A method implements static profiling with graph neural networks. The method includes executing a block model with a control flow graph to generate a block vector corresponding to a block of the control flow graph of source code. The method further includes executing a graph neural network model with the control flow graph and the block vector to generate a graph vector. The method further includes executing a feed-forward neural network with the graph vector to generate a branch-frequency prediction. The method further includes incorporating the branch-frequency prediction into a code profile.
User-level privacy preservation is implemented within federated machine learning. An aggregation server may distribute a machine learning model to multiple users each including respective private datasets. Individual users may train the model using the local, private dataset to generate one or more parameter updates. Prior to sending the generated parameter updates to the aggregation server for incorporation into the machine learning model, a user may modify the parameter updates by applying respective noise values to individual ones of the parameter updates to ensure differential privacy for the dataset private to the user. The aggregation server may then receive the respective modified parameter updates from the multiple users and aggregate the updates into a single set of parameter updates to update the machine learning model. The federated machine learning may further include iteratively performing said sending, training, modifying, receiving, aggregating and updating steps.
A system and method for extracting structured key information from diverse document types using large multimodal models (LMMs) is disclosed. The invention employs a zero-shot analysis to identify candidate keys within an input document, then selects a document schema from a document schema database based on the identified keys. The LMM is prompted with the selected document schema to generate structured key-value pairs, with field constraints enforced by the document schema. Relationships among extracted keys are mapped to a graph representation, enabling robust handling of complex document layouts. The system supports nested structures, tabular data, and alias definitions for fields, and can update document schemas based on ground truth feedback. The resulting structured output is provided in a machine-readable format, enabling reliable and scalable document understanding across varied domains such as invoices, health cards, and driving licenses.
Systems, methods, and other embodiments associated with efficient allocation of live connections for real-time transcriptions of virtual meetings are described. In one embodiment, an example method includes preemptively establishing a set of live connections to an automatic speech recognition service that are available for use, and fewer than the participants of a virtual meeting. In response to a participant of the virtual meeting becoming active, the method dedicate one WebSocket connection from the set of WebSocket connections to real-time transcription of an individual audio stream from the participant. The method labels transcription results received back through the one live connection with a username of the participant. And, the method injects the labeled transcription results back into the virtual meeting for display in a user interface.
Hierarchical gradient averaging is performed as part of training a machine learning model to enforce subject level privacy. A sample of data items from a training data set is identified and respective gradients for the data items are determined. The gradients are then clipped. Each subject's clipped gradients in the sample are averaged. A noise value is added to a sum of the averaged gradients of each of the subjects in the sample. An average gradient for the entire sample is determined from the averaged gradients of the individual subjects with the added noise value. This average gradient for the entire sample is used for determining machine learning model updates.
Skew handling techniques are provided in parallel execution for even load balancing and scaling. In a compile-time solution, a dynamic sampling query is issued to detect partition skew. The compile-time solution determines the number of skewed partitions and uses a hybrid distribution scheme where skewed partitions use a random distribution and non-skewed partitions use the original server mapping. In a runtime solution, producer server processes create partition mapping vectors that contain partition mapping information. Each producer server process sends its partition mapping vector to the query coordinator (QC). The QC receives the partition mapping vectors from the producer server processes, merges the vectors, and determines a skew result based on the merged mapping vectors and sends the skew result to the producer server processes. The producer server process can alter distribution of skewed partitions based on the skew result.
G06F 16/27 - Réplication, distribution ou synchronisation de données entre bases de données ou dans un système de bases de données distribuéesArchitectures de systèmes de bases de données distribuées à cet effet
H04L 67/1008 - Sélection du serveur pour la répartition de charge basée sur les paramètres des serveurs, p. ex. la mémoire disponible ou la charge de travail
7.
Defining And Using Reusable Modules To Generate Form Control Code
Techniques for defining and using reusable modules to generate form control code are disclosed, including: displaying a form control implementation interface for applying form control functions to forms; receiving via the form control implementation interface: a first user input selecting a form control function of the form control functions; a second user input selecting one or more input parameters, for the form control function, that are to be extracted from the target form; a third user input selecting a target field of a target form, one or more attributes of the target field to be modified via execution of the form control function; generating form control code that extracts the one or more input parameters from form data received for the target form and applies the form control function to the one or more input parameters to modify the one or more attributes of the target field.
An approach of performing data center failover using an address that indicates a backup data center. The address includes common names indicating a data center with a domain and a backup datacenter with a replica of the domain. A cloud service provider can receive the address, establish a connection with an available data center, and failover to the backup data center if the data center with the connection becomes unavailable.
G06F 11/20 - Détection ou correction d'erreur dans une donnée par redondance dans le matériel en utilisant un masquage actif du défaut, p. ex. en déconnectant les éléments défaillants ou en insérant des éléments de rechange
G06F 11/07 - Réaction à l'apparition d'un défaut, p. ex. tolérance de certains défauts
9.
MECHANISMS FOR LOGICAL REFERENTIAL INTEGRITY IN RELATIONAL DATABASE MANAGEMENT SYSTEMS
Techniques provide mechanisms for logical referential integrity in relational database management systems. A child record with a foreign key can be inserted into a child table out of order from inserting parent record with the primary key without foreign key violations. When the child record is inserted a referential integrity check is performed to determine whether the primary key exists in a parent table. An integrant field for each child record is set to indicate whether the primary key exists. Query results can filter out records that do not have referential integrity.
Various embodiments of the present technology generally relate to systems and methods for providing a subscription resource engine to suppress transmission of event notifications to originating network functions (NFs). For example, a subscription resource engine may detect an event operation associated with a consumer NF within a network. The consumer NF may be subscribed to a producer NF such that the subscription allows the consumer NF to receive event notifications from the producer NF. The subscription resource engine may determine that the event operation is related to the subscription and generate a subscription resource header based on the subscription. Based on the subscription resource header, the subscription resource engine may then generate an indication to suppress transmission of an event notification to the consumer NF. The indication to suppress transmission of the event notification may cause the producer NF to skip transmission of the event notification to the consumer NF.
H04L 51/212 - Surveillance ou traitement des messages utilisant un filtrage ou un blocage sélectif
H04W 8/18 - Traitement de données utilisateur ou abonné, p. ex. services faisant l'objet d'un abonnement, préférences utilisateur ou profils utilisateurTransfert de données utilisateur ou abonné
11.
Maintaining Relevant Communication History In Association With An Entity For Concurrent Display Of The Communication History And The Entity
Techniques for concurrently presenting a data object and messages related to the data object are disclosed. The system accesses a data object comprising a set of content. The system extracts and displays the set of content from the data object in a Graphical User Interface (GUI). The system determines that a set of messages is stored in association with the data object. The system presents the set of messages concurrently in the GUI with the set of content based on (a) the set of content from the data object being displayed and (b) the set of messages being stored in association with the data object. When presenting a different set of content, the system presents another set of messages associated with the different set of content. The system maps a new message to a data object when the new message is associated with the data object.
Techniques for dynamically resizing frames in a graphical user interface (GUI) window are disclosed. The system presents a GUI window on a display screen. The system presents, within the GUI window, a first frame that displays a first set of content and a second frame that displays a second set of content. The frames share space within the GUI window. The system detects a content transition, modifying the first set of content, corresponding to the first frame. In response to the content transition, the system increases the size of the first frame and decreases the size of the second frame without changing the size of the GUI window.
G06F 3/04845 - Techniques d’interaction fondées sur les interfaces utilisateur graphiques [GUI] pour la commande de fonctions ou d’opérations spécifiques, p. ex. sélection ou transformation d’un objet, d’une image ou d’un élément de texte affiché, détermination d’une valeur de paramètre ou sélection d’une plage de valeurs pour la transformation d’images, p. ex. glissement, rotation, agrandissement ou changement de couleur
13.
ML-Based Selection Of Components Of A Graphical User Interface Based On User-Defined Queries
Techniques for information to present to a user in response to a data type of a query result for a query received from the user are disclosed. The system executes the user provided query to generate the query result. Based on (a) the first data type of the query result and/or (b) characteristics of the user, the system selects a second data type and determines one or more values corresponding to the second data type that are not responsive to the first query. The system concurrently presents the query result corresponding to the first data type and the set of one or more values corresponding to the second data type.
Techniques are disclosed for consistent and scalable replication between source and target data stores in heterogeneous data environments. In one aspect, a method includes receiving, by a data storage system, a source data write from a source data store. The source data write is executed on a replica data store and a transaction is determined based on one or more data operations of the source data write. A router identifies one or more materializers based on the transaction and a mapping between a first schema and a second schema. The materializers generate one or more semantic objects based on the transaction, and the semantic objects are transmitted to target data stores. For each semantic object, the materializers generate a watermark based on a replica data store read timestamp. The data storage system may receive multiple writes causing race conditions that are resolved based on the watermarks and read timestamps.
Techniques are disclosed for touch-aware authorization and access control in hybrid data systems, including data systems supporting hybrid relational-document data models. In one aspect, a method includes receiving a query and determining a data path based on the query. The data path can include a set of touched paths of data in a data system. A touched path of the set of touched paths can be used to access a different touched path of the set of touched paths. Each touched path can be evaluated based on one or more access control policies to determine whether at least one touched path violates one or more access control policies. If at least one touched path violates one or more access control policies, access control of the data can be enforced by controlling the execution of the query on the data system.
A computer-implemented method includes receiving a query in natural language, generating an input for a large language model, the input including a prompt generated based on the query, and identifying a plurality of slots associated with a plurality of sections of a content item. The method further includes generating a query result based on the input, the query result including a subset of the plurality of slots selected, extracting one or more document chunks from a database storing a plurality of document chunks as one or more relevant document chunks associated with the query result, formatting the relevant document chunks into a response to the query, and providing the response to a client system. The plurality of document chunks is generated by dividing each content item of a plurality of content items into the plurality of document chunks based on sections within each content item.
G16H 10/60 - TIC spécialement adaptées au maniement ou au traitement des données médicales ou de soins de santé relatives aux patients pour des données spécifiques de patients, p. ex. pour des dossiers électroniques de patients
17.
EFFICIENT QUERYING WITH DIVERSELY ENCODED CLINICAL DATA
Techniques are disclosed for querying with semantic code expansion in a clinical data system. In one aspect, a method includes receiving a query containing a predicate specifying a clinical code and a semantic expansion parameter indicating a request for approximate matching. A vector embedding associated with the specified clinical code is retrieved from a pre-computed embedding index. A similarity search is performed in a vector space to identify semantically similar codes. Exact code mappings are retrieved for the specified clinical code from a mapping registry. A rewritten query predicate is generated include the exact code mappings and the semantically similar clinical code mappings. The rewritten query is executed against a clinical data store to retrieve results matching the exact and/or semantically similar codes. The results are annotated to distinguish between exact and semantic matches.
G16H 50/70 - TIC spécialement adaptées au diagnostic médical, à la simulation médicale ou à l’extraction de données médicalesTIC spécialement adaptées à la détection, au suivi ou à la modélisation d’épidémies ou de pandémies pour extraire des données médicales, p. ex. pour analyser les cas antérieurs d’autres patients
G06F 16/22 - IndexationStructures de données à cet effetStructures de stockage
Techniques are disclosed for constraint-driven query routing in heterogeneous data environments with disparate data stores. In one aspect, a method includes receiving a query in a first programming language and associated with one or more constraints. The constraints can include freshness, feasibility, divergence, and/or execution time. An intent of the query is identified, and a dry run of the query is performed to evaluate whether a data store satisfies the constraints. An optimal data store is selected based on the dry run. A query result is generated by determining whether the query in the first programming language can be executed on the optimal data store. The query is executed on the optimal data store if the first programming language is executable on the optimal data store. Otherwise, the query is converted to a second query in a second programming that can be executed on the optimal data store.
A computer-implemented method includes receiving a query in natural language, generating an input for a generative large language model, the input including a prompt generated based on the query, and identifying a plurality of slots associated with a plurality of sections of a content item. The method further includes generating a query result based on the input, the query result including a subset of the plurality of slots selected, extracting one or more document chunks from a database storing a plurality of document chunks as one or more relevant document chunks associated with the query result, formatting the relevant document chunks into a response to the query, and providing the response to a client system. The plurality of document chunks is generated by dividing each content item of a plurality of content items into the plurality of document chunks based on sections within each content item.
G16H 10/60 - TIC spécialement adaptées au maniement ou au traitement des données médicales ou de soins de santé relatives aux patients pour des données spécifiques de patients, p. ex. pour des dossiers électroniques de patients
20.
SYSTEMS AND METHODS FOR GENERATING CLINICAL HANDOFF SUMMARIES
A computer-implemented method includes receiving a query to provide a summary of patient-specific information regarding a condition for a particular patient. The method includes determining a category for the query, retrieving data relevant to the query from an electronic health record (EHR) database, including at least structured and unstructured content, and processing and filtering the data as retrieved based on the category. The method further includes generating, by a generative machine learning model, a narrative summary including a first portion of filtered data and some of the unstructured content, generating a structured summary including a second portion of filtered data, including some of the structured content, and formatting the narrative summary and the structured summary into an output. Determining the category for the query includes selecting the category from a plurality of categories, and processing performed for a first category differs from processing performed for a second category.
G16H 10/60 - TIC spécialement adaptées au maniement ou au traitement des données médicales ou de soins de santé relatives aux patients pour des données spécifiques de patients, p. ex. pour des dossiers électroniques de patients
G06F 16/9038 - Présentation des résultats des requêtes
21.
AGENT EXECUTION PLANNING FOR AGENTIC DIGITAL ASSISTANT
Agentic digital assistant methods and systems for generating a response to a user query are disclosed. A computer-implemented method includes accessing a query, obtaining an agent execution plan that identifies one or more agent actions to be executed and an order in which the one or more agent actions are to be executed, executing the agent execution plan to obtain one or more results for the one or more agent actions, and generating a response to the query using the one or more results.
Agentic digital assistant methods and systems for generating a response to a user query are disclosed. A computer-implemented method includes accessing a query, executing planner modules in parallel to generate respective executable actions to retrieve information for answering the query, using a primary planner module of the planner modules to generate an execution plan for executing the executable actions, executing the executable actions per the execution plan to generate a set of results for the executable actions, and generating a response to the query using the set of results.
Computer-implemented techniques are disclosed for constructing, augmenting, and utilizing graph-structured or other linked representations of data elements and associations derived from one or more sources to enable accurate, timely enrichment and analysis across multi-stage or other processing workflows. An intermediate representation of patient-specific data can be obtained. The intermediate representation can be processed to extract condition-related and medication-related information relevant to a patient encounter. Outputs can be processed to further filter and contextualize subsets of the condition-related and medication-related information. One or more filtering techniques including a knowledge-graph-based filtering technique can be applied. A clinical summary that includes facts derived from the intermediate representation can be generated.
G06F 16/901 - IndexationStructures de données à cet effetStructures de stockage
G06F 16/28 - Bases de données caractérisées par leurs modèles, p. ex. des modèles relationnels ou objet
G06N 3/042 - Réseaux neuronaux fondés sur la connaissanceReprésentations logiques de réseaux neuronaux
G16H 15/00 - TIC spécialement adaptées aux rapports médicaux, p. ex. leur création ou leur transmission
G16H 50/70 - TIC spécialement adaptées au diagnostic médical, à la simulation médicale ou à l’extraction de données médicalesTIC spécialement adaptées à la détection, au suivi ou à la modélisation d’épidémies ou de pandémies pour extraire des données médicales, p. ex. pour analyser les cas antérieurs d’autres patients
24.
Using Physiological Sensor Data From Biomedical Sensors For Automated Task Status Tracking And Management
Techniques digitally tracking completion of user tasks based on physiological data obtained from biomedical sensors are disclosed. One or more embodiments manage a schedule of tasks for a specific user and presents the schedule to the user or to other users that have authorization to access the schedule (e.g., an authorized health care professional). In some embodiments, a system dynamically tracks the status of the tasks by obtaining physiological data from biomedical sensors of health monitoring devices. The physiological data may indicate if a particular task has been completed, as well as details related to the task. The system updates a record corresponding to the task based on the physiological data, thereby enabling the system to dynamically provide timely up-to-date tracking details regarding the schedule of tasks to the user.
In some aspects, techniques may include monitoring a primary load of a datacenter and a reserve load of the datacenter. The primary load and reserve load can be monitored by a computing device. The primary load of the datacenter can be configured to be powered by one or more primary generator blocks having a primary capacity, and the reserve load of the datacenter can be configured to be powered by one or more reserve generator blocks having a reserve capacity. Also, the techniques may include detecting that the primary load of the datacenter exceeds the primary capacity. In addition, the techniques may include connecting the reserve generator blocks to at least one of the primary generator blocks and the primary load using a computing device switch.
H02J 9/06 - Circuits pour alimentation de puissance de secours ou de réserve, p. ex. pour éclairage de secours dans lesquels le système de distribution est déconnecté de la source normale et connecté à une source de réserve avec commutation automatique
Techniques are described for managing secure connections (e.g., tunnels) between different endpoints using a pod of servers. Instead of computing devices connecting to a single server at a service IP address, the connections are spread among the different servers in the pod that can be reached using a public IP address.
A Productivity Assistant System (PAS) is described that uses specially-trained ML models (e.g., artificial neural networks (ANNs)) to predict a next action to be performed for a sequence of interactions made by a user with one or more applications or services. The predicted action is customized to that user or to a group of users to which the user belongs. Techniques are described for training and using one or more such machine learning models.
A source system migrates a virtual machine to a destination system by transferring an execution state of the virtual machine. To transfer the execution state, the source system generates a continuation element that includes a continuation capturing a state of a thread and a continuation root providing an entry point for resuming executing the thread at the state. Additionally, to transfer the execution state, the source system determines a set of elements that are reachable from the continuation root and generates a migration package that includes the continuation element and the set of elements. The migration package is transmitted to the destination system, and the destination system resumes executing the virtual machine at the execution state by loading the continuation and the set of elements and commencing executing the thread at the state based on the continuation and the set of elements.
G06F 9/455 - ÉmulationInterprétationSimulation de logiciel, p. ex. virtualisation ou émulation des moteurs d’exécution d’applications ou de systèmes d’exploitation
A system manages network connectivity in cloud environments. The system receives a high-level connectivity description comprising entities and flows representing connections between entities. Based on the description, the system derives a network configuration and applies it to configure resources in a cloud network environment. The system detects modifications, additions, or removals of resources in the cloud environment. In response to detected changes, the system modifies the network configuration to maintain consistency with the high-level description while accounting for resource alterations. The system generates a modified network configuration and applies it to at least one resource in the cloud environment.
Techniques for autonomous assignment of medical codes are disclosed. A natural-language health record is processed, to identify a portion of an extracted text. By applying a binary classification on the portion, a codability of the portion is identified. In response to a positive codability, two or more codes are assigned to the portion, by applying a multi-label classification to the portion. By applying a probability model, a first probability score indicative of a probability of a combination of the two or more codes being assigned to a single encounter is determined. By applying a language model, a second probability score indicative of the two or more codes assigned to the health record being correct is determined. A final probability score is assigned. In response to the final probability score being higher than a threshold, generation of an insurance record is caused, based on the two or more codes.
G16H 10/60 - TIC spécialement adaptées au maniement ou au traitement des données médicales ou de soins de santé relatives aux patients pour des données spécifiques de patients, p. ex. pour des dossiers électroniques de patients
G16H 15/00 - TIC spécialement adaptées aux rapports médicaux, p. ex. leur création ou leur transmission
The present disclosure describes solutions to a confused identity security vulnerability that can arise during multicloud operations. In an example method, a second cloud environment receives, from a first cloud environment, a request to perform an operation involving a service manager for a cloud service offered in the second cloud environment. The request can include a first identifier and a URL. The second cloud environment outputs a network request to a network location identified by the URL. The second cloud environment receives a response including a second identifier, which it compares to the first identifier. Upon determining, based upon the comparing, that the second identifier matches the first identifier, the second cloud environment performs processing enable performance of the operation involving the service manager. Upon determining that the second identifier does not match the first identifier, the second cloud environment rejects the request to perform the operation.
H04L 9/32 - Dispositions pour les communications secrètes ou protégéesProtocoles réseaux de sécurité comprenant des moyens pour vérifier l'identité ou l'autorisation d'un utilisateur du système
Agentic digital assistant methods and systems for generating a response to a user query are disclosed. A method includes receiving a query; identifying, from the input, one or more key phrases associated with one or more medical concepts; expanding the one or more medical concepts to include one or more other medical concepts that are related to the one or more medical concepts; identifying one or more portions of a database schema associated with the one or more medical concepts and the one or more other medical concepts; generating a prompt that comprises an instruction, the one or more portions of the database schema, and an utterance associated with the natural language component; transmitting the prompt to a machine learning model; receiving, from the machine learning model, a query result that includes information to answer the query; and providing the query result to the computing device associated with the user.
G16H 50/70 - TIC spécialement adaptées au diagnostic médical, à la simulation médicale ou à l’extraction de données médicalesTIC spécialement adaptées à la détection, au suivi ou à la modélisation d’épidémies ou de pandémies pour extraire des données médicales, p. ex. pour analyser les cas antérieurs d’autres patients
33.
SYSTEMS AND METHODS FOR GENERATING MEDICATION SUMMARIES
A computer-implemented method includes receiving a request to provide a summary of patient-specific information regarding a medication for a particular patient and retrieving data relevant to the request from a plurality of sources, the retrieved data including at least structured and semi-structured content, harmonizing the retrieved data for at least data structure and semantic alignment to generate harmonized data, filtering the harmonized data to extract fields in the harmonized data relevant to the request, generating an input for a generative machine learning model, the input including a prompt generated based on the request, generating, by the generative machine learning model, a query result associated with the input, the query result including a subset of the filtered, harmonized data, formating the query result into an output, the output including the summary of patient-specific information regarding the medication, and providing the output to a client system.
G16H 15/00 - TIC spécialement adaptées aux rapports médicaux, p. ex. leur création ou leur transmission
G16H 10/60 - TIC spécialement adaptées au maniement ou au traitement des données médicales ou de soins de santé relatives aux patients pour des données spécifiques de patients, p. ex. pour des dossiers électroniques de patients
G16H 20/10 - TIC spécialement adaptées aux thérapies ou aux plans d’amélioration de la santé, p. ex. pour manier les prescriptions, orienter la thérapie ou surveiller l’observance par les patients concernant des médicaments ou des médications, p. ex. pour s’assurer de l’administration correcte aux patients
G16H 50/70 - TIC spécialement adaptées au diagnostic médical, à la simulation médicale ou à l’extraction de données médicalesTIC spécialement adaptées à la détection, au suivi ou à la modélisation d’épidémies ou de pandémies pour extraire des données médicales, p. ex. pour analyser les cas antérieurs d’autres patients
34.
METHODS AND SYSTEMS FOR GENERATING A LOGICAL EXECUTION PLAN WITH DEPENDENCIES FOR EXECUTING AGENT ACTIONS IN AN AGENTIC DIGITAL ASSISTANT
Agentic digital assistant methods and systems for generating a response to a user query are disclosed. A method includes accessing a query regarding a person and database schema information for a database storing an electronic record for the person. An input for a generative machine learning model is generated based on the query and the database schema information. The input is provided to the generative machine learning model to an execution plan for executing actions in one or more stages to generate a set of results for use in generating a response to the query. The execution plan is executed to obtain the set of results. A response to the query is generated using the set of results.
Agentic digital assistant methods and systems for generating a response to a user query are disclosed. A computer-implemented method includes accessing a query, obtaining an agent execution plan that identifies one or more agent actions to be executed and an order in which the one or more agent actions are to be executed, executing the agent execution plan to obtain one or more results for the one or more agent actions, and generating a response to the query using the one or more results.
Techniques for generating knowledge-adapted content based on a knowledge classification of a user are disclosed. A system determines that a trigger condition is satisfied for requesting a knowledge-adapted content element for augmenting information for display on a user interface. In response to determining that the trigger condition is satisfied, the system generates, in real time, an input prompt element for requesting the knowledge-adapted content element and directs the input prompt element to a machine learning (ML) model to generate the knowledge-adapted content element. The knowledge-adapted content element includes machine-generated content pertaining to the target concept. The system receives the knowledge-adapted content element from the ML model and augments the information at least by concurrently displaying the information and machine-generated content on the user interface.
G16H 10/60 - TIC spécialement adaptées au maniement ou au traitement des données médicales ou de soins de santé relatives aux patients pour des données spécifiques de patients, p. ex. pour des dossiers électroniques de patients
37.
Selecting Attributes to Display for a Set of Search Results
Techniques for presenting a set of machine-learning predicted database record attributes for displaying in response to a request to filter a set of query results are disclosed. In response to a query, a system returns a set of system results. Based on receiving a selection of a query filter, the system presents a set of filtered query results. For a record in the set of filtered query results, the system presents a value for at least one default attribute and a value for at least one machine-learning predicted attribute. The system supplements the default attributes by applying a machine learning model to a set of filter data. The machine learning model generates a set of predicted attributes to present together with the default attributes for the set of query results.
G06F 16/38 - Recherche caractérisée par l’utilisation de métadonnées, p. ex. de métadonnées ne provenant pas du contenu ou de métadonnées générées manuellement
38.
CONCURRENT AND NON-BLOCKING OBJECT DELETION FOR CROSS-REGION REPLICATIONS
Techniques are described for enabling concurrent and non-blocking replication object deletion during cross-region replications. In some embodiments, in a target file system, a target replication pipeline as part of a cross-region replication, and a deletion pipeline operate in parallel. The deletion pipeline deletes processed objects reaching the last pipeline stage of the target replication pipeline after each checkpoint in the target replication pipeline. In some embodiments, after a non-recoverable failure during the cross-region replication, the cross-region replication can be restarted from the beginning (i.e., fresh restart) without waiting for its unused objects in the Object Store to be deleted by utilizing a generation number associated with each object to delete the unused objects in a background process while allowing deleting processed objects as normal for the freshly restarted cross-region replication.
G06F 11/14 - Détection ou correction d'erreur dans les données par redondance dans les opérations, p. ex. en utilisant différentes séquences d'opérations aboutissant au même résultat
G06F 3/06 - Entrée numérique à partir de, ou sortie numérique vers des supports d'enregistrement
39.
METHODS, SYSTEMS, AND COMPUTER READABLE MEDIA FOR ROUTING, ALTERNATE ROUTING, AND LOAD BALANCING OF SERVICE-BASED INTERFACE (SBI) REQUEST MESSAGES TO NETWORK FUNCTIONS (NFs) LOCATED IN DIFFERENT REGIONS USING SERVICE COMMUNICATION PROXY (SCP)
A method for routing, alternate routing, or load balancing of SBI request messages among NFs that are members of the same NF set and located in different regions using an SCP includes actively learning NF topology information of NFs that are located in the different regions. The method further includes receiving an SBI request message and determining an NF set-Id of a first target NF of the SBI request message, the first target NF being located in a first region. The method further includes using the NF set-Id and the NF topology information to identify at least one second target NF that is in the same NF set as the first target NF and located in a different region from the first region. The method further includes routing, alternate routing, or load balancing the SBI request message to a least one of the first target NF and the at least one second target NF.
H04W 28/084 - Équilibrage ou répartition des charges entre les entités de virtualisation des fonctions de réseau [NFV]Équilibrage ou répartition des charges entre les entités de calcul en périphérie, p. ex. calcul en périphérie multi-accès
40.
MEMORY ISOLATION OF AN IN-DATABASE VIRTUAL MACHINE USING MEMORY PROTECTION KEYS
Disclosed herein are approaches to isolate the execution of an embedded programming language virtual machine (VM) in a multi-tenant database management system (DBMS). At least a portion of a shared memory area may be associated with a memory protection key. A VM embedded in a database process of a DBMS may initiate execution of a user program. Execution of the database process may transition to a privileged mode, which may enable access to the at least a portion of the shared memory area by the VM. The VM may access the at least a portion of the shared memory area. Execution of the database process may transition to an unprivileged mode and disable access to the shared memory area by the VM. Further, a signal handler may receive a signal from a DBMS, wherein the signal interrupts a VM executing a user program in a database process, and the signal handler executes in the database process. The signal handler may write, to a protection key rights register for user pages (PKRU register), a particular PKRU value associated with a particular access permission to a shared memory area of the DBMS. The signal handler may handle the signal and write, to the PKRU register, a runtime PKRU value. The runtime PKRU value may be associated with a runtime access permission to the shared memory area.
G06F 16/25 - Systèmes d’intégration ou d’interfaçage impliquant les systèmes de gestion de bases de données
G06F 9/455 - ÉmulationInterprétationSimulation de logiciel, p. ex. virtualisation ou émulation des moteurs d’exécution d’applications ou de systèmes d’exploitation
41.
METHODS, SYSTEMS, AND COMPUTER READABLE MEDIA FOR NETWORK ANALYTICS DATA DIRECTOR (NADD)-INFORMED AUTOMATIC CONFIGURATION OF MAXIMUM RESPONSE TIMES
A method for network analytics data director (NADD)-informed configuration of a 3gpp-Sbi-Max-Rsp-Time header value includes receiving, at the NADD and from network functions (NFs), NF configuration details and copies of service-based interface (SBI) messages transmitted to and received by the NFs, determining service operation processing times of the NFs, and communicating, to an NF service consumer, NF analytics data including the service operation processing times and the NF configuration details. The method further includes automatically determining, by the NF service consumer and using the NF analytics data, a 3gpp-Sbi-Max-Rsp-Time header value for an SBI request message, adding, by the NF service consumer, the 3gpp-Sbi-Max-Rsp-Time header value to the SBI request message, and transmitting, by the NF service consumer, the SBI request message to a destination.
Techniques for embodiments generating ETL code for transforming normalized database tables, i.e., snowflake schema, and metadata from an operational database into star schema denormalized dimensions are disclosed. The system accesses metadata associated with the normalized database tables and analyzes the metadata to identify tables and relationships between the tables. Identifying tables includes identifying fact tables and dimension tables referenced by the fact tables. Pattern matching may be used to identify the tables within the normalized dimensions. The tables and the relationships between the tables are parsed to generate an abstract representation, i.e., abstract syntax tree, of the normalized database tables. The system generates an intermediate representation from the abstract representation that defines operations for denormalizing the normalized dimensions. The system renders the operations defined in the intermediate representation into ETL code for creating denormalized dimensions from the normalized dimensions in the operational database.
Techniques for an autonomous edit process for medical claims are disclosed. An electronic claim associated with a patient encounter is retrieved, along with a flag indicative of the claim being erroneous, and an error report identifying an error condition within the claim. A plurality of heterogeneous electronic medical records associated with the patient encounter is retrieved, the plurality including structured billing codes, structured data, semi-structured data, and/or free-text clinical notes. A feature-extraction engine transforms the plurality of heterogeneous electronic medical records into a unified machine-readable representation including semantic embeddings, which are processed by a trained machine learning (ML) model, to generate a mapping between the error condition and one or more spans within the unified representation. The ML model identifies documentary evidence within the one or more spans that satisfies a model-learned evidentiary relevance condition, and generates one or more machine-formatted corrective actions to resolve the error condition.
G16H 10/60 - TIC spécialement adaptées au maniement ou au traitement des données médicales ou de soins de santé relatives aux patients pour des données spécifiques de patients, p. ex. pour des dossiers électroniques de patients
44.
METHOD AND SYSTEM FOR HIGHLY SCALABLE ASYNCHRONOUS MAINTENANCE OF AN INVERTED INDEX
Provided is an improved approach to implement maintenance of inverted indexes, where asynchronous maintenance is performed in a very efficient and highly scalable manner. As a result, it becomes possible to keep search results as close to current as possible, thus allowing the asynchronous maintenance to be fast enough to reach full transactional consistency.
A Productivity Assistant System (PAS) is described that uses specially-trained ML models (e.g., artificial neural networks (ANNs)) to predict a next action to be performed for a sequence of interactions made by a user with one or more applications or services. The predicted action is customized to that user or to a group of users to which the user belongs. Techniques are described for training and using one or more such machine learning models.
The present disclosure describes solutions to a confused identity security vulnerability that can arise during multicloud operations. In an example method, a second cloud environment receives, from a first cloud environment, a request to perform an operation involving a service manager for a cloud service offered in the second cloud environment. The request can include a first identifier and a URL. The second cloud environment outputs a network request to a network location identified by the URL. The second cloud environment receives a response including a second identifier, which it compares to the first identifier. Upon determining, based upon the comparing, that the second identifier matches the first identifier, the second cloud environment performs processing enable performance of the operation involving the service manager. Upon determining that the second identifier does not match the first identifier, the second cloud environment rejects the request to perform the operation.
Techniques disclosed herein pertain to region building for cloud networks and, particularly, for region building process improvements. The techniques include accessing first configuration instructions for building a physical region of a cloud service provider and executing the first configuration instructions. Executing the first configuration instructions causes a first graph that includes nodes to be traversed. A second graph for replacing the first graph can be selected from among candidate graphs. The candidate graphs are generated by reducing an execution time associated with a node of the nodes of the first graph. Second configuration instructions that include instructions for traversing the second graph are generated and executed. Executing the second configuration instructions causes a second graph that includes the nodes to be traversed.
Techniques are disclosed for restricting access to a computing resource in a manner that does not block the performance of other operations in a multi-thread computing environment. A software gate receives a request from a thread for permission to access a computing resource. Responsive to receiving the request, the software gate determines that a dynamic permit limit currently prevents the request from being granted. The software gate returns a data structure indicating that the request is incomplete, adds the request to a queue of pending requests, and releases the thread. Once released, the thread is free to perform other operations while the request is pending. If the request subsequently becomes allowable, the software gate grants the request, removes the request from the queue, and updates the data structure to indicate the request is complete.
G06F 21/44 - Authentification de programme ou de dispositif
49.
METHODS, SYSTEMS, AND COMPUTER READABLE MEDIA FOR FACILITATING PROCESSING OF INTER-PUBLIC LAND MOBILE NETWORK (PLMN) MESSAGES RELATING TO EXISTING SUBSCRIPTIONS
A method for facilitating processing of inter-public land mobile network (PLMN) messages relating to existing subscriptions includes receiving, at a first network function (NF) repository function (NRF), an inter-PLMN subscription creation request message for creating a subscription. The method further includes determining, by the first NRF, that the first NRF includes a resource for which the subscription is requested. The method further includes, in response to determining that the first NRF includes a resource for which the subscription is requested, generating, by the first NRF, a subscription creation response message indicating creation of the subscription. The first NRF includes, in the subscription creation response message, a hint indicating that the subscription is located on the first NRF and forwards the first NRF, the subscription creation response message towards a consumer NF.
