PatClaim.ai (forthcoming AI-related & AI-based work with patent claims)
- Including using text of patent claims as proxies for patent spec as a whole
- Patent mining based on patent claims (the legally-operative “business end” of patents often overlooked in traditional research methods).
- Patent searching based on semantic similarity of claim language, rather than solely on keywords, classifications, citations, etc.
- Custom patent-claim LLM which, given patent claim text as input, outputs predictions e.g. likely grant vs. non-grant, likely litigated vs. non-litigated
- Using large language models (LLMs) and neural networks trained specifically on patent data to generate sentence embeddings (mathematical representations of text) for claim comparisons.
- Patent “landscaping” using patent claim sentence embeddings, and SentenceTransformer models (like PatentSBERTa), to map claims into a vector space, potentially identifying “holes” in the space (un-claimed technological areas).
- Uncovering prior art using patent-claim semantic similarity (claims are a limited form of prior art, but use claims as proxies for earlier disclosures, where the earlier disclosure is already represented in claim language that might better match the current claim at issue)
- Applying machine learning to patent-portfolio ranking, based on the claims
- Possible case study of patents of QQQS holdings; compare QQQS to QQQ — does diff QQQS minus QQQ reflect value of patents?; to what extent is methodology based on patent claims?
- AI summary from Google Finance: “How does QQQS determine what are ‘promising patent portfolios’? [QQQS tracks NCX, which uses proprietary methodology from IPR Strategies; methodology] moves beyond simple patent counts to measure intellectual property values more effectively…. Relative value: The index focuses on the relative value of a company’s patents compared to its market capitalization. This allows QQQS to identify smaller companies where intellectual property may be undervalued by the broader market. Focus on future growth: The core theory is that high-quality, valuable patents are an early indicator of future growth potential and competitive advantages. These patents represent future assets and potential future revenue streams.”
- QQQS tracks Nasdaq Innovators Completion Cap Index (NCX), which Nasdaq states uses a proprietary methodology from a third-party patent valuation firm, IPR Strategies: “NCX’s methodology leverages a third-party dataset built by IPR Strategies — a leader in the science of patent valuation — to identify the 200 non-NDX/NGX companies listed on Nasdaq that have the highest-valued patent portfolios in relation to their market capitalizations (excluding financials)”
- Nasdaq paper, 2022
- Nasqad current NCX weighting (xls): TXG, EGHT, ACRS, ADPT, AEIS, AEHR, AVAV, AGIO, API, AKBA, etc. (index uses equal weighting; QQQS Top 10 holdings currently listed as QURE, VTYX, CLPT, NVTS, NKTR, PSNL, AQST, OMER, KOPN, VUZI)
- IPR Strategies on QQQS launch, Oct. 2022: “First patent value-based Innovation ETF QQQS launched“
- IPR Strategies paper, Nov. 2023: “IP-Value Factor-Based Strategy Show Outperformance in Nearly All Investigated Global Indices“
- IPR Strategies Aug. 2023: “IPR-Strategies releases new patent valuation dataset model” (“Improved patent valuation: The patent valuation formula has been further developed and improved … In addition, patent claims are more thoroughly analyzed and considered in the valuation of each single patent making the value prediction more accurate.”) [TODO: confirm whether NCX, QQQS based on this newer model]
- See also NQIPL (Nasdaq International Patent Leaders Index; “Using alternative data in the form of patent valuation estimates, NQIPL is designed to offer investors an equity benchmark that in many ways looks and feels like an international version of the Nasdaq-100”; methodology paper doesn’t reference patent claims); top holdings: TSMC; Tencent; Samsung; ASML; SAP; Toyota; SK hynix; Novartis; SoftBank; AstraZeneca; PATN ETF
- Developing tools to query PTO PatentsView API and Google Patents with SQL (BigQuery).
- Examining patent-related AI, such as PQAI (AT&T, GreyB; API)
- Sessions with ChatGPT (and Anthropic’s Claude) developing Python code to work with patent claims and patent-litigation data; “brainstorming”; discussing machine learning implementation and interpretability.
- Claim chart drafting using AI tools such as ChatGPT for infringement and invalidity contentions, and drafts of expert reports.
