19 results about "Data enrichment" patented technology

A method and system are disclosed for improving the accuracy, efficiency, and scalability of data interpretation and extraction of structured data from unstructured documents utilizing Large Language Models (LLMs). Applicable in finance, healthcare, legal, and government contexts, the disclosed invention addresses limitations of conventional Optical Character Recognition (OCR), machine learning, and LLM-based methods. In particular, the system and method incorporate feedback loops for continuous learning and leverage pre-processing, contextual tagging, customized prompt engineering, and post-processing to achieve robust data extraction. By integrating data enrichment techniques, the invention manages the inherent complexities of multilingual documents and evolving content standards.

A first application generates a first user interface to display a pricing quote to a user. The first application is configured to detect a request by the user for an enhanced view of the pricing quote. In response to detecting such a request, the first application retrieves first data from a second application and second data from a pricing and configuration service. Also, the first application populates a temporary database with the first data and the second data. Additionally, the first application performs data enrichment of the first data and the second data to create third data by translating technical information of the first data and the second data into a user-friendly notation. Finally, the first application generates a second user interface to display, to the user, the third data associated with the pricing quote.

An Al-driven system and method for healthcare lead nurturing are disclosed. A system includes a data enrichment module configured to integrate internal and external data sources to enhance customer profiles, and a stigma prediction engine employing machine learning algorithms to analyze behavioral patterns and predict stigma levels. The system further includes a multi-channel communication module for delivering personalized content via email, SMS, chat, and voice calls. The system further includes an Al audiologist virtual assistant for real-time interactions, configured to adapt dynamically based on user input.

The invention relates to the technical field of physical training and data processing, and discloses a physical training multi-dimensional data analysis system fused with radio and television technology, comprising a main time code reference module used for generating a global time code signal; the video acquisition and synchronization module is used for outputting a synchronous video stream carrying a timestamp; the context-aware network module is used for providing identity and position contexts of athletes; the athlete data unit is used for collecting athlete sensor data; and the central processing and analysis module adopts a multi-path merging and sorting algorithm, fuses all data streams into a single unified training event stream which is absolutely ordered according to time, and completes data enrichment. According to the method, through hardware-level synchronization and a unified event stream architecture, submillisecond-level accurate synchronization of multi-source data, automatic association of context information and simplification of a data processing architecture are realized.

The present invention, BCCBuddy, is an automated workflow management system that dynamically triggers multi-step workflows using email recipient fields (To, CC, and BCC). Unlike traditional automation tools relying on static triggers or basic email parsing, BCCBuddy executes multiple workflows simultaneously, each mapped to distinct recipient fields. The system incorporates advanced email parsing, a configurable storage system for workflow mapping, and optional data enrichment through natural language processing (NLP). By integrating seamlessly with third-party platforms such as CRM, task management, and scheduling tools, the invention supports high-volume email processing and complex automation scenarios. Its intuitive design ensures accessibility for technical and non-technical users alike, eliminating the need to modify email content or user behavior. BCCBuddy redefines email-driven automation by streamlining workflows, reducing manual effort, and enhancing productivity across industries and applications.

Systems and procedures are provided for extracting relations from textual data, comprising the collection (502) of labeled textual data from diverse sources, including digital archives and online repositories, each source comprising sentences annotated with detailed grammatical structures. Initial relational data are generated from the grammatical structures by applying advanced parsing and machine learning techniques using a sophisticated rule-based algorithm (506). Training sets are generated (508) to improve the diversity and complexity of a relational dataset by applying data enrichment techniques to the initial relational data. A neural network model is trained (510) using an array of semantically equivalent but syntactically varied prompt templates designed to test and refine the linguistic capabilities of a model.A final relation extraction output is determined (512) by implementing a voting-based decision system that integrates statistical analysis and uses a weighted voting mechanism to optimize extraction accuracy and reliability.

Methods, systems, and apparatus, including computer-readable media, for data preparation and modeling using artificial intelligence. In some implementations, a system receives user input that indicates one or more data sets. In response, the system generates a set of recommendations for data modeling, data preparation, or data enrichment for the one or more data sets, where at least one recommendation in the set of recommendations is generated using one or more artificial intelligence and / or machine learning (AI / ML) models. The system provides the set of recommendations for display in the user interface in association with one or more interactive controls to accept or dismiss the recommendations. In response to receiving user input the system updates the data model or the one or more data sets to apply an update corresponding to an accepted recommendation. The system provides the updated data set or updated data model to a chatbot or other application.

