Open Access. Powered by Scholars. Published by Universities.®
Articles 2461 - 2490 of 4185
Full-Text Articles in Entire DC Network
Predictive Capacity Planning: Estimating Hardware Requirements For Customer Deployments, Mahesh Chidrawar, Sanjeev Panem Jaya, Umesh Prabhakar Gaikwad
Predictive Capacity Planning: Estimating Hardware Requirements For Customer Deployments, Mahesh Chidrawar, Sanjeev Panem Jaya, Umesh Prabhakar Gaikwad
Defensive Publications Series
Achieving suitable dimensioning for accurately determining the hardware specifications necessary for new mobile gateway deployments is a complex challenge, involving numerous variables whose interrelationships must be understood in the context of specific use cases. This research aims to develop a model for estimating the necessary capacity for product dimensioning and deployments on existing hardware infrastructure. By leveraging data from current customer labs, the study seeks to facilitate effective capacity planning for clients. The model will be constructed through a comprehensive analysis of various customer datasets, enabling the prediction of the hardware requisites tailored to specific deployment needs. The objective is …
Automated Self-Improving And User-Responsive Video Chapter Generation Using Generative Artificial Intelligence, Ilia Labzovsky, Danny Karmon
Automated Self-Improving And User-Responsive Video Chapter Generation Using Generative Artificial Intelligence, Ilia Labzovsky, Danny Karmon
Defensive Publications Series
Traditional video segmentation methods that offer limited granularity, static chapter structures. Such solutions lack deep semantic understanding and do not incorporate user interaction data for video segmentation, leading to inefficient information discovery and a poor user experience. This disclosure describes the use of generative artificial intelligence (genAI) to automatically create and dynamically refine smart chapters for long-form videos. These chapters are rich in detail, including titles, summaries, key concepts, and preview clips. With user permission, user interaction data can be used to dynamically adapt the chapter structure for individual viewers or the overall population. The techniques transform static videos into …
Augmenting Photo Reminiscing Experiences With Llm-Generated Narrations, Ji Hun Kim, Abdulrahman Khan, Tommy Meaney, Tim Zhu, Zhou Zhou
Augmenting Photo Reminiscing Experiences With Llm-Generated Narrations, Ji Hun Kim, Abdulrahman Khan, Tommy Meaney, Tim Zhu, Zhou Zhou
Defensive Publications Series
This disclosure describes techniques that leverage a large language model (LLM) to generate custom narrations when arranging sets of photos in the form of a visual experience for users to reminisce on thereby personalizing and adding depth to the reminiscing experience. Per the techniques, a set of photos from a user’s photo library to be featured in a visual experience is identified. Photo pixels, metadata, and other relevant data are used to populate a prompt that is to an LLM. The response from the LLM is parsed and associated with photos to be featured in the visual experience. The LLM-generated …
Real-Time Feedback On User-Written Generative Artificial Intelligence Model Prompts, Gulmohar Khan, Ramprasad Sedouram, Karthik Srinivas
Real-Time Feedback On User-Written Generative Artificial Intelligence Model Prompts, Gulmohar Khan, Ramprasad Sedouram, Karthik Srinivas
Defensive Publications Series
A common user interface for consumer use of generative artificial intelligence (genAI) tools is a chat interface, where a user can input their request as a text prompt within an input field and where the genAI model generates a response that answers the prompt. When users are insufficiently skilled at prompting, they may require multiple interactions with the genAI tool before a satisfactory result is generated. There are no mechanisms that guide the users to provide a prompt that uses good prompting strategies to express their intent. Also, user-provided prompts may include errors, gaps, or other issues that impact the …
System And Method For Thermochromic Indicator Payment Card, Mohit Koul, Rohit Mukherjee
System And Method For Thermochromic Indicator Payment Card, Mohit Koul, Rohit Mukherjee
Defensive Publications Series
The present disclosure related to a system and method for thermochromic indicator payment card. The system and method more specifically implements thermochromic technique. The payment card is embedded with thermochromic material that can change colour in response to temperature fluctuation, providing a passive and dynamic visual indicator for transaction verification and security alerts. Further, the present disclosure enables real time transaction awareness and fraud detection without relying on electronic alerts and external power sources.
