Google Patents Archive

Isang curated archive ng Google at Microsoft search patents na inorganisa ayon sa inventor o research lead na nag-ship sa kanila. Bawat track ay sumusunod sa isang engineer sa pamamagitan ng kanilang patent lineage upang makita ng reader kung paano nag-evolve ang research thread sa continuations, divisional applications, at grants. Ang mga patents ay may annotations sa ranking surface na hinihipo nila, query understanding, ranking, retrieval, behavioral signals, knowledge graph, o AI search.

Patent tracks ayon sa inventor

• 65 Google Patents to Help You Understand How Search Engines Work! (69 patents) — Isang curated collection ng foundational at advanced patents na nile-file ng Google.
• Steven Baker — Google Search Patents (49 patents) — 49 granted search-engine patents plus 2 published applications ng Google researcher Steven D. Baker, sumasaklaw sa query understanding (synonyms, n-gram, geographic, cross-language), query refinement, meaningful stopword detection, answer passage scoring (featured snippets), query–document similarity, list co-occurrence, at embedding-based personalized search.
• Amit Singhal — Google Search Patents (22 patents) — 22 search-engine patents ng Amit Singhal, dating Google SVP of Search (2000-2016), sumasaklaw sa interleaving at result composition, query rewriting, semantic query understanding, link-quality scoring, location-aware ranking, language at country biasing, commercial-intent detection, at advertisement serving. Ang 8-patent meaningful-stopwords family ay co-invented kasama si Steven Baker at nidocument sa Steven Baker section upang iwasan ang duplicate content.
• Prabhakar Raghavan — Google Search Patents (16 patents) — 16 search/IR patents ni Prabhakar Raghavan, head ng Google Search 2020-2024 (dating Yahoo VP at IBM Almaden research scientist), sumasaklaw sa personalized PageRank, web community detection, adaptive ranking, multi-engine result merging, hierarchical taxonomy, subspace clustering, information filtering, customizable navigation, user-generated content surfacing, location-quality signals mula sa travel patterns, at temporal metadata analysis. Sumasaklaw sa kanyang IBM, Yahoo, at Google research.
• Nitin Gupta — Google Search Patents (32 patents) — 32 search/SEO patents ni Nitin Gupta (Google Search Quality engineer) sumasaklaw sa query suggestions at templates, reranking at personalization, answer-generating neural networks, content-knowledge generation, ang latest 2026 generative-search retrieval patent para sa AI Overviews era, at isang direct SEO patent sa domain names at websites. Ang final 5 patents ay co-invented kasama si Steven Baker at nidocument sa kanyang section upang iwasan ang duplicate content.
• Navneet Panda — Google Search Patents (15 patents) — 15 search-quality patents ni Navneet Panda, ang Google engineer na ang content-quality work ay naging basis para sa Panda algorithm update (2011). Sumasaklaw sa site quality scoring, pagpredikta ng site quality bago ma-accumulate ang behavioral data, ranking ng search results sa pamamagitan ng combined relevance at site-quality signals, website duration performance, locally significant queries, query revision, at image-concept learning.
• Anand Shukla — Google Search Patents (16 patents) — 16 search/IR patents ni Anand Shukla, Google Search Quality engineer, kasama ang cutting-edge 2024 "Search with stateful chat" patent na nag-bridge sa classical search hanggang LLM-based conversational search. Sumasaklaw din sa in-depth article surfacing, query pattern matching, query topic mapping, content channel curation, at ang enhanced-search-feed family. Ang final 3 patents ay co-invented kasama si Steven Baker sa personalized-feed embeddings work at nidocument sa kanyang section.
• Krishna Bharat, Google Search Patents (48 patents) — 48 search-engine patents ni Krishna Bharat, founding inventor ng Google News at creator ng Hilltop algorithm. Sumasaklaw sa Hilltop link-analysis algorithm, query rewriting gamit ang entity detection, semantic units sa queries, query at search augmentation, document ranking sa pamamagitan ng usage statistics, image selection para sa news search, editorial-opinion ranking, search engine coverage estimation, query-specific bookmarking, at entity-aware ad rendering. Sumasaklaw mula 2001 hanggang 2024.
• Srinivasan Venkatachary, Google Search Patents (25 patents) — 26 search-engineering patents ni Srinivasan Venkatachary, co-inventor sa Generative Summaries para sa Search Results patent family na sumusuporta sa AI Overviews at Search Generative Experience. Sumasaklaw din sa enhanced search feeds batay sa user interests, candidate-answer-passage scoring, context-scoring adjustments para sa answer passages, implicit question query identification, keyword-based conversational voice search, query generation gamit ang document structural similarity, resource identification mula sa organic at structured content, content channel curation, at ang emoji classifier. Sumasaklaw mula 2013 hanggang 2025.
• Jeromy William Henry, Google Search Patents (25 patents) — 25 search-engine patents ni Jeromy William Henry sumasaklaw sa Knowledge Panels (ang entity card sa tabi ng search results), Structured Entity Information Pages (ang standalone entity-detail surface), Personalized Entity Information Pages, SERP image sizing, at music-search secondary link presentation. Bawat patent sa kanyang corpus ay SEO-relevant. Sumasaklaw mula 2013 hanggang 2025.
• Ramanathan V. Guha, Google Search Patents (33 patents) — 33 search-engine patents ni Ramanathan V. Guha, co-founder ng Schema.org at lead inventor sa Google Custom Search Engine. Sumasaklaw sa Programmable Search Engine family, trust-based search ranking, spam detection para sa programmable contexts, social-signal propagation, classifying search results, customized web summaries, filtering sa pamamagitan ng user annotations, contextual searching sa pamamagitan ng query intersections, query identification, at ang generating-tasks UI family. Sumasaklaw mula 2003 hanggang 2019.
• Marc Najork, Google Search Patents (25 patents) — 25 search/IR patents ni Marc Najork na sumasaklaw sa kanyang Microsoft Research era (web crawler architecture, spam-resistant ranking, content evaluation, hyperlink databases) at kanyang Google era (form information extraction, ML-driven semantic similarity, active learning, document activity logs para sa relevance training). Kasama ang isang foundational cross-listed patent kasama ang 65 Google Patents collection sa document activity logs. Sumasaklaw mula 2005 hanggang 2025.
• Jeffrey Dean, Google Search Patents (78 patents) — 78 search at IR-infrastructure patents ni Jeffrey Dean, Google Senior Fellow at Chief Scientist, co-founder ng Google Brain, at architect ng MapReduce at BigTable. Kasama ang foundational Document Scoring families (Query Analysis, Content Update, Inception Date, Link-Based Criteria), Historical Data Patent (cross-listed kasama ni 65gp), Anchor Tag Indexing, Content Snippets gamit ang Tokenspace, near-duplicate at mirrored-host detection, topic-drift prevention, distributed ML models, at ang foundational MapReduce at BigTable infrastructure na ginagawang posible ang web-scale search. Sumasaklaw mula 2000 hanggang 2025.
• Paul Haahr, Google Search Patents (63 patents) — 63 Google search at ranking patents ni Paul Haahr, long-tenured Google search ranking engineer. Sumasaklaw sa canonical link-quality patents (Bharat/Singhal/Haahr "Determining quality ng linked documents"), document-classification ranking, node-independence anti-PBN signal, ang rumored quality-measure patent (cross-listed kasama ang 65gp pat-48), ang multi-stage query processing pipeline, ang framework para sa pag-evaluate ng web search scoring functions (ang documentary back-stop sa kanyang SMX talk), preferred sites, reference contexts, synthetic descriptive text para sa images, at ang multi-family query refinement at stopword-detection clusters. Co-authored kasama si Singhal, Dean, Bharat, Cutts, Henzinger, Simon Tong, at Acharya. Sumasaklaw mula 2008 hanggang 2019.
• Hyung-Jin Kim, Google Search Patents (23 patents) — 23 Google search at ranking patents ni Hyung-Jin Kim. Lead inventor sa Navboost / implicit-feedback ranking family (US 8,661,029 + 5 continuations hanggang US 11,816,114 active 2023 IP), ang system na isinalahis sa 2024 DOJ Google antitrust trial. Sumasaklaw din sa presentation-bias model, temporal score adjustments, document-change ranking, user cohort grouping, similar-query borrowing, variant generalized queries, propagating query classifications, at ang cross-listed Quality Measure patent (kasama si Paul Haahr, 65gp pat-48). Co-authored kasama si Simon Tong, Noam Shazeer, Michelangelo Diligenti, Adrian Corduneanu, Henele Adams, Andrei Lopatenko. Filings 2006-2021.
• Pandu Nayak, Google Search Patents (13 patents) — 13 Google search patents ni P. Pandurang Nayak, Google Chief Scientist para sa Search. Ang patent record ay sumasaklaw sa query-revision at query-understanding stack mula 2003-2015, integration ng multiple query-revision models, confidence-scored revision, known-highly-ranked-queries borrowing, concept-context substitution, phrase-restricted synonyms, click-driven synonym identification, acronym expansion. Co-authored kasama si David R. Bailey, Ben Gomes, Trystan Upstill, Kedar Dhamdhere, Thomas Strohmann. Ang kanyang neural-era contributions (RankBrain, BERT, MUM, DOJ testimony) ay intentionally unpatented at nabubuhay sa blog posts at trial transcripts.
• Daniel Egnor, Google Local Search Patents (20 patents) — 20 Google local-search at geographic-relevance patents ni Daniel Egnor, ang architect ng Google Maps / Local Pack ranking layer. Lead inventor sa US 8,046,371 "Scoring local search results batay sa location prominence", ang patent na tumutukoy sa "Prominence" factor sa official Local Search Ranking documentation ng Google. Sumasaklaw din sa location-sensitivity ranking (Relevance factor, kasama si Amit Singhal), geographical-relevance indexing, ambiguous-geographic-reference resolution, authoritative-document identification, business-listing categorization, at GIS ambiguous-search processing (kasama si Keyhole founder John Hanke). Co-authored kasama si Singhal, Haahr, Greenfield, John Hanke. Filings 2003-2019.
• Yossi Matias, Google SERP Features & Assistant Patents (47 patents) — 47 Google patents ni Yossi Matias, Google VP Engineering. Lead inventor sa foundational People-Also-Ask patent US 9,213,748 "Generating related questions para sa search queries". Built Autocomplete (2004), Trends, weather/sports/dictionary OneBoxes. Sumasaklaw sa People-Also-Ask family, Knowledge Panel entity scaffolding (grouping related entities, central-entity identification, fact estimation, textual entity attributes, entity-quote search), ang Trends/analytics layer (campaign-competitive analysis, ranking graphical visualizations), ang Duplex automated-call infrastructure, ang personalized chatbot at LLM-output entailment patents na sumusuporta sa Google Assistant's 2024-2026 generative era, at contextual / third-party SERP-features. Co-authored kasama si Yaniv Leviathan (Duplex), Gal Chechik, Ziv Bar-Yossef, Eyal Segalis, Avinatan Hassidim. Filings 2013-2026.
• Andrei Broder, Cross-Vendor Search Patents (25 patents) — 25 search-engine at IR patents ni Andrei Broder, ang cross-vendor inventor na namuo ang near-duplicate detection sa DEC/AltaVista, co-invented ang foundational CAPTCHA sa Compaq, itinayo ang AltaVista web-page ranking system, nag-author ng web-page-decay signal sa IBM, at namuo ang ad-matching/SERP-features/CAPTCHA-evolution infrastructure sa Yahoo. Lead inventor sa foundational MinHash/shingling patent (US 6,349,296), ang fingerprint-collision-probability technique (US 5,974,481), ang document-resemblance shingling patent (US 6,230,155), at ang foundational CAPTCHA patent (US 6,195,698). Sumasaklaw mula 1999 hanggang 2024.
• Monika Henzinger, Google Search Patents (50 patents) — 98 search at IR patents ni Monika Henzinger, ang first Director ng Research sa Google. Co-inventor sa foundational near-duplicate-page-detection patent (US 6,138,113, kasama si Dean) at isang extensive set ng document-scoring families (Query Analysis, Content Update, Inception Date, Link-Based Criteria, Historical Data, lahat ay cross-listed kasama ang Dean's canonical articles). Independent contributions, detecting-duplicate-files family kasama si William Pugh, ang AltaVista connectivity-server, ang connectivity-at-content-ranking patents, document-freshness determination, semantic-distance ranking, query-semantic-information ranking, anchor-text cross-language IR, in-context searching, hypertext-browser assistant, at usage-statistics-driven document retrieval. Sumasaklaw mula 1997 hanggang 2017.
• Trystan Upstill, Google Site-Quality & HCU-Era Search Patents (28 patents) — 28 Google search at ranking patents ni Trystan Upstill. Co-inventor sa Panda patent (US 9,135,307 = 65gp pat-39) at lead inventor sa HCU-era follow-up US App 2019/0155948 "Re-ranking resources batay sa categorical quality", ang patent na nag-anchor sa post-Panda site-quality lineage. Sumasaklaw din sa authoritative-results discovery, search-result ranking at re-scoring, affiliated-domains detection, resource-attribute extraction mula sa site address, local search at POI identification, synonyms at query rewriting, navigational-intent resources, language identification mula sa link context, at ang recent LLM-era UI components patent. Filings 2012-2026.
• Susan Dumais, Microsoft IR & Search Patents (18 patents) — 61 search at IR patents ni Susan Dumais, Microsoft Technical Fellow at Gerard Salton Award winner. Co-inventor sa foundational Latent Semantic Indexing patent (US 4,839,853, Bellcore 1989), ang conceptual ancestor ng bawat dense-embedding retrieval system. Sa Microsoft Research ang kanyang trabaho ay sumasaklaw sa automated SERP satisfaction measurement, preference-judgment click models, activity-based-context ranking, time-aware ranking na may temporal dynamics, personalized navigation at search results, per-user domain-expertise determination, web-page-change × revisitation freshness, implicit device-related query reformulation, authority ranking. Sumasaklaw mula 1989 hanggang 2026. Ang mga co-authors ay kasama si Jaime Teevan, Eric Horvitz, Ryen White, Adam Fourney, Joshua Goodman, Eric Brill.
• Christopher Burges, Microsoft Research Learning-to-Rank Patents (16 patents) — ~30 unique US inventions (84-entry WIPO corpus) ni Christopher J. C. Burges, ang Microsoft Research scientist na nag-invent ng RankNet/LambdaRank/LambdaMART learning-to-rank lineage. RankNet (US 7,689,615) ang pairwise neural ranker, LambdaRank (US 7,617,164) ang arbitrary-cost extension na nag-optimize sa NDCG, LambdaMART (US 12/032,697) ang gradient-boosted ensemble na nanalo sa Yahoo LTR Challenge 2010. Sumasaklaw din sa behavioral-variability adaptation, web-page multi-graph analysis, link-spam smoothing, margin-tree boosting, k-NN probability estimation. Sumasaklaw mula 2001 hanggang 2019. Ang mga co-authors ay kasama si Robert Ragno (LambdaRank), Irina Matveeva, Timo Burkard.
• Eric Brill, Microsoft Research Search & NLP Patents (20 patents) — 20 Microsoft Research patents ni Eric Brill, ang NLP at search scientist na kilala para sa query speller patents at ang noisy-channel correction model. Lead inventor sa US 7,254,774 "Systems at methods para sa improved spell checking", ang foundational noisy-channel query speller patent. Sumasaklaw din sa string-to-string spell-correction transformations (US 7,290,209 / 7,366,983), behavioral-variability accounting sa web search (US 7,743,047), popularity-data ranking, mining web search user behavior, page-biased search, cost-benefit Q&A composition, at user-intent discovery. Filings 2005-2010.
• Tomas Mikolov, Google word2vec Patents (2 patents) — 2 foundational Google patents ni Tomas Mikolov sumasaklaw sa word2vec, ang continuous word-embedding learning model na tinukoy ang embeddings era. Co-invented kasama si Kai Chen, Greg Corrado, at Jeff Dean. US 9,037,464 ang original grant, US 9,740,680 ang continuation na may mas malawak na claim scope. Ang parehong patents ay nagbabahagi ng parehong disclosure (application 13/841,640) at magkasama ay bumubuo sa buong word2vec patent family. Cross-listed conceptually sa Dean's section (Dean ang co-inventor at gumagana sa mas malawak na Brain/distributed-models infrastructure).
• Noam Shazeer, Google Transformer, MoE & Search Patents (18 patents) — ~46 captured Google patents (buong portfolio ay estimated sa 120-150) ni Noam Shazeer. Co-inventor sa foundational Transformer attention architecture (US 10,452,978 kasama si Vaswani, Polosukhin, Uszkoreit, Jones, Gomez, Kaiser, Parmar), ang Sparsely-Gated Mixture-of-Experts scaling approach (US 11,769,055 kasama si Dean, Hinton, Le, Mirhoseini), ang Switch Transformer (US 12,093,829), at ang Navboost implicit-feedback ranking family (US 8,661,029 kasama si Kim/Tong/Diligenti, cross-listed). Sumasaklaw din sa LLM-in-assistant response generation, distributed tensor computations (GSPMD/Pathways), at ang 2008-2015 pre-Transformer large-scale ML ranking infrastructure. Sumasaklaw mula 2008 hanggang 2025.
• Ashish Vaswani, Google Transformer, Attention-Based Vision & Fast Decoding Patents (6 patents) — 4 captured new canonical articles plus 2 cross-listings mula sa Noam Shazeer section. Lead author ng foundational Transformer patent (US 10,452,978, cross-listed). First inventor sa local-self-attention computer vision patent (US 20210390410, HaloNet) na nag-scale ng attention sa long sequences at images. Co-inventor sa MultiModel (US 20200089755, ang unified multi-task multi-modal architecture na ang structural ancestor ng MUM at Gemini), ang fully-attentional computer vision system (US 20250292560, ang ViT lineage), at ang fast-decoding-with-discrete-latents method (EP 3732627) na ginagawang economically viable ang LLM inference sa search-scale latency. Sumasaklaw mula 2017 hanggang 2025.
• Jakob Uszkoreit, Google Transformer Architect, Universal Transformer & Vision Transformer Patents (6 patents) — 3 new canonical articles plus 3 cross-listings mula sa Vaswani at Shazeer sections. Co-author ng "Attention Is All You Need" (US 10,452,978, cross-listed). Co-inventor sa Universal Transformer (US 10,740,433, ang adaptive-depth Transformer na may per-token Adaptive Computation Time), ang Decomposable Attention NLI model (EMNLP 2016, ang load-bearing pre-Transformer proof na attend-compare-aggregate ay nagpalit ng recurrence), at ang Vision Transformer (ViT, ICLR 2021, ang patch-tokenization architecture na nag-anchor sa modernong visual search at multimodal grounding). Ang mga cross-listings ay sumasaklaw sa original Transformer, MultiModel (ang unified multi-task multi-modal ancestor ng MUM/Gemini), at ang Image Transformer. Sumasaklaw mula 2016 hanggang 2021.
• Simon Tong, Google Click-Driven Ranking, Population Signals & Anti-Spam Patents (9 patents) — 5 new canonical articles plus 4 cross-listings mula sa Kim, Shazeer, at 65 Google Patents sections. Ang Tong ay co-inventor kasama si Marc Pearson at Sergey Brin sa related-query ranking patent (US 7,505,964), kasama si Pearson sa population-information ranking patent (US 7,454,417), solo sa country biasing (US 20040254932), kasama si Bem/Harik/Levenberg/Shazeer sa pre-Transformer large-scale ML ranking infrastructure (US 7,222,127), at kasama si Ghemawat/Piscitello/Cutts sa user-document-removal patent (US 8,417,697, ang structural ancestor ng Personal Blocklist at crowd-sourced spam signals). Ang mga cross-listings ay sumasaklaw sa Navboost implicit-feedback family, ang CTR-as-ranking-factor patent, ang historical-data patent, at ang large-dataset-ranking continuation. Sumasaklaw mula 2003 hanggang 2017+.
