Entity Resolution Patents: Leaders, Trends & White Space 2026
Filing growth compares 2021 (1,439 records) with 2024 (1,233) — a three-year span. 2024 is the most recent year we treat as complete: publication lags filing by roughly 18 months, so 2025 onwards are still filling in and any growth rate that ends there would understate the field. Top-5 share is the combined record count of the five largest assignees divided by all 763,853 records in scope (CR5), not by the ranked leaders only.
What the entity resolution patent record actually shows
Entity resolution — also called entity matching or record linkage — is the problem of deciding whether two data objects from different sources describe the same real-world thing. It sits underneath knowledge graphs, customer data platforms, fraud detection and supply-chain visibility tools, which is why the 763,853 records in scope span far more than one industry vertical. The search captures anything indexed against entity resolution, entity matching, record linkage, or the co-occurrence of “entity” and “resolution” in the title, abstract, claims or description.
That breadth means the ranking is not dominated by a single specialist vendor. Large diversified technology filers occupy the top of the assignee list, but the concentration figures below show a field where most of the claim space still sits outside the leading names — useful context before assuming any one company controls the core methods.
Filing trend, technology composition and where records are filed
Three views of the same 763,853-record dataset: how filing volume has moved year over year, which IPC subclasses carry the work, and where applicants have chosen to file first.
A decade of filing activity, with 2022 as the high point
Annual filings rose from 929 in 2017 to a peak of 1,580 in 2022, then eased to 1,233 by 2024 — a -14% move across that three-year span. 2025 and 2026 figures are still low because publication lags filing by roughly 18 months; they are not evidence of a declining field.
Technology composition across eight leading IPC subclasses
G06F (electric digital data processing) leads at 1.0% of all 763,853 records, with H04L, H04W and G06Q each under 0.7% and G06N, H04N, G06T and G06K trailing further behind. Because a single record can carry several IPC codes, these shares add up to more than 100% — the pattern to read is relative weight, not a partition of the field.
Shares are the percentage of the 763,853 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
Go deeper on Knowledge Graphs & Graph Computing: Entity Resolution Patent Landscape with Eureka
This page is one run against one query. Ask Eureka your own question about knowledge graphs & graph computing: entity resolution patent landscape and every answer comes back with the patent numbers behind it.
Try EurekaThe most-cited prior art and a representative claim
US9996607B2 — Entity resolution between datasets (International Business Machines Corporation, 2018-06-12)
The patent describes a deterministic model for entity resolution: define the entity to be resolved, select two datasets for comparison, define matching predicates for their attributes to select candidate matches, then apply a precedence rule to narrow those candidates to a final matched subset. Running the model applies the matching predicates and precedence rule across the datasets to produce resolved entity pairs.This is one of the clearest deterministic-matching claim sets in the corpus and a useful reference point for how narrow or broad a matching-predicate claim can be drafted.
| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | US20170235848A1 | System and method for fuzzy concept mapping, voting ontology crowd sourcing, and technology prediction | 1,473 |
| 2 | US6131095A | Method of accessing a target entity over a communications network | 1,339 |
| 3 | US9516053B1 | Network security threat detection by user/user-entity behavioral analysis | 1,297 |
| 4 | US20040064351A1 | Increased visibility during order management in a network-based supply chain environment | 1,295 |
| 5 | US20040220926A1 | Personalization services for entities from multiple sources | 1,150 |
| 6 | US7403942B1 | Method and system for processing data records | 960 |
| 7 | US20050010653A1 | Content distribution system for operation over an internetwork including content peering arrangements | 925 |
| 8 | US9286413B1 | Presenting a service-monitoring dashboard using key performance indicators derived from machine data | 793 |
| 9 | US20060253584A1 | Reputation of an entity associated with a content item | 739 |
| 10 | US20150310188A1 | Systems and methods of secure data exchange | 728 |
Citation counts favour older filings simply because they have had longer to accumulate citations inside this corpus — read them as a signal of influence on the field, not as a ranking of current importance.
Each row carries its publication number; clicking a row searches Eureka by that number.
Put your own technology through the same analysis
Eureka on the web
When you want the answer in the next five minutes.
The agent works the prompt against patents and technical literature, citing every source.
Run your analysis now →MCP server & REST API
When it has to run inside your own pipeline.
Patent search, landscape analysis and assignee resolution as MCP tools. Drop them into any agent framework, or call REST directly.
Browse MCP servers →Reading the concentration, the trend and the filing geography together
None of these figures are surprising in isolation. Together they describe a field where a handful of large filers hold a meaningful but partial share, filing activity has plateaued rather than collapsed, and most activity still clears through one office.
Leadership is real but partial
The five leading assignees together account for 94,396 records, 12.4% of the 763,853 in scope; the top 10 add up to 141,502, or 18.5%. That leaves more than four-fifths of the field outside the ranked leaders, spread across the other 90 names in the ranking and the long tail beyond it.
Volume has eased off its 2022 peak
Annual filings rose steadily from 929 in 2017 to 1,580 in 2022, then pulled back to 1,233 by 2024 — a -14% move over that three-year window. That pattern reads as claim space filling in around core matching and linkage methods, not as the underlying technology losing relevance.