Tax Rule Validation Using Large Language Models, Sandeep Chandna, Misha Narinsky
Tax Rule Validation Using Large Language Models, Sandeep Chandna, Misha Narinsky
Defensive Publications Series
The process of mapping products to jurisdiction-specific tax categories and validating rates against evolving legislation can involve complex, manual effort. A system is described that may leverage Large Language Models (LLMs) to assist with these tasks. The system can be configured to perform functions such as interpreting natural-language text from tax laws to classify products from an internal catalog. Additionally, the system can source and ingest public tax data to compare a company's internal tax rule configurations against external information sources. By automating the cross-referencing of internal product data with external legal and rate information, such a system can facilitate …
Establishing A Debug Channel Using Conductive Components Of An Electronic Device To Create Switchable Communication Paths, Averill Chang, Zoro Chang
Establishing A Debug Channel Using Conductive Components Of An Electronic Device To Create Switchable Communication Paths, Averill Chang, Zoro Chang
Defensive Publications Series
This disclosure describes a technique for establishing a debug channel over a wired connection by repurposing hardware components of an electronic device to create switchable communication paths for data transfer. In particular, internal debug signals can be selectively routed to non-dedicated external hardware contacts, for example, by using switching mechanisms. This approach may provide a stable and reliable debugging channel, potentially reducing the need for dedicated, corrosion-prone debug pins and mitigating certain performance and cost considerations associated with some wireless communication techniques.
A System For Recommending Advertising Campaign Settings Based On Performance Data Analysis, Ankit Aggarwal
A System For Recommending Advertising Campaign Settings Based On Performance Data Analysis, Ankit Aggarwal
Defensive Publications Series
Advertisers may experience difficulty in defining effective targeting and configuration settings for digital promotional campaigns, which can result in performance that does not meet certain goals for scale and return on investment (ROI). A system can utilize an artificial intelligence agent to analyze historical campaign data. The system may correlate past campaign configuration changes, such as adjustments to bids, budgets, and targeting criteria, with corresponding performance metrics including, for example, impressions, redemptions, and cost-per-action. Based on this correlation analysis, the system can formulate tailored recommendations for campaign settings to help meet specific advertiser goals. This process can contribute to improved …
Generating Context-Aware Linearized Representations Of Tabular Data For Language Models, Ashutosh Dwivedi
Generating Context-Aware Linearized Representations Of Tabular Data For Language Models, Ashutosh Dwivedi
Defensive Publications Series
Large Language Models (LLMs) may experience challenges when processing information from hypertext markup language (HTML) tables, as providing table data without its original on-page context can lead to suboptimal comprehension and factuality. A disclosed technology can address this with a two-step offline process. First, a context generation module can create a descriptive summary for a table by analyzing its page title, table title, and nearby caption text. Second, a linearization module can use this context to convert the structured HTML table into a descriptive, natural language paragraph. The resulting linearized text passages may enable LLMs to better comprehend and integrate …
Long-Life Index Protocol — Open Specification Of Incentive Smart Contracts For Durability And Predictive Maintenance Of Physical Assets, Xavier Pillet
Long-Life Index Protocol — Open Specification Of Incentive Smart Contracts For Durability And Predictive Maintenance Of Physical Assets, Xavier Pillet
Defensive Publications Series
This document, produced with the assistance of ChatGPT o3 and ChatGPT 5 Thinking, is released under the Apache 2.0 licence. It is a voluntary defensive publication (prior art) and therefore enters the prior art upon release under the applicable patent statutes : EPC Art. 54(2) (European Patent Convention), French IPC Art. L 611-11 (CPI), 35 U.S.C. §102(a) (United States Patent Act), Chinese Patent Law Art. 22(5) (中华人民共和国专利法), and Japanese Patent Act Art. 29(1) (特許法). It discloses 100 enabling inventions across devices & sensors, assays & QA, algorithms (TinyML/FL), closed-loop controllers, materials & packaging, NDT/imaging, privacy/standards, UX/workflows, and supply-chain. Proposals cover …
Cloud Audits: A Technical Framework For Financial Institutions In A Multi-Cloud Era, Ramkumar Bharathan
Cloud Audits: A Technical Framework For Financial Institutions In A Multi-Cloud Era, Ramkumar Bharathan
Defensive Publications Series
This paper provides a comprehensive, forward-looking analysis of technology auditing within the financial services sector, specifically for multi-cloud infrastructures encompassing AWS, Azure, and Google Cloud Platform (GCP). It presents a detailed technical framework for modern audit practices, moving from traditional, reactive methodologies to a proactive, continuous assurance model. The framework details technical audit procedures for core cloud domains including Identity and Access Management (IAM), data protection, and network security, with a focus on provider-specific tooling and logging. It also provides a strategic roadmap for auditing emerging technologies such as Artificial Intelligence (AI), Distributed Ledger Technology (DLT), and Post-Quantum Cryptography (PQC). …
Interplanetary Rail Launcher: Open Architecture And Free Technology Bricks (Magseal, Mhd, Hts) – Full Report And Associated Data, Xavier Pillet
Interplanetary Rail Launcher: Open Architecture And Free Technology Bricks (Magseal, Mhd, Hts) – Full Report And Associated Data, Xavier Pillet
Defensive Publications Series
This document, produced with the assistance of ChatGPT o3 and ChatGPT 5 Thinking, is released under the Apache 2.0 licence. It is a voluntary defensive publication (prior art) and therefore enters the prior art upon release under the applicable patent statutes: EPC Art. 54(2) (European Patent Convention), French IPC Art. L 611-11 (French Intellectual Property Code), 35 U.S.C. §102(a) (United States Patent Act), Chinese Patent Law Art. 22(5) (中华人民共和国专利法), and Japanese Patent Act Art. 29(1) (特許法). It discloses 140+ enabling inventions for an interplanetary rail-launcher: MagSeal liquid-metal gates, large-Ø plasma windows, staged MHD vacuum pumps, deployable ReBCO shields, adaptive MHD …
Social-Impact Technologies For Fair Wealth Distribution — 1990–2025, Xavier Pillet
Social-Impact Technologies For Fair Wealth Distribution — 1990–2025, Xavier Pillet
Defensive Publications Series
This document, produced with the assistance of ChatGPT o3 and ChatGPT 5 Thinking, is released under the Apache 2.0 licence. It is a voluntary defensive publication (prior art) that enters the public domain of prior art upon release under the applicable patent statutes: EPC Art. 54(2) (European Patent Convention), French IPC Art. L 611-11 (French Intellectual Property Code), 35 U.S.C. §102(a) (United States Patent Act), Chinese Patent Law Art. 22(5) (中华人民共和国专利法), and Japanese Patent Act Art. 29(1) (特許法). The compendium enumerates 114 system-level, enabling inventions across inclusive fintech, digital identity, energy microgrids/DC, health/edtech, agriculture, logistics, and civic tech. Each proposal …
Browser-Integrated Bidding System For Third-Party Ai Agent Tools, Jason Mayes
Browser-Integrated Bidding System For Third-Party Ai Agent Tools, Jason Mayes
Defensive Publications Series