This disclosure provides a mechanism for the enrichment of sparse datasets using language models. By training language models on the specific distribution of known values in a dataset, missing values can be predicted, and the predicted values added, thereby resulting in a more complete dataset. This method also facilitates the enhancement and augmentation of datasets by predicting values for new properties that were not previously available. The approach proves particularly effective at scale, transforming large sparse datasets into more complete and enhanced datasets. Masking language modeling may be employed to train language models capable of generating representations of technical data. Training data includes corpuses of technical data that may be represented as text strings. These pretrained models are fine-tuned to predict various properties. The resulting models can predict missing values in large technical datasets, providing valuable data for guiding scientific research.

The present disclosure provides an approach of generating a request to obtain information corresponding to a data sample. The approach produces, by a processing device, sample metadata using an artificial intelligence (AI) model trained to analyze the data sample and generate the sample metadata. In turn, the approach enriches the data sample based on the sample metadata to produce an enriched data sample.

This disclosure provides a mechanism for the enrichment of sparse datasets using language models. By training language models on the specific distribution of known values in a dataset, missing values can be predicted, and the predicted values added, thereby resulting in a more complete dataset. This method also facilitates the enhancement and augmentation of datasets by predicting values for new properties that were not previously available. The approach proves particularly effective at scale, transforming large sparse datasets into more complete and enhanced datasets. Masking language modeling may be employed to train language models capable of generating representations of technical data. Training data includes corpuses of technical data that may be represented as text strings. These pretrained models are fine-tuned to predict various properties. The resulting models can predict missing values in large technical datasets, providing valuable data for guiding scientific research.

Data enrichment and identity translation techniques using probabilistic data structures are described. In one or more examples, a plurality of datasets are received from a plurality of entities. Each dataset has a plurality of dataset records describing a respective audience. A plurality of sets of sketches are generated as probabilistic data structures, respectively, based on the plurality of datasets. A result is formed by processing a query. The result includes at least one sketch having a probabilistic data structure generated based on one or more of the plurality of sets of sketches. An entity is identified from the plurality of entities corresponding to the at least one sketch and entity is exposed for display in a user interface.

Embodiments of the present application disclose a data enrichment method and device for an engine and an electronic device. After obtaining historical test data of a target engine, the historical test data is input into a pre-trained data enrichment model to obtain target test data output by the data enrichment model. The target test data includes at least one target parameter of the target engine that is not measured in historical tests. The data of the target engine that is not measured in the historical tests under corresponding historical working conditions can be quickly obtained without repeating the historical tests of the target engine, and the method is applicable to engines of any type.

The invention discloses a component basic information data cleaning method, and belongs to the technical field of data management. The method comprises the following steps: formulating basic information data standard specifications of components; constructing a data dictionary database; performing standardized preprocessing on the multi-source heterogeneous component data; performing basic attribute cleaning of uniqueness verification based on a dictionary rule mapping table; performing multi-condition comparison cleaning on the basic attributes of the non-dictionary class; merging, de-duplication and endowing unified codes; and data enrichment and system online are realized. According to the invention, the problems of multi-source isomerism, non-uniform standard and low cleaning efficiency of component data in the prior art are solved, efficient and accurate component basic information data cleaning is realized, and high-quality data support is provided for a digital supply chain system.

Techniques for performing sequential data enrichment are presented herein. In some examples, a data enrichment system includes an input pipeline and a set of enrichment routines. In some examples, a datagram is retrieved from the input pipeline and an enrichment type associated with the retrieved datagram is determined. In some examples, after determining the enrichment type that is associated with the retrieved datagram, the datagram is processed by the enrichment routine that corresponds to the enrichment type to determine an enriched datagram. In some examples, the enriched datagram is stored in the input pipeline, which can enable determining a new enrichment type associated with the enriched datagram or storing the enriched datagram in an output pipeline if it is determined that the enriched datagram is not associated with any enrichment types.

