Why do some SEPs not show an ES score?
A Note on How Essentiality Scoring Has Evolved
Interested in a proof of concept?
Overview
Standard Essentiality of a patent has long been a complex issue, often leading to conflicts, negotiations, and lawsuits. Traditionally, determining whether a patent may be essential to a standard requires significant investment in time, money, legal counsel, and subject-matter expertise. For example, assessing the potential essentiality of a portfolio of 100 patents can take 10 experts at least two weeks and cost over $150,000.
To address this, LexisNexis IPlytics has developed the Essentiality Score (ES) — a faster, more efficient way to predict potential patent essentiality. Within seconds, ES provides a score (1–100%) and score category, indicating how likely it is that a patent matches to the section of a standard.
We are currently scoring across five technology spaces, covering approximately 15 standards:
- Cellular (2G - 5G and bridging, including Cellular IoT)
- Audio Codes: (MPEG-H 3D, OPUS, AAC, IVAS, EVS)
- Video Codes: (AVC, HEVC, VVC, VP9, AV1)
- Wireless Charging Standards: (Qi1, Qi2)
- Wi-Fi: (Wi-Fi 1-6, Wi-Fi 7)
- Memory: (HBM, GDDR, LPDDR, and DDR)
This is achieved through semantic matching of independent patent claims to standard sections, and this methodology and approach is described in this article.
Note: A separate, LLM-based Essentiality Score has been introduced for the Memory technology package. The DRAM Memory patent database (covering HBM, GDDR, LPDDR, and DDR in the Memory Technology Package) is the first to use this LLM-based approach. To learn more about the LLM-Based Essentiality Score, click here.
How ES Works
Semantic Matching
Matches a patent to to the relevant section(s) of a standard, based on contextual meaning rather than shared vocabulary alone.
Similarity Scoring
A semantic similarity score (ES - as shown in the column above) is displayed for each patent and the machine learning algorithms determine which has similar language to that of a section of a standard. You can then sort and filter to identify the most likely essential patents by standard or portfolio.
Please note that a range from 40% to 100% is shown. Anything below an ES score of 40% is not displayed.
A column can also be added to the table view, Essentiality Score Category (ESC). The score band is illustrative of the predicted essentiality as follows. :
Band |
Meaning (short)
|
Example next actions
|
🟩 High ES |
High likelihood of being truly essential. |
Queue for expert confirmation/claim charting; include in licensing binders; monitor renewals closely. |
🟨 Medium ES |
Mixed signals; some may be partially essential. |
Route to analyst review; check feature/Release matching; decide escalate/defer based on business context. |
🟥 Low ES |
Low likelihood of true essentiality. |
Spot-check a small sample; monitor only for strategic owners/features. |
Portfolio Comparison
Refine results and compare portfolios or individual patents based on their ES.
Side-by-Side Analysis
View the most semantically matching independent patent claim alongside the relevant standard section.
Why do some patent families not show an ES score?
The IPlytics ‘Essentiality Score’ (ES) is a fast, efficient way to assess how closely a patent family matches a given standard. Within seconds, ES provides a score (1–100%) indicating how likely a patent is essential to the standard it was declared against. This is achieved through semantic matching of independent patent claims to standard sections.
The ES score is shown in the data view, as seen here:
There are a few reasons why an ES score may not be shown next to a publication:
- ES can only score English original documents (e.g., US or EP)
- While Chinese translated patents are scored with ES, the accuracy is not confirmed
- ES only reports the most semantically relevant claim section combinations. ES now selects the top three most relevant claim-section pairs and excludes the rest. Any other relevant standard sections are not reported.
- There may be a very low correlation between the claims and the standard and ES below 40% is not displayed.
A Note on How Essentiality Scoring Has Evolved
ES (previously named SES - Semantic Essentiality Score) has been continually refined since its introduction. The current Semantic Essentiality Score 2.0, replaced an earlier Latent Semantic Analysis (LSA) model, which occasionally flagged patents as relevant simply because they shared vocabulary with a standard, even when describing unrelated technology.
The current Word2Vec/TF-IDF/SVD framework resolved this by understanding semantic context rather than word overlap, meaningfully reducing false positives and the amount of manual review required:
We have now also introduced the LLM-Based Essentiality Scoring methodology for the Memory Technology Package. Click here to learn more about this methodology. Note that this score is also also shown under "ES" or Essentiality Score in the product.
Use Cases
For Portfolio Management:
- Identify the most and least likely essential patents in your or another’s portfolio.
-
Determine your relative share of the overall technology landscape
For Licensing & Business Development:
- To cut through the noise of over-declaration in the cellular space and determine what is most likely to be essential
- In spaces of under-declaration (Video, Audio, Wi-Fi, Qi), to understand which entities have relevant portfolios.