Currently, artificial intelligence agents may need to use multiple proprietary tools from different third parties to complete a complex user-assigned task. This creates a fragmented process for payment and tool selection, as there is no centralized mechanism for discovering, comparing, and paying for these agentic tools. In an example implementation, a web browser is configured to facilitate a bidding process for third-party agent tools. For example, when a specific function is required by an agent, suitable tools are identified by the web browser from a central marketplace or via indexes on websites. These tool providers can then bid for the …
Pulsepower — Public-Domain Release Of Open Specifications For Bn-Diamondoid Capacitor, Self-Healing Cu-Graphene Busbar And Opto-Fmc, Xavier Pillet
Pulsepower — Public-Domain Release Of Open Specifications For Bn-Diamondoid Capacitor, Self-Healing Cu-Graphene Busbar And Opto-Fmc, Xavier Pillet
Defensive Publications Series
This document, produced with the assistance of ChatGPT o3 and ChatGPT 5 Thinking, is released under the Apache 2.0 licence. It is a voluntary defensive publication (prior art) and enters the prior art upon release under the applicable patent statutes: EPC Art. 54(2) (European Patent Convention), French IPC Art. L 611-11 (French Intellectual Property Code), 35 U.S.C. §102(a) (United States Patent Act), Chinese Patent Law Art. 22(5) (中华人民共和国专利法), and Japanese Patent Act Art. 29(1) (特許法). The work discloses enabling architectures, materials, controls, QA, imaging, scheduling, and service workflows for ultra-high-power pulsed systems: BN-diamondoid capacitors, self-healing Cu-graphene busbars, and opto-FMC high-voltage …
On-The-Fly Adaptation Of Macromoe Llms Using Task-Specific Low-Rank Adapters, Johannes Start, John Lunney
On-The-Fly Adaptation Of Macromoe Llms Using Task-Specific Low-Rank Adapters, Johannes Start, John Lunney
Defensive Publications Series
A mixture of experts (MoE) is a way of building machine learning models (including large language models) that improves efficiency and scalability by activating only a subset of the model parameters during each inference step. The activated model-parameter subset is an expert in the domain of the input query. This disclosure describes techniques for efficient, dynamic specialization of LLMs that are based on a macro-MoE architecture. A plurality of foundational, pre-trained LLMs (‘macro-experts’), a library of distinct low-rank adaptation (LoRA) modules, and an adaptive gating network is provided. Upon receiving an input prompt, the gating network analyzes the prompt and …
Enhanced Optical Supervisory Channel (Osc) Safety Verification For Automatic Laser Shutoff Systems, Anonymous
Enhanced Optical Supervisory Channel (Osc) Safety Verification For Automatic Laser Shutoff Systems, Anonymous
Defensive Publications Series
This disclosure addresses critical safety vulnerabilities associated with Optical Supervisory Channel (OSC) systems used for Automatic Laser Safety Shutoff (ALSO) in optical networks. Existing methods assume correct OSC connectivity based solely on manual provisioning, which poses significant risks of unsafe amplifier operation during fiber disruptions. This disclosure introduces two novel methods—OSC_noConfig and OSC Connection Validation (CV)—to robustly verify OSC connectivity. The OSC_noConfig method automatically detects missing OSC provisioning, simplifying safety shutoff criteria to a single-term threshold, thereby ensuring safe amplifier shutdown even when OSC messages are unreliable. The OSC CV method utilizes precise on-off keying signaling to confirm correct OSC …
Ultra-Light Diamond Lattices For Orbital And Terrestrial Uses: State Of The Art, 1g/0g Technological Feasibility And Strategic Roadmap (2025–2035), Xavier Pillet
Defensive Publications Series
This document, produced with the assistance of ChatGPT o3 and GPT-5 Thinking, is released under the Apache 2.0 licence. It is a voluntary defensive publication (prior art) and therefore enters the prior art upon release under the applicable patent statutes: EPC Art. 54(2) (European Patent Convention), French IPC Art. L 611-11 (French Intellectual Property Code), 35 U.S.C. §102(a) (United States Patent Act), Chinese Patent Law Art. 22(5) (中华人民共和国专利法), and Japanese Patent Act Art. 29(1) (特許法). It discloses 103 enabling inventions spanning ISRU materials (HIBE, ENZ–THz, graphene–regolith aerogels), orbital manufacturing, rail-launch subsystems (vacuum/diaphragms, MHD shields, cryo-tribology), QA/standards, logistics and controls. Each …
Next-Generation Isru Materials For Space Infrastructure: Hibe, Enz–Thz And Graphene–Regolith Aerogel, Xavier Pillet
Next-Generation Isru Materials For Space Infrastructure: Hibe, Enz–Thz And Graphene–Regolith Aerogel, Xavier Pillet
Defensive Publications Series
This document, produced with the assistance of ChatGPT o3 and GPT-5 Thinking, is released under the Apache 2.0 licence. It is a voluntary defensive publication (prior art) and therefore enters the prior art upon release under the applicable patent statutes: EPC Art. 54(2) (European Patent Convention), French IPC Art. L 611-11 (French Intellectual Property Code), 35 U.S.C. §102(a) (United States Patent Act), Chinese Patent Law Art. 22(5) (中华人民共和国专利法), and Japanese Patent Act Art. 29(1) (特許法). It discloses 103 enabling inventions spanning ISRU materials (HIBE, ENZ–THz, graphene–regolith aerogels), orbital manufacturing, rail-launch subsystems (vacuum/diaphragms, MHD shields, cryo-tribology), QA/standards, logistics and controls. Each …
Open Blueprint: 2025–2040 Roadmap For A Circular Orbital Megastation — Rebco Maglev Ring, 6d Spirulina–Mycelium Photobioreactor And Mof-Li Water–Air Loop, Xavier Pillet
Defensive Publications Series
This document, produced with the assistance of ChatGPT o3 and GPT-5 Thinking, is released under the Apache 2.0 licence. It is a voluntary defensive publication (prior art) and therefore enters the prior art upon release under the applicable patent statutes: EPC Art. 54(2) (European Patent Convention), French IPC Art. L 611-11 (French Intellectual Property Code), 35 U.S.C. §102(a) (United States Patent Act), Chinese Patent Law Art. 22(5) (中华人民共和国专利法), and Japanese Patent Act Art. 29(1) (特許法). It discloses 160 enabling inventions for a circular orbital megastation: REBCO maglev gravity ring (mechanics, cryo, safety), 6D spirulina–mycelium photobioreactors (control, QA, imaging, biotech), and …
Framework For Governed Orchestration And Adaptive Lifecycle Management Of Hierarchical Artificial Intelligence Agents, Mithun Kumar S R
Framework For Governed Orchestration And Adaptive Lifecycle Management Of Hierarchical Artificial Intelligence Agents, Mithun Kumar S R
Defensive Publications Series
This document describes a framework for governing and managing hierarchical artificial intelligence (AI) agent systems. The framework addresses various limitations of current approaches, such as static architectures, naive delegation, high operational costs, lack of auditability, security vulnerabilities, and poor conflict handling. The framework includes a primary orchestrator node (PON) for task decomposition, a delegation and provisioning engine (DAPE) for agent selection and provisioning, specialized execution agents (SEAs) for task execution in secure sandboxes, a shared context manifold (SCM) for collaboration, a synthesis and reconciliation module (SRM) for output assembly, a causal traceability ledger for auditability, and a performance-driven optimization loop …
Analysis Of User Interaction Events For Conversion Matching Across Products, Ozgur Yonter, Arun Mathew, Alistair Hughes, Zhusong Li
Analysis Of User Interaction Events For Conversion Matching Across Products, Ozgur Yonter, Arun Mathew, Alistair Hughes, Zhusong Li
Defensive Publications Series
Granular user interaction data can be used to guide product development, customize products per user preferences, and other purposes that can provide user benefits. However, it is difficult to efficiently aggregate and compare user interaction data from multiple different products and across different data dimensions in a privacy-preserving manner. This disclosure describes privacy-compliant techniques that generate insights into user behavior by analyzing user journeys cross-contextually across diverse products (including interaction via different modalities and user interfaces). Per the techniques, a conversion matching service (CMS) is built that offers flexibility in defining join algorithms that can be used for attribution. User-permitted …
