidata reconciliation.
Richer sameAs JSON-LD restricted to reconciled entities.
Explicit ambiguity and unavailable-result states.

Limitations

The heuristic misses lowercase brands and common-noun concepts, and can mistake headings or sentence openings for entities. Mention count is not topical importance. Model extraction can vary. The first Knowledge Graph result is not necessarily the intended entity, and the generated ordering is not a semantic judgment. This tool does not compare the page against a required topic model or competitors.

Compare existing entity schema

Paste an existing Person or Organization node to find missing name, URL, description, and identity links after reconciliation.

Frequently asked questions

What is entity coverage?

Here it means the named people, organizations, products, and concepts that appear in the supplied page text. The tool inventories likely names; it does not decide which entities a topic must include.

What is the difference between extraction and reconciliation?

Extraction proposes names found in text. Reconciliation connects a selected name to a stable knowledge-graph identifier. A name can be extracted correctly but reconciled to the wrong real-world entity, so every match needs review.

Does a Knowledge Graph miss mean an entity does not exist?

No. A miss can reflect ambiguity, coverage, quota, or service availability. The tool reports it as unmatched or not evaluated, never as proof of nonexistence.

Can I publish the generated JSON-LD as-is?

Review it first. Only reconciled entities are included, but the top match can still be wrong. Confirm each identity and ensure the entity is genuinely about or mentioned by the page.

Does adding more entities improve AI visibility?

Not automatically. Clear, accurate coverage can reduce ambiguity, but stuffing names or schema that the visible page does not support creates misleading markup rather than authority.

Next stepSchema Markup Validator — verify it with a direct check.

Feature requests for Entity Coverage Analyzer

Upvote what you want most. New ideas can be submitted from the floating Feedback menu; requests appear here once approved, and the most-wanted rise to the top.

Loading…

➕ Request a feature

New requests are reviewed before they appear here.

ツールについて

ページに含まれる可能性の高いエンティティを抽出し、選択した一致を照合して、透明性のあるabout／mentions JSON-LDの下書きを作成します。

機能

ページエンティティの候補抽出
選択した一致の照合と根拠表示
about／mentions構造化データの下書き
入力と選択内容を分けた確認

仕組み

ページのテキストとマークアップから候補を抽出し、選択された一致と未確認項目を分けます。確認済みの選択だけをabout／mentions JSON-LDの下書きへ反映します。

制限事項

候補は入力ページからの推定であり、知識グラフでの認識、検索結果、ランキングを保証しません。
下書きはレビュー用で、実体の同一性や公開時のリッチリザルト掲載を自動承認しません。

よくある質問

抽出されたエンティティは正しいですか？

いいえ。候補です。実体、表記、対象ページとの関連を確認してから使用してください。

JSON-LDを自動公開しますか？

いいえ。下書きを確認してから、対象ページと構造化データの要件に合わせて手動で公開してください。

Googleのナレッジパネルを保証しますか？

いいえ。エンティティ候補と構造化データを整理しますが、Googleの表示や認識は保証しません。