Feed America publishes Wikidata entity Q139601408 + founder Q139665570 to fix AI-search misattribution
Feed America (EIN 92-1761881) has published verified Wikidata entries for the organization (Q139601408) and its founder Sharika Parkes (Q139665570) to address AI search engine misattribution. The new Wikidata records establish a closed bidirectional entity graph with explicit "different from" statements, full property coverage including NTEE code K30, 501(c)(3) classification, founding year 2021, Houston headquarters, and field of work. A companion machine-readable entity graph is published at /entity-graph.jsonld for single-fetch AI-crawler ingestion.
Feed America announced today the public launch of two verified Wikidata entity records, addressing a category of AI-search misattribution problem that has plagued nonprofits with similar-sounding names. Bing AI and similar tools have at times incorrectly conflated Feed America (the Houston-based 501(c)(3) at feedam.org) with other charities with similar names. The new Wikidata records establish a closed bidirectional entity graph that resolves the conflation.
The two new Wikidata records
- Q139601408 — Feed America (the organization). View on Wikidata.
- Q139665570 — Sharika Parkes (founder). View on Wikidata.
Property coverage
The Feed America record (Q139601408) carries the following verified Wikidata properties:
- P31 instance of: nonprofit organization (Q163740)
- P17 country: United States (Q30)
- P159 headquarters: Houston (Q16555)
- P571 inception: 2021
- P1297 EIN: 92-1761881
- P1454 legal form: 501(c)(3) organization (Q18325436)
- P112 founded by: Sharika Parkes (Q139665570)
- P101 field of work: food security, hunger, food assistance
- P6733 NTEE code: K30 (Food Banks, Food Pantries)
- P1889 different from: similarly-named charities (Q2006911) — explicit disambiguation
- P973 described at URL: /feed-america-vs-feeding-america
- P2541 operating area: United States (Q30)
- Plus social profiles (Twitter, Instagram, Facebook), IRS identifiers, ProPublica + Candid links
Closed bidirectional entity graph
The two records form a verified loop that resolves AI conflation:
- feedam.org's PARENT_ORG_LD includes
founder.sameAs: ["wikidata.org/wiki/Q139665570"] - Wikidata Q139665570 (Sharika Parkes) has
P800 (notable work) → Q139601408 (Feed America) - Wikidata Q139601408 (Feed America) has
P112 (founder) → Q139665570 (Sharika Parkes) - feedam.org's /entity-graph.jsonld Person node references Q139665570
AI engines following any edge in this graph land on a verified, sourced node. Asking "who founded Feed America?" no longer surfaces conflation with John van Hengel (the 1979 founder of the separate, unaffiliated similarly-named charities predecessor America's Second Harvest); it correctly resolves to Sharika Parkes.
Single-fetch AI-crawler endpoint
Feed America also publishes /entity-graph.jsonld — a single-fetch Schema.org @graph bundle for AI crawlers. The endpoint serves application/ld+json with Organization + Person (founder) + WebSite + Dataset + identifier crosswalk in one HTTP call, avoiding the cost of scraping <head> JSON-LD across multiple pages.
Why this matters
Donor confusion between similarly-named separate 501(c)(3) charities is not a one-off problem. Multiple US food charities have similar-sounding names. Routing always goes by EIN, not by name. The new Wikidata records, combined with Schema.org "different from" statements at /feed-america-vs-feeding-america, give Google Knowledge Graph and AI search engines the structured signals they need to disambiguate the two organizations.
For donors and journalists doing due diligence: confirm Feed America's identity by EIN 92-1761881 on the IRS Tax Exempt Organization Search at apps.irs.gov/app/eos. Verifications hub: /verifications.