A data enrichment method includes: receiving event information associated with an event in a target of interest; based at least in part on the received event information, analyzing data associated with the target of interest to identify: an impact of the event within the target of interest, and existing one or more mitigation measures associated with the event; and outputting information about an event, the output information including the identified impact and the identified one or more mitigation measures, where the event information includes information about an abnormal behavior detected in the target of interest or information about a vulnerability detected in the target of interest, and wherein the analyzed data is based on a plurality of sources.

A computer system for organizational maturity and competency development comprising: a processor; memory; computer-readable instructions that cause the system to: receive a dataset concerning organizational goals; generate a maturity acceleration plan (MAP) with maturity metrics for entities; generate performance acceleration models (PAMs) based on the MAP, each PAM can include measurement criteria representing knowledge, competence, and proficiency; store a hierarchical data structure with MAP at first level, PAMs at second level, and accolades at third level; monitor user interactions detecting accolade completion by tracking training course completion, platform feature usage, data element creation, data enrichment, and specified actions; display a visual dashboard showing organizational maturity progress, deployed MAPs, and completed PAMs. The system can provide an AI algorithm with pre-defined parameters and organization rules to suggest additional aspects of the maturity acceleration plan and utilizes the AI algorithm to provide a new or modified maturity acceleration plan.

Systems and methods are disclosed for implementing a low-latency data stream monitoring system. The data stream monitoring system may obtain raw data from a data source as soon after the data is generated, and may classify the data according to different topics. The topics may be published in a publish-subscribe messaging model, and data enrichment systems may subscribe to the topics to receive data for enrichment. The data enrichment systems may supplement or replace the raw data with additional information, and may further classify or reclassify the enriched data into different topics. The enriched data may then be published to an alert generation system, which may apply various criteria to the enriched data to determine that alerts should be generated, generate the alerts, and publish or transmit the alerts to client devices. Individual data streams, topics, enrichments, criteria, and alarms may be added, removed, or modified as required.

The application relates to the technical field of network operation and maintenance, in particular to an automatic operation and maintenance method and system based on artificial intelligence. The method collects detailed log information containing security events in real time through a security information and event management system configured at a target endpoint, then uses an agent artificial intelligence system to clean, label and structure process the log data, generates standardized metadata, maps the metadata with an MITRE ATT&CK knowledge base, identifies the technical features and behavior patterns of attacks, compares the enriched log data with external threat intelligence sources, analyzes the TTP of known attack groups, generates a threat intelligence correlation report, uses a large language model to generate a targeted response plan, generates an executable command sequence according to the response plan, connects the agent executor with the server through a WebSocket protocol, executes the command in a POSIX shell environment, captures and returns the execution result, verifies the execution result, performs necessary command optimization, records the optimized response to a vector database, automatically matches historical events through a vector retrieval mechanism, triggers a predefined response process, and realizes continuous threat monitoring and adaptive response. The application can solve the problem that the existing security operation and maintenance cannot form a closed loop of terminal executor, data enrichment and historical event recall.

A method, system and computer program product of processing a document from an electronic_historical digital archive and enabling electronic semantic searching, may include: electronic_pre-processing the electronic_document; electronically enriching, and enabling electronically_semantic searching of the electronic document, which may digitize and make machine readable. Electronic pre-processing is a quality check of the process and not of the results. Electronic pre-processing means the translation / transformation quality of the document may or may not require alterations after translation / transformation. Electronic p re-processing embodiments may make searchable by LLMs, electronic score quality of electronically searchable document, electronically update the electronically searchable document to create a better score quality, and may include retrieval-augmented generation (RAG)-encoding, embedding LLM scoring vector and object semantics; and / or store in, an electronic database such as, e.g., an electronic database capable of storing vector data. Embodiments may receive an electronic semantic search query; provide electronic search results having LLM scoring vector and object semantics embeddings corresponding comparingly to the semantic search, and electronically enable creating software data enrichment agents to enable electronic searching for interesting historical facts; electronically building inferences with or without electronic prompts; or electronically_discovering new history such as, e.g., electronically identifying at least one connection between a plurality of said documents in the electronic historical digital archive.