Bahasa

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
Report a bug

Found something broken in Entity Coverage Analyzer? Let us know what happened — this goes straight to a private triage queue, not a public list.

What will be sent 
 No tool inputs, uploads, pasted source, complete results, query parameters, or URL fragments are attached automatically. You can edit or remove the selected passage above. Browser and anti-abuse metadata is processed for spam prevention. 
  
 
  
 
 
  
 
 
  
 
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
  1. Paste the visible text or HTML from one page.
  2. Choose Extract entities for the local capitalized-name heuristic, or the optional model-assisted extraction.
  3. Review the proposed names and select Reconcile IDs only for entities you intend to check against Google KG and Wikidata.
  4. 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.
 
   
 
 
   
 
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.

People also explore
ToolGoogle Knowledge Graph ExplorerToolAI Commerce ValidatorToolRich-Result Eligibility CheckerArticleEntity & Identity SchemaArticleEntity SEOArticleKnowledge Graph SEO
 
 
 
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.
 
     
  
  
 
 
 
 
   
  
Tentang alat
Laporan contoh Deterministik lokal contoh
Bagaimana ini berfungsi lokal jalur counts capitalized sequences dari up untuk five ketentuan, menghapus small hentikan daftar, dan sorts oleh mentions. opsional AI jalur asks dibatasi endpoint untuk candidate menyebut. Google KG dan Wikidata lookups adalah terpisah dan pengguna-triggered. Dihasilkan WebPage JSON-LD mencakup hanya ditautkan entitas: pertama di bawah tentang, subsequent cocok di bawah mentions, dan reconciled KG, Wikidata, Wikipedia, dan diklaim social URLs di bawah sameAs.
Dapat I terbitkan dihasilkan JSON-LD sebagai-adalah?
Fitur
  • Tidak. miss dapat reflect ambiguity, cakupan, quota, atau service availability. alat laporan ini sebagai unmatched atau tidak dievaluasi, tidak pernah sebagai proof dari nonexistence.
  • Batasan heuristic misses lowercase brands dan umum-noun concepts, dan dapat mistake headings atau sentence openings untuk entitas. Mention count tidak topical importance. Model extraction dapat vary. pertama Knowledge Grafik hasil tidak necessarily dimaksud entity, dan dihasilkan ordering tidak semantik judgment. alat tidak bandingkan halaman terhadap wajib topic model atau competitors.
  • Entity Cakupan Analyzer
  • Fitur Lokal extraction yang berfungsi tanpa API atau model. Opsional model-assisted candidate extraction dengan terlihat provenance. Per-entity, pada-demand Google KG dan Wikidata rekonsiliasi. Richer sameAs JSON-LD dibatasi untuk reconciled entitas. Eksplisit ambiguity dan tidak tersedia-hasil keadaan.
  • Bagaimana untuk gunakan ini Tempel terlihat teks atau HTML dari satu halaman. Pilih Mengekstrak entitas untuk lokal capitalized-nama heuristic, atau opsional model-assisted extraction. Tinjau proposed menyebut dan pilih Cocokkan IDs hanya untuk entitas Anda intend untuk periksa terhadap Google KG dan Wikidata. Periksa ambiguous hasil sebelum menggunakan dihasilkan tentang dan mentions JSON-LD.
Cara kerja
Melakukan Knowledge Grafik miss mean entity tidak exist? Di sini ini berarti named people, organizations, produk, dan concepts yang muncul di yang diberikan halaman teks. alat inventories kemungkinan menyebut; ini tidak putuskan yang entitas topic harus termasuk. Apa adalah entity cakupan? Apa Anda dapatkan heuristic — proposed oleh lokal capitalized-nama Pola. teramati — parsed dari titik-di-waktu model-assisted respons, tidak independently terverifikasi. ditautkan — pencarian dikembalikan stable identifier; konfirmasi identitas manually. ambiguous — lebih banyak daripada satu candidate adalah dikembalikan dan pertama harus tidak menjadi assumed benar.
Batasan
  • Tidak ada sameAs tautan muncul di JSON-LD until pengguna melakukan pencarian dan accepts identifier. contoh dengan sengaja berhenti sebelum rekonsiliasi karena nyata cocok depends pada langsung service dan human identitas tinjau.
  • Tinjau ini pertama. Hanya reconciled entitas adalah disertakan, tetapi top cocok dapat masih menjadi salah. Konfirmasi setiap identitas dan pastikan entity adalah genuinely tentang atau mentioned oleh halaman.
  • Entity JSON-LD Capitalized-nama mencocokkan adalah fallback, tidak knowledge grafik. KG miss adalah “tidak dievaluasi” atau unmatched, tidak pernah proof entity tidak exist. Tinjau setiap dihasilkan skema reference sebelum publishing.
  • Mengekstrak terlihat named entitas, cocokkan ones Anda pilih, dan draft transparan JSON-LD.
Pertanyaan umum
Tidak secara otomatis. Jelas, accurate cakupan dapat mengurangi ambiguity, tetapi stuffing menyebut atau skema yang terlihat halaman tidak mendukung membuat menyesatkan markup daripada daripada authority.
Apa adalah perbedaan antara extraction dan rekonsiliasi?
Gratis, tanpa pendaftaran. Temukan kemungkinan entitas di Anda halaman teks, lalu cocokkan dipilih menyebut terhadap Google Knowledge Grafik dan Wikidata. Lokal extraction berfungsi ketika AI jalur tidak tersedia.
Home › Alat › Entity Cakupan Analyzer
Bandingkan existing entity skema Tempel existing Person atau Organization node untuk temukan hilang nama, URL, deskripsi, dan identitas tautan setelah rekonsiliasi. Temukan entity gaps Buka di Skema Generator →
Cloudflare · 2 mentions · heuristic Workers AI · 1 mention · heuristic Acme Analytics · 1 mention · heuristic
Mengekstrak entitas Coba AI extraction
Untuk “Cloudflare operates Workers AI. Cloudflare provides services untuk Acme Analytics.”, lokal extraction mengembalikan: