Fastest
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 →https://www.patsnap.com/resources/blog/rd-blog/knowledge-graphs-and-graph-computing-graph-query-optimization-patent-landscape-patent-landscape/ · Patsnap · data cut-off 2026-07-31 · downloaded from the live page
Patent Landscape · Graph Query OptimizationGraph Query Optimization Patents: Who Files, What They Claim, and Where the Field Is Still Open
50
Published Records
44%
Top-5 Share of All Records
CN
Leading Jurisdiction
41
Active Filers Ranked
Top-5 share is the combined record count of the five largest assignees divided by all 50 records in scope (CR5), not by the ranked leaders only.
Published byPatsnap Research··6 min readSourced from Patsnap Eureka
Overview
What this landscape covers
Graph query optimization sits at the intersection of database systems and knowledge graph infrastructure: query planning, index selection, execution-path tuning and, increasingly, retrieval strategies that feed large language models. This landscape tracks 50 published records filed or published between 2015 and the 2026-07-31 cut-off, drawn from a search string built around graph query optimization and tuning language paired with database, data processing, knowledge graph and information retrieval context terms. Publication lags filing by roughly 18 months, so the most recent filing years understate real activity.
The scope spans classic relational and graph-database query planning as well as newer knowledge-graph-driven retrieval work, including schema-aware query generation for generative AI pipelines. Records come from a mix of enterprise software vendors, research institutes and universities, with receiving-office activity concentrated in China.
Filing activity and technology mix, 2015-2026
Source: Patsnap Eureka. Assignee ranking and totals. Derived from a Patsnap search on Knowledge Graphs & Graph Computing: Graph Query Optimization Patent Landscape covering 2015–2026, data cut-off 2026-07-31. Counts reflect published records only and shift as new filings publish.Run this in Eureka MCP
The numbers
Filing trend and technology composition
Two views of the same 50-record set: how filings have moved year over year, and which IPC subclasses carry the claim language.
Filing trend, 2017-2026
Recorded filings rose from 2 in 2017 to a peak of 15 in 2024, before the count for 2026 (2 so far) reflects the partial year and the usual publication lag rather than a slowdown.
IPC subclass composition
G06F (electric digital data processing) covers 94.0% of the 50 records, confirming that graph query optimization is claimed primarily as a data-processing discipline. G06N (AI-based computing) appears in 36.0% of records, a meaningful but secondary share, while G06Q, H04L, G06T and G06V each cover a small slice tied to specific applications such as business process data, network transmission or image-linked graph data.
Shares are the percentage of the 50 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
Source: Patsnap Eureka. Filing trend and technology composition. Derived from a Patsnap search on Knowledge Graphs & Graph Computing: Graph Query Optimization Patent Landscape covering 2015–2026, data cut-off 2026-07-31. Counts reflect published records only and shift as new filings publish.
Go deeper on Knowledge Graphs & Graph Computing: Graph Query Optimization Patent Landscape with Eureka
This page is one run against one query. Ask Eureka your own question about knowledge graphs & graph computing: graph query optimization patent landscape and every answer comes back with the patent numbers behind it.
Try EurekaKey filings
Representative and most-cited records
Representative filing
Knowledge graph query optimization for retrieval augmented generation
MICROSOFT TECHNOLOGY LICENSING, LLC
Filed by Microsoft Technology Licensing, this application describes a chat system that identifies entities from a user message, has a first generative language model draft a structured query against a knowledge-graph schema, executes that query, and hands the retrieved entity data to a second generative model to produce a response grounded in the knowledge graph.Published 2026-07-16 — the most recent record in scope, illustrating how graph query optimization claims are now being written directly around retrieval-augmented generation pipelines.
Most-cited records in scope
| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | CN119941237A | 一种基于多维度数据体系的充电站运行维护方法 | 16 |
| 2 | CN118277638A | 企业信息管理方法及系统 | 16 |
| 3 | CN112637263A | 一种多数据中心资源优化提升方法、系统和存储介质 | 10 |
| 4 | CN118469309A | 知识图谱驱动的岩溶区铁路智能选线方法、介质及设备 | 8 |
| 5 | US20180246929A1 | Ontology-based graph query optimization | 8 |
| 6 | CN121210525A | 基于多化智能体协作的图查询处理方法及系统 | 5 |
| 7 | US20230004559A1 | Ontology-based graph query optimization | 5 |
| 8 | CN118760775A | 一种知识图谱查询方法、装置、设备及存储介质 | 4 |
| 9 | CN120216650A | 一种问答优化方法、装置、设备及存储介质 | 3 |
| 10 | KR101731579B1 | Database capable of intergrated query processing and data processing method thereof | 3 |
Citation counts favour older records in the corpus and should be read as a signal of influence, not of current technical importance.
Patent titles are shown in the language they were filed in, not translated, so that each record stays verifiable against the original filing — a translated title will not match in Eureka or in any national register. Each row carries its publication number; clicking a row searches Eureka by that number.
Source: Patsnap Eureka. Citation counts and representative records. Derived from a Patsnap search on Knowledge Graphs & Graph Computing: Graph Query Optimization Patent Landscape covering 2015–2026, data cut-off 2026-07-31. Counts reflect published records only and shift as new filings publish.Run this in Eureka MCP
Run it yourself
Put your own technology through the same analysis
Where to run it
For builders
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 data
What the concentration and geography mean for strategy
The ranking and receiving-office split point to a field with one dominant filer, a long tail, and a filing base weighted heavily toward China.