Open Void-Fab Stack: Holo-Atomic Micro-G Printing, Real-Time Di!Ractive Ai, And Nano-Plasma Patch-Healing, Xavier Pillet
Open Void-Fab Stack: Holo-Atomic Micro-G Printing, Real-Time Di!Ractive Ai, And Nano-Plasma Patch-Healing, Xavier Pillet
Defensive Publications Series
This document, produced with the assistance of ChatGPT o3 and ChatGPT 5 Thinking, is released under the Apache 2.0 licence. It is a voluntary defensive publication and therefore enters the prior art upon release under the applicable patent statutes: EPC Art. 54(2) (European Patent Convention), French IPC Art. L 611-11 (French Intellectual Property Code), 35 U.S.C. §102(a) (United States Patent Act), Chinese Patent Law Art. 22(5) (中华人民共和国专利法), and Japanese Patent Act Art. 29(1) (特許法). This work discloses enabling concepts for orbital fabrication and in-situ repair: holo-atomic micro-g printing, real-time diffractive AI control, and a nano-plasma Patch-Healer torch, plus 12+ hybrid …
Regenerative Planet: Open-Source Catalog Of Trl 3–9 Technologies For Climate Resilience And Biodiversity Restoration – 2025 Edition, Xavier Pillet
Regenerative Planet: Open-Source Catalog Of Trl 3–9 Technologies For Climate Resilience And Biodiversity Restoration – 2025 Edition, Xavier Pillet
Defensive Publications Series
This document, produced with the assistance of ChatGPT o3 and ChatGPT 5 Thinking, is released under the Apache 2.0 licence. It is a voluntary defensive publication (prior art) covering thirty implementable innovations spanning devices, materials, processes, control algorithms, MRV stacks, and system integrations for climate resilience and biodiversity restoration. By public disclosure, it enters the state of the art under the applicable patent statutes : EPC Art. 54(2) (European Patent Convention), French IPC Art. L 611-11 (French Intellectual Property Code), 35 U.S.C. §102(a) (United States Patent Act), Chinese Patent Law Art. 22(5) (中华人民共和国专利法), and Japanese Patent Act Art. 29(1) (特許法). …
Equalitarian Dao V1.0 — Technical Specification & Source Code, Xavier Pillet
Equalitarian Dao V1.0 — Technical Specification & Source Code, Xavier Pillet
Defensive Publications Series
This document, produced with the assistance of ChatGPT o3 and ChatGPT 5 Thinking, is released under the Apache 2.0 licence. It is a voluntary defensive publication (prior art) and therefore enters the prior art upon release under the applicable patent statutes: EPC Art. 54(2) (European Patent Convention), French IPC Art. L 611-11 (French Intellectual Property Code), 35 U.S.C. §102(a) (United States Patent Act), Chinese Patent Law Art. 22(5) (中华人民共和国专利法), and Japanese Patent Act Art. 29(1) (特許法). This work discloses 48 enabling proposals across L1/L2 consensus and PQC (Stake & People 50/50, tree-signatures), fiscal protocols (polynomial rebase with ZK proof-of-execution), DeFi …
Custom Test Point Insertion For Improved Rom Test Coverage, Krunal Siddhapathak, Ankit Garg
Custom Test Point Insertion For Improved Rom Test Coverage, Krunal Siddhapathak, Ankit Garg
Defensive Publications Series
Read-only memories (ROM) incorporate a scan chain to assist in detecting manufacturing defects. This scan chain can share a single flip-flop across multiple ROM outputs. However, sharing of scan flip-flops across ROM outputs can create a controllability issue for the ROM fan-out logic. Testing ROM fan-out becomes challenging, as a single control element drives four outputs, each with potentially different requirements for fault excitation and detection. This disclosure describes techniques that insert custom test points to improve the controllability and observability of the fan-out and fan-in logic cones of read-only memories (ROMs). By placing test points at the ROM outputs, …