Core computing dominates, AI-specific classes trail
G06F (electric digital data processing) is the largest single subclass at 1.0% of all records, ahead of H04L at 0.7%. G06N, the AI-model computing class, sits at just 0.3%, suggesting most entity resolution claims are still framed as data-processing methods rather than as AI-model architectures.
The US is the dominant first-filing venue
Among the receiving offices tracked, the United States accounts for 13,726 records, well ahead of the EPO at 2,443 and WIPO/PCT at 2,235. Australia, Canada and the UK each sit under 600, indicating that strategic filing decisions in this field are still made with the US market as the primary target.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to knowledge graphs & graph computing: entity resolution patent landscape, with the prior art for and against each one.
Where to take this from here
The dataset points to specific next questions rather than a single conclusion. These are the ones worth running down before committing a filing or freedom-to-operate budget.
Map the white space around G06N-classed claims
AI-model-specific entity resolution sits at only 0.3% of records against G06F's 1.0%, a gap worth testing before assuming the deterministic-matching space is fully occupied.
Explore the technology map in EurekaCheck freedom-to-operate against the leading assignees
With 18.5% of records held by the top 10 assignees, a targeted search against their specific claim families is more useful than a landscape-level read alone.
Run a freedom-to-operate search in EurekaTrack filings past the 2024 cut-off with care
2025 and 2026 volumes are still filling in due to the roughly 18-month publication lag; treat any near-term trend read with that in mind.
Set up filing alerts in EurekaCommon questions about entity resolution patents
The dataset behind this landscape covers 763,853 published records filed under entity resolution, entity matching or record linkage terms between 2015 and mid-2026. That figure counts patent families in the assignee ranking, which is the fairer unit than raw document counts because it neutralises repeated continuations and multi-jurisdiction refilings of the same invention. Not every record is an active granted patent — the corpus includes applications at every stage, from filed to granted to abandoned.
The assignee ranking lists 100 companies, led by a top filer with 25,906 records, with the fifth-place holder at 14,204 and the tenth at 7,413. The five leading assignees together hold 94,396 records, 12.4% of the 763,853 records in scope, which means the field is not tightly held by one company even though a clear leader exists. The remaining 87.6% of records are spread across the other ranked names and a long tail of smaller filers not captured in the top 100.
Filings rose from 929 in 2017 to a peak of 1,580 in 2022, then eased to 1,233 by 2024, a -14% change over that three-year span. That is a plateau after a period of growth, not a collapse, and it likely reflects core matching and linkage methods becoming more thoroughly claimed rather than declining commercial interest. Figures for 2025 and 2026 understate real activity because patent publication typically lags filing by about 18 months, so recent years always look artificially low in any trend chart.
Electric digital data processing (IPC class G06F) is the largest single technology class, covering 1.0% of the 763,853 records in scope, followed by digital information transmission (H04L) at 0.7% and wireless communication networks (H04W) at 0.4%. Business, commerce and administrative data processing (G06Q) and AI-model computing (G06N) trail further behind at 0.4% and 0.3% respectively. Because a record can carry more than one IPC class, these percentages do not sum to 100%; they show relative weight across overlapping technology areas rather than a clean partition of the field.
US9996607B2, assigned to International Business Machines Corporation and granted in 2018, is a useful reference point: it claims a deterministic model that defines an entity to be resolved, selects two datasets for comparison, applies matching predicates to the datasets' attributes to select candidate matches, and then applies a precedence rule to narrow those candidates to a final matched subset. That structure — matching predicates followed by a precedence rule — is a common pattern across deterministic entity resolution claims in this corpus, and it is worth checking any new filing against it for overlap. Probabilistic or machine-learning-based matching approaches, by contrast, are less represented in the highest-cited records, which points to where some newer claim strategies may still find open ground.
Research Knowledge Graphs & Graph Computing: Entity Resolution Patent Landscape in depth with Eureka
Go past this page: query the whole knowledge graphs & graph computing: entity resolution patent landscape corpus yourself, in your own scope.
Every answer comes back with patent numbers you can open.
Disclaimer. This page is generated from Patsnap Eureka data drawn from a limited snapshot of global patent and scientific-literature records, and is provided for general information and reference only.
Patent data carries inherent limitations: recent filings (typically the most recent 18–24 months) are under-counted due to standard publication lag; counts may be reported at either a patent-family or a patent-record basis and are not always directly comparable; classification, applicant-name, and citation data may contain errors, duplicates, or omissions; and the underlying search query defines and constrains the scope shown. As a result, the analysis may be incomplete or inaccurate and may not reflect the full technology landscape.
Nothing on this page constitutes an exhaustive prior-art, novelty, freedom-to-operate, or validity search, nor does it constitute legal, financial, investment, or professional advice, and it should not be relied upon as such. Any patent, commercial, or strategic decision should be verified independently and reviewed with qualified patent, legal, and domain professionals. Patsnap makes no warranties, express or implied, as to the accuracy, completeness, or fitness for any particular purpose of the information presented.
Machine translation. Assignee and organisation names originally recorded in Chinese, Japanese or Korean have been rendered into English by an AI translation step so that the tables stay readable. These renderings are best-effort and may not match a company’s registered English name; the original name is what the underlying patent record carries, and it is what any Eureka query launched from this page uses.