Concentration
44.0%
of 50 records held by top 5 assignees
One leader, then a long tail
The leading assignee holds 17 of the 50 records outright. The top 5 combined reach 44.0% of all records in scope, and the top 10 reach 54.0% — meaning roughly half the field is split across 31 other ranked assignees plus unranked single filers.
41 companies appear in the assignee ranking overall.
Geography
34 of 50
records route through a China receiving office
Filing activity is China-weighted
China accounts for 34 of the tracked receiving offices, well ahead of the United States (7), WIPO/PCT (4), Europe (2), India (2) and South Korea (1). Freedom-to-operate work that only checks US and EP registers will miss most of the documented activity.
Receiving-office counts, not family counts.
Technology mix
94.0% vs 36.0%
share of records in G06F vs G06N
Claims still frame as data processing, not AI architecture
94.0% of records carry a G06F classification against 36.0% for G06N, so most graph query optimization claims are anchored in core data-processing language even where an AI model sits in the pipeline. That gap is where retrieval-augmented-generation-specific claims, like the featured filing, are starting to push in.
Classes overlap; shares are of the 50-record total, not of each other.
Peak activity
15 filings
in 2024, the peak year so far
Activity built steadily through 2024
Recorded filings grew from 2 in 2017 to a peak of 15 in 2024. The 2026 figure of 2 filings so far reflects the partial year and publication lag rather than any real drop-off in activity.
Growth rate not stated: too few complete post-lag years to compute reliably.
Eureka AI Agent
Looking for what nobody has claimed yet?
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to knowledge graphs & graph computing: graph query optimization patent landscape, with the prior art for and against each one.
Source: Patsnap Eureka. Co-assignee relationships and derived observations. Derived from a Patsnap search on Knowledge Graphs & Graph Computing: Graph Query Optimization Patent Landscape covering 2015–2026, data cut-off 2026-07-31. Counts reflect published records only and shift as new filings publish.Run this in Eureka MCP
Next steps
Where to take this analysis
The dataset points to a concentrated leader, a China-weighted filing base, and claim language still anchored in core data processing rather than AI-model architecture. Each of those raises a different follow-up question.
Map the leader's claim scope
With one assignee holding 17 of 50 records, understanding exactly which query-planning steps and data structures its claims cover is the first filter for any new filing in this space.
Explore assignee claims in EurekaCheck China-first prior art
34 of 50 receiving-office records originate in China. Prior-art searches limited to US and EP registers will systematically miss more than two-thirds of the documented filing activity.
Run a China-weighted prior art searchTrack the RAG-driven filings
The most recent record in scope frames graph query optimization around retrieval-augmented generation. Watching this sub-branch for new entrants is worthwhile before it consolidates.
Monitor emerging RAG-related filingsSource: Patsnap Eureka. Forward-looking reading of the same dataset. Derived from a Patsnap search on Knowledge Graphs & Graph Computing: Graph Query Optimization Patent Landscape covering 2015–2026, data cut-off 2026-07-31. Counts reflect published records only and shift as new filings publish.Run this in Eureka MCP
FAQ
Common questions about graph query optimization patents
The assignee ranking behind this landscape lists 41 companies across the 50 records in scope. Filing activity is concentrated at the top: the leading assignee alone holds 17 records, and the top 5 combined account for 44.0% of all 50 records. Below that, the ranking thins into a long tail of assignees with one or two filings each, so most participants in this field are not repeat filers.
One assignee leads the ranking with 17 of the 50 records in scope, well ahead of the fifth-ranked assignee, which holds a single record. That gap between the leader and the rest of the ranked field is the clearest concentration signal in this dataset. It suggests that anyone assessing freedom to operate in this space should look closely at that leader's claim scope before filing new work.
China accounts for 34 of the tracked receiving-office filings, far ahead of the United States at 7, WIPO/PCT filings at 4, Europe and India at 2 each, and South Korea at 1. This means the bulk of documented graph query optimization activity is being filed through Chinese offices rather than through the US or European systems that many Western freedom-to-operate searches default to. A search limited to US and EP registers alone would miss most of the activity captured here.
Recorded filings rose from 2 in 2017 to a peak of 15 in 2024, showing a clear multi-year build-up in activity. The apparent drop to 2 filings in 2026 is not a real decline — it reflects the partial year and the roughly 18-month lag between filing and publication, which always understates the most recent period. A reliable growth rate cannot be computed yet because too few complete years remain once that lag is accounted for.
94.0% of the 50 records in scope carry a G06F (electric digital data processing) classification, making it the dominant technical framing for this field. G06N, covering AI-based computing, appears in 36.0% of records, reflecting a meaningful but secondary overlap with machine-learning and generative-model techniques. Smaller shares touch G06Q (business data processing), H04L (digital transmission), G06T and G06V (image-related processing), pointing to specific application niches rather than core claim territory.
Answers are grounded in the same dataset. Derived from a Patsnap search on Knowledge Graphs & Graph Computing: Graph Query Optimization Patent Landscape covering 2015–2026, data cut-off 2026-07-31. Counts reflect published records only and shift as new filings publish.Run this in Eureka MCP
Research Knowledge Graphs & Graph Computing: Graph Query Optimization Patent Landscape in depth with Eureka
Go past this page: query the whole knowledge graphs & graph computing: graph query optimization 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.