Generative Vision Bot: Democratizing Advanced Image Analysis With Genai For Non-Ml Experts, Wenfeng Gu, Zhen Zhu, Danutha Samael, Chin Hock Soh, Yee Ping Wong, Seng Hin Chaim, Randy Chinn, Muhammad Yazid Mohamed Kamil, Jun Lee Kok, Yonggang Zhang, Sthitie Bom
Generative Vision Bot: Democratizing Advanced Image Analysis With Genai For Non-Ml Experts, Wenfeng Gu, Zhen Zhu, Danutha Samael, Chin Hock Soh, Yee Ping Wong, Seng Hin Chaim, Randy Chinn, Muhammad Yazid Mohamed Kamil, Jun Lee Kok, Yonggang Zhang, Sthitie Bom
Defensive Publications Series
This disclosure introduces Generative Vision Bot, a self-service, prompt-driven image analysis platform powered by generative vision-language models. It enables non-ML professionals to perform advanced computer vision tasks such as defect detection, microscopy analysis, and scientific image interpretation without the need for large labeled datasets, coding expertise, or lengthy model development cycles. By leveraging few-shot and zero-shot learning, the system generates meaningful insights from raw images in under three hours using natural-language prompts and minimal annotation (zero to five examples). The platform includes task-specific templates (FREE, TAG, RACE, TRACE), automatic prompt optimization, and an intuitive interface designed for usability and …
Repository Of Standard Annotation Codes To Annotate Llm-Generated Software Code, David Pearl
Repository Of Standard Annotation Codes To Annotate Llm-Generated Software Code, David Pearl
Defensive Publications Series
Large language models are capable of automatically generating functional software code in response to user prompts. However, in current practice, LLM-generated code typically does not include comments and/or supporting information that links specific parts of the generated code to corresponding practices and/or intent of the end users. This disclosure describes a comprehensive, extensible repository of standard codes that can be used to annotate no-code or low-code software solutions where the application functionality is implemented using LLM-generated software code. The standard codes can be used to obtain LLM-generated code that includes appropriate details regarding end user practices and/or intent corresponding to …
Machine Learning Driven Ranking To Control Degree Of Personalization In Search Results, Eric Lin, Devora Berlowitz, Olive Guo
Machine Learning Driven Ranking To Control Degree Of Personalization In Search Results, Eric Lin, Devora Berlowitz, Olive Guo
Defensive Publications Series
An important problem in introducing personalization to query-based search retrieval and ranking is how to balance the relevance of results to the query with the affinity of the results to the user as oftentimes there is a trade-off between the two. This disclosure describes techniques to train a machine learning model to personalize ranking of search results in a controllable manner. With user permission, training data is constructed by analyzing query logs to identify manual query refinements followed by "long clicks," forming triplets of {original query, manually refined query, click}. These reflect personalized search results that the user found of …
Generating Shared Experience Visual Indicators From Aggregated Physiological Data, Sayantan Paul
Generating Shared Experience Visual Indicators From Aggregated Physiological Data, Sayantan Paul
Defensive Publications Series
A system for generating shared experience visual indicators can operate by collecting consented, limited physiological data from a group of users in a co-viewing environment. Such data may include heart rate information from wearable devices or audio data from a computing device (e.g., a smartphone, tablet) processed for emotion detection. The system can anonymize this data, for instance, by associating it with a temporary session identifier and then aggregating the anonymized data. If a collective response is detected across a configured threshold of users, such as a synchronized change in heart rate, the system can generate a subtle, optional visual …