H04W 8/18 - Traitement de données utilisateur ou abonné, p. ex. services faisant l'objet d'un abonnement, préférences utilisateur ou profils utilisateurTransfert de données utilisateur ou abonné
H04B 7/08 - Systèmes de diversitéSystèmes à plusieurs antennes, c.-à-d. émission ou réception utilisant plusieurs antennes utilisant plusieurs antennes indépendantes espacées à la station de réception
H04L 12/28 - Réseaux de données à commutation caractérisés par la configuration des liaisons, p. ex. réseaux locaux [LAN Local Area Networks] ou réseaux étendus [WAN Wide Area Networks]
H04L 41/04 - Architectures ou dispositions de gestion de réseau
H04L 67/1021 - Sélection du serveur pour la répartition de charge basée sur la localisation du client ou du serveur
H04W 4/60 - Services basés sur un abonnement qui utilisent des serveurs d’applications ou de supports d’enregistrement, p. ex. boîtes à outils d’application SIM
H04W 8/20 - Transfert de données utilisateur ou abonné
In accordance with an embodiment, described herein are systems and methods for automated identification of churned client entities (e.g., product purchasers or users, cloud service subscribers, or other types of client entities), generally referred to herein as customers; predictive assessment of customer attrition likelihood; and determination of temporal windows for strategic intervention. The system can be used, for example, to predict if a client (e.g., a customer) will churn; determine a churn timeframe or action window for possible action to address the churn; and/or automatically identify which clients or customers may have already churned. Data or information describing churned clients or customers can be used by the system to automatically determine and/or perform an action directed to particular clients or customers; or can be returned in the form of displayed reports or other data visualizations.
G06Q 10/0637 - Gestion ou analyse stratégiques, p. ex. définition d’un objectif ou d’une cible pour une organisationPlanification des actions en fonction des objectifsAnalyse ou évaluation de l’efficacité des objectifs
G06Q 30/01 - Services de relation avec la clientèle
G06Q 40/06 - Gestion de biensPlanification ou analyse financières
Provided is an event queue to implement an improved approach to coordinate the intake of work items that are received and operated upon by entities to maintain an inverted index. An improved approach is provided to implement maintenance of inverted indexes, where asynchronous maintenance is performed in a very efficient and highly scalable manner. Also disclosed is an improved approach to coordinate DML and DDL operations between maintenance processing entities and user processing entities.
Provided is an improved approach to implement maintenance of inverted indexes, where asynchronous maintenance is performed in a very efficient and highly scalable manner. As a result, it becomes possible to keep search results as close to current as possible, thus allowing the asynchronous maintenance to be fast enough to reach full transactional consistency.
Embodiments optimize hotel room pricing by generating a causal model including an estimate of a causal effect of a hotel room price on a demand of the hotel room. Embodiments receive historical hotel room reservation data and select one of a plurality of predictive models based at least on the causal model. Embodiments then map the price of the hotel room to the demand of the hotel room.
G06Q 10/02 - Réservations, p. ex. pour billetterie, services ou manifestations
G06Q 10/04 - Prévision ou optimisation spécialement adaptées à des fins administratives ou de gestion, p. ex. programmation linéaire ou "problème d’optimisation des stocks"
G06Q 30/0283 - Estimation ou détermination de prix
Techniques are described that enable, in a multi-region cloud environment, information regarding one or more tenancy sessions that a network access program (e.g., a browser) participates in to be efficiently stored in a centralized location. The centrally stored sessions information can then be used for various purposes such as for restricting the number of tenancy sessions using a network access program, sessions cleanup, and other sessions-related tasks. In certain implementations, the centrally stored sessions information is used to prevent the network access program from opening multiple sessions for the same tenancy. In such implementations, for a particular tenancy, the network access program is allowed to have only one active session for the particular tenancy at a time. The centrally stored sessions information facilitates efficient sessions management including session cleanup after a session is closed.
A system captures a facial image of a human face, and at approximately a same time as capturing the facial image of the human face, the system captures a three-dimensional spatial representation of the human face. The system determines that the facial image corresponds to the three-dimensional spatial representation, and responsive to determining that the facial image corresponds to the three-dimensional spatial representation, the system generates a certification corresponding to the facial image. The system stores and/or transmits the certification for use in a process for authenticating the facial image.
G06V 40/16 - Visages humains, p. ex. parties du visage, croquis ou expressions
G06F 21/32 - Authentification de l’utilisateur par données biométriques, p. ex. empreintes digitales, balayages de l’iris ou empreintes vocales
G06V 10/74 - Appariement de motifs d’image ou de vidéoMesures de proximité dans les espaces de caractéristiques
H04L 9/32 - Dispositions pour les communications secrètes ou protégéesProtocoles réseaux de sécurité comprenant des moyens pour vérifier l'identité ou l'autorisation d'un utilisateur du système
Systems, methods, and other embodiments associated with modification of existing infrastructure designs by large language models (LLMs) are described. In one embodiment, a method includes accessing an existing graph of compute infrastructure. The existing graph represents a design of the compute infrastructure. The method includes accessing changed infrastructure requirements for the compute infrastructure that differ from the design. The changed infrastructure requirements are in human language. The method includes automatically generating a modified graph from the existing graph and the changed infrastructure requirements using an LLM. The LLM has been trained to generate new graph portions where the compute infrastructure is affected by the changed infrastructure requirements. The method includes converting the modified graph into a deployment specification. And, the method includes executing the deployment specification to automatically configure a target computer system to have modified compute infrastructure that conforms to the changed infrastructure requirements.
Techniques for managing tasks and resources using a map-based GUI are disclosed. A system determines locations and available actions associated with the tasks. The system generates a map representing a geographic region and markers indicating the locations of the tasks within the region. In response to the selection of a marker corresponding to a particular task, the system presents an interactive tooltip corresponding to the available action. Responsive to the selection of the graphic element, the system calls an API that performs the action while concurrently presenting the map.
Techniques may include receiving a request from a tenant to perform an operation with respect to a containerized application. The containerized application can be one of a plurality of containerized applications that are executing on the one or more second computing devices, and where workloads are assigned to the containerized applications by a service provider computing device. In addition, the techniques may include obtaining an identity for the request. The techniques may include providing the identity and the request to a resource manager computing device that is configured to query an access management computing device to determine whether the identity is permitted to perform the operation. The operation may include a change to a parameter of the one or more second computing devices. The techniques may include receiving a response to the request. The response can indicate whether the identity is permitted to perform the operation.
Disclosed is an improved approach to implement a learning-based recommendation system which provides a recommendation for the operator access, e.g., for the proper duration and privilege required when a new operator access request is raised for accessing the customer resource.
G06F 21/57 - Certification ou préservation de plates-formes informatiques fiables, p. ex. démarrages ou arrêts sécurisés, suivis de version, contrôles de logiciel système, mises à jour sécurisées ou évaluation de vulnérabilité
Embodiments optimize hotel room overbooking limits for reservations of hotel rooms of a hotel. Embodiments receive historical reservation data and determine an upgrade offer acceptance probability as a function offer price based on the historical reservation data. Embodiments determine a premium category occupancy distribution based on the historical reservation data and determine a basic category cancellation distribution based on the historical reservation data. Embodiments determine an optimal upgrade price as a function of overbooked rooms from the upgrade offer acceptance probability and determine a marginal revenue as a function of overbooked rooms based on the determined premium category occupancy distribution and the determined optimal upgrade price as a function of overbooked rooms. Embodiments determine a marginal loss as a function of overbooked rooms from the basic category cancellation distribution.
The present disclosure generally relates to systems and methods for intelligently predicting which tests to run on modified source code of an enterprise application. More specifically, the present disclosure relates to systems and methods that build a model using machine-learning algorithms to predict which tests to run on modified source code. The prediction may be based on the modification made to the application code of the enterprise application.
A system captures a facial image of a human face, and at approximately a same time as capturing the facial image of the human face, the system captures a three-dimensional spatial representation of the human face. The system determines that the facial image corresponds to the three-dimensional spatial representation, and responsive to determining that the facial image corresponds to the three-dimensional spatial representation, the system generates a certification corresponding to the facial image. The system stores and/or transmits the certification for use in a process for authenticating the facial image.
A defragmentation process is provided for contiguous persistent or volatile memory regions. The defragmentation process selects and moves extents, updates extent maps, and ensures all read/write operations are consistent and uninterrupted. The defragmentation process can be applied to online maintenance defragmentation, online on-demand defragmentation, or offline defragmentation. The illustrative embodiments provide a source-destination mapping algorithm that allows for optimal defragmentation outcome with least amount of space relocation. In some embodiments, a cost-based greedy algorithm is used for source-destination mapping. Quiesce and unquiesce mechanisms allow for fine-grained access control for the extent currently being relocated by defragmentation.
Techniques are described herein for performing generation of vector-based recommendations for entities. One or more embodiments address the challenge of generating recommendations for entities lacking sufficient recommendation generation input data. Unlike traditional collaborative filtering methods, various embodiments leverage content filtering techniques to make tailored and relevant recommendations for entities. By analyzing entity attributes and generating entity attribute vectors in an N-dimensional space, this system identifies the most suitable transactions by determining the closest neighbors in this space, thereby enhancing the accuracy and effectiveness of recommendations for entities associated with insufficient recommendation generation input data within the platform.
A defragmentation process is provided for contiguous persistent or volatile memory regions. The defragmentation process selects and moves extents, updates extent maps, and ensures all read/write operations are consistent and uninterrupted. The defragmentation process can be applied to online maintenance defragmentation, online on-demand defragmentation, or offline defragmentation. The illustrative embodiments provide a source-destination mapping algorithm that allows for optimal defragmentation outcome with least amount of space relocation. In some embodiments, a cost-based greedy algorithm is used for source-destination mapping. Quiesce and unquiesce mechanisms allow for fine-grained access control for the extent currently being relocated by defragmentation.
G06F 16/215 - Amélioration de la qualité des donnéesNettoyage des données, p. ex. déduplication, suppression des entrées non valides ou correction des erreurs typographiques
G06F 16/27 - Réplication, distribution ou synchronisation de données entre bases de données ou dans un système de bases de données distribuéesArchitectures de systèmes de bases de données distribuées à cet effet
66.
PARTITIONING OF INVERTED FILE (IVF) VECTOR INDEXES IN A DATABASE SYSTEM
Techniques for partitioning an inverted file (IVF) vector index in a database system are provided. In one technique, a partitioned IVF index is generated based on vectors in a base table using the following process. First, a first clustering technique is performed on vectors in a first subset of the base table. Such clustering results in a first plurality of clusters of vectors. For each cluster, a centroid of that cluster is identified, the centroid is stored in a first partition of a centroids table of the partitioned IVF index, the vectors in that cluster are stored in a first subpartition of a clusters table of the partitioned IVF index (where the first subpartition corresponds to the first partition of the centroids table), and a first association is stored between the centroid and the first subpartition. This clustering repeats for each subset of the base table.
Techniques for data synchronization using transaction identifications within objects are disclosed. In some embodiments, a method comprises the following: executing a first data synchronization process for synchronizing data objects comprising corresponding transaction identifications (IDs) from a source data repository to a destination data repository, wherein an interruption occurs in the first data synchronization process; identifying a first transaction ID for the first data synchronization process that was last processed prior to the interruption; identifying a second transaction ID that is subsequent to the first transaction ID in a sequence of transaction IDs; identifying a second set of one or more data objects that each comprise the second transaction ID; and executing a second data synchronization process for synchronizing the second set of one or more data objects by copying the second set of one or more data objects from the source data repository to the destination data repository.
G06F 16/27 - Réplication, distribution ou synchronisation de données entre bases de données ou dans un système de bases de données distribuéesArchitectures de systèmes de bases de données distribuées à cet effet
Techniques are provided for cross-region replication of block volume data. The techniques include a method wherein a computer system implements operations including creating a first snapshot of a block volume at a first geographic region and at a first logical time, the block volume including a plurality of partitions, transmitting first snapshot data to an object storage system at a second geographic region, creating a second snapshot of the block volume at the first geographic region and at a second logical time, generating a plurality of deltas, transmitting a plurality of delta data sets corresponding to the plurality of deltas to the object storage system at the second geographic region, generating a checkpoint at least in part by aggregating object metadata associated with the plurality of deltas and the first snapshot, receiving a restore request to generate a restore volume, and generating the restore volume from the checkpoint.
G06F 16/27 - Réplication, distribution ou synchronisation de données entre bases de données ou dans un système de bases de données distribuéesArchitectures de systèmes de bases de données distribuées à cet effet
A framework for establishing ephemeral privileged access workstations for providing access to production workstations is described. For example, an ephemeral privileged access workstation may be provisioned just in time at a request from a client device for accessing a production workstation.
A memorability prediction system (MPS) is described for predicting the image memorability of an input image while considering the contribution of sub-images and pixels of the input image. In some embodiments, a visual transformer-based memorability prediction network in the MPS may include three machine learning (ML) models responsible for processing different parts of the input image, namely, the whole input image (referred to as the main image), partitioned images (referred to as diced images or sub-images) of the input image, and pixels of the input image. In further embodiments, a relationship between the main image and one or more sub-images may be identified and passed between two ML models. In some embodiments, the three ML models may generate intermediate information to be combined to result in a final memorability score of the input image.
G06V 10/82 - Dispositions pour la reconnaissance ou la compréhension d’images ou de vidéos utilisant la reconnaissance de formes ou l’apprentissage automatique utilisant les réseaux neuronaux
G06V 10/26 - Segmentation de formes dans le champ d’imageDécoupage ou fusion d’éléments d’image visant à établir la région de motif, p. ex. techniques de regroupementDétection d’occlusion
G06V 10/774 - Génération d'ensembles de motifs de formationTraitement des caractéristiques d’images ou de vidéos dans les espaces de caractéristiquesDispositions pour la reconnaissance ou la compréhension d’images ou de vidéos utilisant la reconnaissance de formes ou l’apprentissage automatique utilisant l’intégration et la réduction de données, p. ex. analyse en composantes principales [PCA] ou analyse en composantes indépendantes [ ICA] ou cartes auto-organisatrices [SOM]Séparation aveugle de source méthodes de Bootstrap, p. ex. "bagging” ou “boosting”
71.
TECHNIQUES FOR PARTITIONING IMAGES FOR MACHINE LEARNING MODELS
A memorability prediction system (MPS) is described for predicting image memorability of an input image while considering the contribution of sub-images and pixels of the input image. In some embodiments, the input image may be partitioned (also referred to as diced) randomly into multiple sub-images. Various techniques for dicing the input images are described. In certain embodiments, the input image may be partitioned randomly into one or more segments in both the x-dimension (e.g., width) and the y-dimension (e.g., height). The segments in both dimensions may be combined to generate sub-images in different sizes. In some embodiments, objects in the input image may be identified, and the sub-images may be re-arranged or re-oriented according to an arrangement configuration to achieve higher probability of partitioning or preserving one or more objects. In some embodiments, a pre-defined partition may be used for one or more regions of the input image.
Techniques for validating use of data in training of machine learning (ML) models are disclosed. Synthetic data is by generated, by sampling from a statistical distribution of user data. The user data and the synthetic data are fed to an inference endpoint of the ML model. First results are generated by the ML model, based on the user data; and second results are generated by the ML model, based on the synthetic data. A statistical analysis is conducted, based at least in part on the first results and the second results. A determination is made as to whether the user data was used for training the ML model, based at least in part on conducting the statistical analysis. An indication of the determination as to whether the user data was used for training the ML model is displayed on a user interface.
Techniques for partitioning an inverted file (IVF) vector index in a database system are provided. In one technique, a partitioned IVF index is generated based on vectors in a base table using the following process. First, a first clustering technique is performed on vectors in a first subset of the base table. Such clustering results in a first plurality of clusters of vectors. For each cluster, a centroid of that cluster is identified, the centroid is stored in a first partition of a centroids table of the partitioned IVF index, the vectors in that cluster are stored in a first subpartition of a clusters table of the partitioned IVF index (where the first subpartition corresponds to the first partition of the centroids table), and a first association is stored between the centroid and the first subpartition. This clustering repeats for each subset of the base table.
The present disclosure relates to cloud-based centralized clinical data exchange (CDeX) techniques leveraging a unified interoperability interface for sharing selectively and/or dynamically medical records of subjects with other participants. In some aspects, techniques may be provided to facilitate, support or perform notifying participants or external entities (e.g., regulators, payers, insurance companies or other entities) in real-time or near real-time when a subject encounter occurs and/or throughout the encounter by transmitting admission, discharge, and transfer (ADT) messages. The unified interoperability interface may enable data sharing by establishing subject-specific and/or participant-specific communication channels that can be initiated by either party, provided that the participants are on-boarded and registered within the CDeX system.
G16H 40/67 - TIC spécialement adaptées à la gestion ou à l’administration de ressources ou d’établissements de santéTIC spécialement adaptées à la gestion ou au fonctionnement d’équipement ou de dispositifs médicaux pour le fonctionnement d’équipement ou de dispositifs médicaux pour le fonctionnement à distance
G16H 10/60 - TIC spécialement adaptées au maniement ou au traitement des données médicales ou de soins de santé relatives aux patients pour des données spécifiques de patients, p. ex. pour des dossiers électroniques de patients
Systems, methods, and machine-readable media may facilitate programmable data trimming. Metadata associated with a collective operation may be determined by an application of a server. The metadata may specify a job identifier corresponding to a unit of work to be completed in conjunction with the collective operation, a collective type of the collective operation, and/or an ordering mode for packets corresponding to the collective operation. The metadata associated with the collective operation may be sent by the application to a network interface card (NIC). The NIC may be caused by the application to transmit a data packet with the metadata embedded in a cookie of the data packet to a switch of a network fabric to cause the switch to use a selected network path and/or selected load-balancing for the collective operation based on one or more of the job identifier, the collective type, and/or the ordering mode.
Techniques for receiving, storing, analyzing, and utilizing raw content are disclosed. The system receives raw text from a user, including a first tag and second tags. The system identifies and removes formatting attributes in the raw text, including whitespace characters, to generate a normalized input. The system stores the normalized input in target dataset(s) and parses the normalized input for the first tag and the second tags. In this case, the first tag corresponds to one or more topics, and the second tags represent a computing system associated with corresponding portions of the normalized input. The system analyses the first tag and the second tags to identify the topics. The system generates one or more topic maps for the target dataset(s) based on the first tag and the one or more second tags. The topic map(s) include one or more references to content items within the target dataset(s).
Techniques disclosed herein relate generally to text classification and include techniques for fusing word embeddings with word scores for text classification. In one particular aspect, a method for text classification is provided that includes obtaining an embedding vector for a textual unit, based on a plurality of word embedding vectors and a plurality of word scores. The plurality of word embedding vectors includes a corresponding word embedding vector for each of a plurality of words of the textual unit, and the plurality of word scores includes a corresponding word score for each of the plurality of words of the textual unit. The method also includes passing the embedding vector for the textual unit through at least one feed-forward layer to obtain a final layer output, and performing a classification on the final layer output.
H04L 51/02 - Messagerie d'utilisateur à utilisateur dans des réseaux à commutation de paquets, transmise selon des protocoles de stockage et de retransmission ou en temps réel, p. ex. courriel en utilisant des réactions automatiques ou la délégation par l’utilisateur, p. ex. des réponses automatiques ou des messages générés par un agent conversationnel
In one embodiment, a method includes receiving from a service executing in a service tenancy and by an event broker, a request to modify a rule to deliver a set of events from a first tenancy to the service tenancy. Modify may include at least one of create or update. The method also includes receiving from the service and by the event broker, a proxy token for substantiating the request. The proxy token represents an authority of a user principal of the first tenancy. The method further includes determining, by the event broker, whether modification of the rule is authorized based at least on the authority of the user principal, and subsequent to determining that the modification of the rule is authorized, delivering, by the event broker, the set of events from the first tenancy to the service tenancy according to the rule.
Techniques are disclosed for supporting heterogenous arrays. A method comprises determining, from metadata generated from sample data, a first discriminator and a second discriminator, wherein the first discriminator identifies an occurrence of a heterogeneous array included within received data that follows an open-standard data interchange format, and the second discriminator identifies one or more resource types included within the heterogenous array; receiving data that follows the open-standard data interchange format; determining, based on an occurrence of the first discriminator within the data, a heterogeneous array; determining, based on an occurrence of the second discriminator identified within the heterogeneous array, a resource type identified within the data; determining one or more attributes associated with the resource type from the data; generating a normalized resource from the resource type and the one or more attributes that conforms with an integration model; and performing one or more actions using the normalized resource.
Techniques are disclosed for simplifying extensions. A method comprises receiving data that comprises a standard resource that conforms to a Fast Healthcare Interoperability Resources (FHIR) standard, wherein the standard resource includes one or more standard data elements for representing and exchanging healthcare information; determining that the standard resource further includes an extension that incorporates a non-standard data element into the standard resource, wherein the non-standard data element is defined by a structure definition that is external from the data; converting the standard resource into a normalized resource that conforms with an integration model, wherein the converting includes normalizing the non-standard data element based on a schema of the integration model to generate a normalized extension element that includes an attribute that extends the standard resource, and wherein the normalized resource comprises the one or more standard data elements and the normalized extension element; and providing the normalized resource.
G16H 40/20 - TIC spécialement adaptées à la gestion ou à l’administration de ressources ou d’établissements de santéTIC spécialement adaptées à la gestion ou au fonctionnement d’équipement ou de dispositifs médicaux pour la gestion ou l’administration de ressources ou d’établissements de soins de santé, p. ex. pour la gestion du personnel hospitalier ou de salles d’opération
G16H 70/20 - TIC spécialement adaptées au maniement ou au traitement de références médicales concernant des pratiques ou des directives
H04L 67/12 - Protocoles spécialement adaptés aux environnements propriétaires ou de mise en réseau pour un usage spécial, p. ex. les réseaux médicaux, les réseaux de capteurs, les réseaux dans les véhicules ou les réseaux de mesure à distance
81.
Augmenting Knowledge Not Available in Electronic Health Record System Patient Charts
A system processes clinical guidance to enhance patient care management. The system receives clinical guidance in digital text form from authoritative medical sources. A large language model analyzes the received clinical guidance to extract key information and relationships. The system generates a structured pathway based on the analyzed clinical guidance. The generated pathway represents a comprehensive summary for a specific disease state, organized into logical steps. These steps encompass treatment goals, management strategies, and measures for preventing complications. The system integrates the generated pathway information into an electronic health record (EHR) system. Within the EHR, the system produces a patient chart incorporating the derived pathway information.
G16H 10/60 - TIC spécialement adaptées au maniement ou au traitement des données médicales ou de soins de santé relatives aux patients pour des données spécifiques de patients, p. ex. pour des dossiers électroniques de patients
G16H 70/20 - TIC spécialement adaptées au maniement ou au traitement de références médicales concernant des pratiques ou des directives
82.
Creating Semantic Knowledge Relationships for Electronic Health Record Systems Content
A system and method for enhancing electronic health record (EHR) systems through integration of proprietary and standardized medical terminologies. Embodiments proprietary terminology concepts from electronic sources and identifies matching standardized medical terminology concepts. A knowledge graph is generated, incorporating both proprietary and standardized concepts along with their relationships. The system receives patient data, including problem lists, medication lists, and lab results. This data is analyzed against the knowledge graph to generate suggestions for the EHR system. These suggestions are then provided to the EHR system for display, enhancing clinical decision support and improving patient care.
G06N 5/025 - Extraction de règles à partir de données
G16H 10/60 - TIC spécialement adaptées au maniement ou au traitement des données médicales ou de soins de santé relatives aux patients pour des données spécifiques de patients, p. ex. pour des dossiers électroniques de patients
G16H 20/10 - TIC spécialement adaptées aux thérapies ou aux plans d’amélioration de la santé, p. ex. pour manier les prescriptions, orienter la thérapie ou surveiller l’observance par les patients concernant des médicaments ou des médications, p. ex. pour s’assurer de l’administration correcte aux patients
G16H 70/40 - TIC spécialement adaptées au maniement ou au traitement de références médicales concernant des médicaments, p. ex. leurs effets secondaires ou leur usage prévu
83.
PROXY-BASED TECHNIQUES FOR AUTHORIZING CROSS-REALM REQUESTS
Techniques are disclosed for using a proxy service to generate resource principals corresponding to a cross-realm request. A request to perform an operation in a target realm (TR) may be received by the proxy service of a host realm (HR). The request may comprise identity data that indicates an identifier of the requestor in one or more identity realms (e.g., in at least the TR). The proxy service of the HR may establish a trusted connection with a proxy service of the TR. The proxy service of the HR may transmit request data that indicates the identity of the requestor within the TR, causing the proxy service in the TR to generate a resource principal object corresponding to the identity of the requestor in the TR, whereby the resource principal object is used to execute (or to attempt execution of) the requested operation in the TR.
Techniques are disclosed for restricting access to a computing resource in a manner that does not block the performance of other operations in a multi-thread computing environment. A software gate receives a request from a thread for permission to access a computing resource. Responsive to receiving the request, the software gate determines that a dynamic permit limit currently prevents the request from being granted. The software gate returns a data structure indicating that the request is incomplete, adds the request to a queue of pending requests, and releases the thread. Once released, the thread is free to perform other operations while the request is pending. If the request subsequently becomes allowable, the software gate grants the request, removes the request from the queue, and updates the data structure to indicate the request is complete.
Techniques are disclosed for restricting access to a computing resource in a manner that does not block the performance of other operations in a multi-thread computing environment. A software gate receives a request from a thread for permission to access a computing resource. Responsive to receiving the request, the software gate determines that a dynamic permit limit currently prevents the request from being granted. The software gate returns a data structure indicating that the request is incomplete, adds the request to a queue of pending requests, and releases the thread. Once released, the thread is free to perform other operations while the request is pending. If the request subsequently becomes allowable, the software gate grants the request, removes the request from the queue, and updates the data structure to indicate the request is complete.
G06F 21/44 - Authentification de programme ou de dispositif
86.
METHODS, SYSTEMS, AND COMPUTER READABLE MEDIA FOR NETWORK ANALYTICS DATA DIRECTOR (NADD)-ASSISTED DYNAMIC CONFIGURATION OF HYPERTEXT TRANSFER PROTOCOL (HTTP) PARAMETER SETTINGS AT NETWORK FUNCTIONS (NFs)
A method for NADD-assisted dynamic configuration of HTTP parameter settings at NFs includes receiving, at the NADD, SBI message feeds from a plurality of producer NFs. The method further includes determining, by the NADD and from at least one of the SBI message feeds, an HTTP parameter setting for one of the producer NFs. The method further includes communicating, by the NADD, the HTTP parameter setting to the producer NF. The method further includes receiving, by the producer NF and from the NADD, the HTTP parameter setting. The method further includes using, by the producer NF, the HTTP parameter setting to control traffic flow from a consumer NF to the producer NF.