• Michelangelo Diligenti, Google Click-Data Quality & Implicit Feedback Patents (5 patents) — 2 new canonical articles plus 3 cross-listings mula sa Kim at 65 Google Patents sections. Ang Diligenti ay inventor sa Detecting Click Spam patent (US 8,694,374, attribute-deviance anomaly detection na sumusuporta sa ranking signal) at sa Click Model That Accounts para sa User Intent (US 20120143789, intent-conditional click weighting). Ang kanyang Navboost / presentation-bias / CTR-as-ranking-factor co-inventorships (cross-listed) ay nag-anchor sa kanya sa implicit-feedback ranking family. Ang portfolio ay nakatuon sa click-data quality, ang layer sa pagitan ng raw user behavior at ang ranking signal na sinusupportahan nito. Sumasaklaw mula 2007 hanggang 2019+.
• Jaime Teevan, Microsoft Personalized Search & Search-Task Patents (5 patents) — 4 new canonical articles plus 1 cross-listing mula sa Dumais section. Ang Teevan ay co-inventor kasama si Dumais at Horvitz sa foundational personalized-search patent (US 7,693,818) na nag-re-rank ng global candidates laban sa per-user profile vector. Inventor sa search-task identification (US 8,326,824, cross-session query clustering sa multi-step tasks), sa re-finding vs new-finding classification (US 8,756,219, pag-stabilize ng prior clicked results para sa return visits), at sa microtask search decomposition (US 10,062,103, pag-decompose ng high-level goals sa resumable micro-steps). Ang cross-listing ay sumasaklaw sa personalized-navigation (Dumais canonical). Sumasaklaw mula 2003 hanggang 2018+.
• Eric Horvitz, Microsoft Decision-Theoretic Search & Attention-Aware Information Delivery Patents (3 patents) — 3 new canonical articles. Ang Horvitz ay inventor sa decision-theoretic ranking sa ilalim ng uncertainty (US 7,124,129, relevance bilang probability distribution na may expected-utility selection), sa cost-benefit reasoning sa search-result presentation (US 7,194,464, ang mechanical basis para sa featured snippets at AI Overviews), at sa attention-aware proactive information delivery (US 6,934,917, ang zero-query surfaces lineage sa likod ng Discover, Now cards, at AI assistants). Sumasaklaw mula 2002 hanggang 2006+.
• Ryen White, Microsoft Search Behavior & Search-Trails Patents (3 patents) — 3 new canonical articles. Ang White ay inventor sa search-trail / successful-path mining (US 7,904,440, ranking destinations na nagtatapos sa successful cross-user trails), sa exploratory-versus-lookup query phase classification (US 8,041,711, paghahatid ng phase-tuned ranking at result formats), at sa cross-session task continuation (US 9,116,996, pag-stitch ng queries sa mga sessions at devices sa isang continuing task). Sumasaklaw mula 2007 hanggang 2015+.
• Azalia Mirhoseini, Google ML Infrastructure & Chip-Design Patents (3 patents) — 2 new canonical articles plus 1 cross-listing mula sa Shazeer section. Ang Mirhoseini ay co-inventor sa RL-based chip floorplan placement patent (US 11,475,278, ang Nature 2021 paper na nag-design ng TPU floorplans sa loob ng oras sa halip na linggo) at sa device-placement-for-distributed-ML patent (arXiv 1706.04972, ang RL controller na natututo kung paano mag-split ng malalaking models sa GPUs/TPUs upang mabawasan ang training time). Ang parehong feed back sa compute-cost economics na nagpapahintulot sa Google na maglingkod ng Transformer ranking sa search-scale. Ang cross-listing ay sumasaklaw sa foundational MoE patent (Shazeer canonical). Sumasaklaw mula 2017 hanggang 2021+.