- To determine comparable licensing terms, using relative shares of the overall technology landscape.
Note: IPlytics ES is not a replacement for expert claim charting but serves as a valuable first step in evaluating potential essentiality of patent portfolios.
Strategic Patent Portfolio Analysis & Licensing Opportunities
How to combine IPlytics with ES to find high‑value patents, surface licensing opportunities, and track competitors.
Organizations can pair IPlytics’ rich patent analytics with the ES score to identify high‑value assets, size likely‑essential subsets, and guide licensing or collaboration strategies—while continuously monitoring competitor activity.
Key Components
-
IPlytics Platform
Comprehensive patent data, advanced search, and analytics for tracking filing trends, matching innovation clusters, and benchmarking technologies. -
ES (Essentiality Scoring)
Quantifies the semantic similarity between independent claims and standard sections to flag patents most aligned with essential technical disclosures and requirements.
Benefits of Integrating IPlytics + ES
-
Enhanced Patent Quality Assessment
• Use ES as an objective signal of claim–standard alignment.
• High‑ES assets strengthen licensing narratives and acquisition targeting. -
Efficient Portfolio Optimization
• Mine large portfolios in IPlytics; set ES thresholds to triage.
• Flag low‑ES assets for deeper review or rationalization; promote top‑ES assets for investment, defense, or commercialization. -
Competitive Intelligence
• Focus on competitors’ high‑ES patents to understand strategic bets, quality focus, and potential vulnerabilities. -
Licensing & Collaboration Opportunities
• Map high‑ES families across jurisdictions; combine with evidence views to support cross‑licensing or technology partnerships. -
Innovation Benchmarking
• Benchmark internal filings against standard sections; feed insights back to R&D and drafting practices to lift ES over time.
Practical Scenario (Example)
- Use IPlytics to search across relevant families in a target domain.
- Filter by domain criteria and sort by ES.
- Review top‑ES patents for strategic fit and technical rigor (claims ↔ standard matching).
- Initiate licensing discussions or adjust filing strategy to improve future ES.
Compelling Use Cases & Workflows
1) SEP Portfolio Triage & Expert Review Prioritization
Goal: Cut review noise; focus expert time on the right families.
How: Filter declared SEPs by standard/release; rank by ES; use claim↔section views to speed checks; set review thresholds and track movement.
Workflow: Inputs (Declared SEPs, claims, standard sections, portfolio meta) → ES filter (rank + ≥X threshold; cluster by feature/release) → Action (route top items to experts; generate evidence packs) → KPI (% confirmed essential; review TAT; expert hours saved)
2) Licensing Pipeline & FRAND Prep
Goal: Build a credible, data‑driven licensing narrative.
How: Cluster high‑ES assets by counterparty/product line/standard (e.g., NB‑IoT, LTE‑M, Cat‑1, V2X); package ES evidence for negotiation binders; monitor ES trends.
Workflow: Inputs (Counterparty features, declared families, jurisdictions, ES) → ES filter (identify high‑ES per feature & counterparty) → Action (binders, outreach, chart prep) → KPI (meeting‑to‑offer rate; royalty uplift; time to first agreement)
3) Launch Risk Scan (Freedom‑to‑Operate Gate)
Goal: Prioritize external owners most likely to impact a launch.
How: For target features, surface third‑party families with higher ES; watchlist and fast‑track charting/design‑arounds; escalate blockers.
Workflow: Inputs (Feature list, release timeline, 3P portfolios, ES) → ES filter (rank by relevance to planned features) → Action (watchlist; deep‑dive; mitigation) → KPI (# critical risks surfaced early; lead time; post‑launch assertions avoided)
4) M&A / Asset Valuation & Portfolio Shaping
Goal: Estimate likely‑essential subset for pricing and diligence.
How: Combine ES‑ranked sets with portfolio KPIs to frame valuation ranges and keep/sell decisions; watch ES movement post‑prosecution as an early signal.
Workflow: Inputs (Target families, jurisdictions, prosecution status, ES, quality KPIs) → ES filter (size high‑ES subset; bucket core/adjacent/legacy) → Action (price ranges; term‑sheet adjustments) → KPI (value per high‑ES family; deal confidence; monetization yield)
5) Undeclared Patent Scouting (Gap & Opportunity)
Goal: Find technically relevant—but undeclared—assets for acquisition or enforcement.
How: Use IPlytics’ undeclared landscapes; filter by ES to surface promising candidates in Wi‑Fi, video, Qi, etc.; route to experts.
Workflow: Inputs ( undeclared flag, ES) → ES filter (high‑ES undeclared by feature/standard) → Action (acquisition shortlist; continuation strategy; enforcement eval) → KPI (acquisition hit rate; chart‑to‑license conversion; time to first assertion)
Interested in a proof of concept?
Book a free consultation: sales@iplytics.com