• Arnold Overwijk, Microsoft Site Topical Authority Patent (1 patents) — 1 canonical article. Ang Overwijk ay co-inventor kasama si Li Xiong, Chuan Hu, at Junaid Ahmed sa Site Topical Authority patent (US 20210004416, ang formula A(s,t) = SR(s) * C(s) * T(s,t)^2 na malinaw na nagsasaad ng per-site per-topic authority sa pamamagitan ng pagsasama ng global site rank, click signal, at squared per-topic page share). Ang pinaka-tiyak na mathematical formulation ng topical authority sa modernong patent literature. Conceptually paired sa Procopio (per-author per-topic authority) at Agent Rank (author-level reputation, 65gp pat-47).
• Michael Procopio, Google Per-Author Per-Topic Authority Patent (1 patents) — 1 canonical article. Ang Procopio ay named inventor sa Topic Authority patent (US 8,458,196). Ang system ay nagsasagawa ng per (author, topic) authority signatures, para sa bawat document d na nag-contribute ang author, sum (authorship% * topic_weight), aggregate sa lahat ng documents ng author upang makabuo ng per-author per-topic expertise score. Ang mathematical quantification ng "ang taong ito ay isang authority sa topikong ito," structurally ang basis para sa E-A-T Expertise signal sa author level. Nag-connect sa Agent Rank (US 7,565,358, 65gp pat-47, ang mas maaga na Google author-reputation patent) at sa Overwijk's site-topical-authority (US 20210004416) upang bumuo ng layered authority model.
• Ori Allon, Google Orion Query-Refinement Patent (1 patents) — 1 canonical article. Ang Allon ay co-inventor sa Orion query-refinement patent (US 8,392,443) kasama si Ugo Di Girolamo, Tomer Shmiel, Alexandre Petcherski, at Tzvika Hartman. Ang Orion algorithm ay nagmula bilang PhD-thesis work ni Allon sa UNSW at nag-acquire ng Google sa 2006. Ang patent ay lumilikha ng corpus-mined query refinements, ibinigay ang query, kunin ang phrase patterns mula sa documents na nakakatugon sa query upang makabuo ng ranked refinement candidates. Ang mechanism sa likod ng "Searches related to," extended SERP snippets, at ang lineage ng People-Also-Ask. Granted 2013.
• Corin Anderson, Google Reasonable Surfer (Foundational) Patent (2 patents) — 1 new canonical article plus implicit cross-listings sa 65 Google Patents collection. Ang Anderson ay FIRST inventor sa foundational Reasonable Surfer patent (US 7,716,225, kasama si Jeff Dean at Alexis Battle), ang 2004-filed grant na nagpapakilala ng non-uniform link-click probability weighted ng per-link features (anchor text, font size, position, surrounding context). Ang continuations (US 8,117,209 sa 65gp pat-45, US 9,305,099 sa 2016) ay lumalaki sa parehong modelo. Ang mechanism para sa differentiated PageRank flow sa mga links sa parehong page.
• Ashok Popat, Google OCR & Image-Text Indexing Patents (1 patents) — 1 canonical article na sumasaklaw sa translation-inspired OCR research ni Popat (Popat et al., 2009 onward, US 8,675,012 patent family para sa selective display ng OCR'd text). Ang methodology ay nag-reframe ng OCR bilang statistical machine translation, image features bilang source channel, character sequences bilang target, joint decoding kasama ang channel at language models ay nag-maximize ng transcription quality. Sumusuporta sa Google Books indexing, Lens text capture, Drive PDF search, at ang image-bound-text indexing pipeline sa likod ng bawat image at infographic sa web.

Bakit ang inorganisasyong ito

Ang patent law ay nag-cluster ng intellectual property ayon sa inventor at assignee, hindi sa topic. Ang pag-group ayon sa inventor ay nagbabalik ng narrative arc, ang dekada ng ranking work ni Amit Singhal, ang BERT-era query handling ni Pandu Nayak, ang news ranking lineage ni Krishna Bharat. Bawat track ay nagbabasa tulad ng research biography na isinulat sa claim language.