Techniques for a container orchestration system are disclosed. A system executes a virtual agent in a cloud network on a virtual node of a container orchestration system. The virtual node hosts multiple container instances within the cloud environment. The system executes a first container instance within the virtual node and connects the first container instance to a first subnet. The system executes a second container instance within the same virtual node and connects the second container instance to a second subnet distinct from the first subnet. The system enables access to the first container instance through the first subnet and enables access to the second container instance via the second subnet. This architecture allows for flexible network configurations within a single virtual node, enhancing resource utilization and network segmentation capabilities in containerized environments.
G06F 9/455 - ÉmulationInterprétationSimulation de logiciel, p. ex. virtualisation ou émulation des moteurs d’exécution d’applications ou de systèmes d’exploitation
88.
METHODS, SYSTEMS, AND COMPUTER READABLE MEDIA FOR NETWORK ANALYTICS DATA DIRECTOR (NADD)-ASSISTED DYNAMIC CONFIGURATION OF HYPERTEXT TRANSFER PROTOCOL (HTTP) PARAMETER SETTINGS AT NETWORK FUNCTIONS (NFS)
A method for NADD-assisted dynamic configuration of HTTP parameter settings at NFs includes receiving, at the NADD, SBI message feeds from a plurality of producer NFs. The method further includes determining, by the NADD and from at least one of the SBI message feeds, an HTTP parameter setting for one of the producer NFs. The method further includes communicating, by the NADD, the HTTP parameter setting to the producer NF. The method further includes receiving, by the producer NF and from the NADD, the HTTP parameter setting. The method further includes using, by the producer NF, the HTTP parameter setting to control traffic flow from a consumer NF to the producer NF.
Embodiments are directed to operating a cloud based product configurator. Embodiments store, as vectorized data in a vector database product, information corresponding to a first product to be configured. While configuring the first product, embodiments receive a query regarding the first product. Embodiments augment the query in response to a context based semantic search of the vector database using the query. Embodiments prompt a large language model (“LLM”) using the augmented query and receiving an LLM response. Embodiments the provide the LLM response in response to the query.
A method generates static defect checkers using a language model. The method includes generating an example representation. The method further includes combining an explanation section, an instruction section, and a description section to generate a prompt. The explanation section includes the example representation, the instruction section includes instruction text, and the description section includes defect description text. The defect description text includes a natural language description of a defect corresponding to the example representation. The explanation section includes operations corresponding to the defect. The instruction text includes instructions in the natural language to generate defect checker code using the example representation and the defect description text. The method further includes executing a language model using the prompt to generate the defect checker code. The defect checker code is in a programming language.
Systems, methods, and computer-readable media are provided for generating an interactive node graph showing process states. A data management system accesses a data structure that includes a set of candidate process records representing a plurality of candidate process states, where each candidate process record represents a candidate process state and one or more sequentially connected candidate process states. The data management system generates an interactive node graph that includes nodes, each node representing a candidate process state. The interactive node graph includes edges representing candidate connections between candidate process states. The data management system causes display of the interactive node graph. The data management system receives input modifying settings for displaying the interactive node graph and causes display of an updated interactive node graph using the settings as modified.
Systems and methods and computer-readable media are provided for generating an interactive node graph showing aggregate node and/or edge metrics. The interactive node graph includes nodes and edges, each representing a process state or a connection between process states. The data management system marks each node graphically based on a first metric type that is based on an aggregation of occurrences of a corresponding process state and marks each edge graphically based on a second metric type that is based on an aggregation of occurrences of a transition between corresponding process states. The data management system causes display of the interactive node graph, receives input modifying a particular metric type of the first metric type or the second metric type, and causes display of an updated interactive node graph based at least in part on the particular metric type as modified.
Systems, methods, and computer-readable media are provided for animating an interactive node graph. A data management system generates an interactive node graph having nodes that represent process states and edges representing connections between process states. The data management system uses a first live metric for determining aggregated node values and a second live metric to use for determining aggregated edge values. The data management system causes display of the interactive node graph according to a selected data slice of a plurality of data slices. Based at least in part on a selection of an option to play the interactive node graph through the plurality of data slices, the data management system updates the display of the interactive node graph to show a different data slices of the plurality of data slices, and, after an amount of time, another different data slice of the plurality of data slices.
Techniques for enabling a service to perform operations corresponding to a subject tenancy on behalf of a governing tenancy are disclosed. The system receives a request from a service for a resource principal token. The request includes a resource principal, a service identifier for the service, and a link identifier that identifies a governance link. The governance link is associated with a governing tenancy, a subject tenancy, and a service. The system evaluates the governance link to determine if the governance link is active. After determining that the governance link is active, the system responds to the request from the service, providing a resource principal token. The resource principal token that is provided to the service forms a basis for authorizing the service to perform operations corresponding to the subject tenancy on behalf of the governing tenancy.
G06F 9/455 - ÉmulationInterprétationSimulation de logiciel, p. ex. virtualisation ou émulation des moteurs d’exécution d’applications ou de systèmes d’exploitation
Systems, methods, and other embodiments associated with LLM-based generation of computing infrastructure are described. In one embodiment, an example method includes accessing infrastructure requirements for compute infrastructure that are in human language. The example method includes translating the infrastructure requirements into a physical infrastructure topology using one or more large language models. The example method includes converting the physical infrastructure topology into an executable deployment specification. And, the example method includes executing the deployment specification to automatically configure a target computer system to have the compute infrastructure described by the infrastructure requirements.
G06F 30/13 - Conception architecturale, p. ex. conception architecturale assistée par ordinateur [CAAO] relative à la conception de bâtiments, de ponts, de paysages, d’usines ou de routes
96.
DETECT REDUNDANT INITIALIZATION CHECKS USING STATIC ANALYSIS
A method detects redundant initialization checks using static analysis. The method includes inlining source code to generate inlined code. The inlined code includes instructions to materialize an element within a scope. The method further includes consolidating the inlined code to form consolidated code by moving the instructions to materialize the element to a point where the element escapes the scope. The method further includes running a points-to analysis on the consolidated code. The method further includes reducing the consolidated code to generate reduced code by removing an initialization check from the consolidated code.
Improved network traffic flow processing techniques are described. In a network device providing multiple processing planes, different processing resources can be allocated to affect efficient and rapid packet processing. This allocation of resources can be upset via receipt of a configuration update. When a configuration update is received, a previously programmed flow can be provisionally invalidated. To prevent the overwhelming of slow path resources, a provisionally invalid flow can continue to be processed according to previous programming by a fast path.
H04L 45/00 - Routage ou recherche de routes de paquets dans les réseaux de commutation de données
98.
METHODS, SYSTEMS, AND COMPUTER READABLE MEDIA FOR PROVIDING ACCESS TO COMMUNICATION NETWORK HEALTH INFORMATION USING COMMUNICATION-NETWORK-AWARE GENERATIVE ARTIFICIAL INTELLIGENCE (AI) RETRIEVAL AUGMENTED GENERATION (RAG) MODEL AND NETWORK FUNCTION (NF)
A method for providing access to communication network health information using a communication-network-aware generative AI RAG model includes receiving, as a first input to the RAG model, a query for communication network health information and receiving, as a second input to the RAG model, at least one feed of communication network health information regarding at least one NF. The method further includes using the query to extract, from the communication network health information regarding the at least one network function, context information for the query for communication network health information, providing the query and the context information as inputs to a base LLM component of the RAG model, and generating, as output, a query response including an indication of the communication network health information requested by the query and in a natural language format.
H04L 41/16 - Dispositions pour la maintenance, l’administration ou la gestion des réseaux de commutation de données, p. ex. des réseaux de commutation de paquets en utilisant l'apprentissage automatique ou l'intelligence artificielle
Systems, methods, and computer-readable media are provided for storing a snapshot of an interactive node graph. A data management system generates an interactive node graph having nodes that represent process states and edges representing connections between process states.
Systems, methods, and computer-readable media are provided for storing a snapshot of an interactive node graph. A data management system generates an interactive node graph having nodes that represent process states and edges representing connections between process states.
The data management system uses a first live metric for determining aggregated node values and a second live metric to use for determining aggregated edge values. Based at least in part on a selection of an option to save the interactive node graph, the data management system stores a snapshot. The snapshot includes the aggregated node values based on the first live metric as of a particular time, the aggregated edge values based on the second live metric as of the particular time, a first mapping between the first live metric and the nodes, and a second mapping between the second live metric and the edges. The stored snapshot is loadable without access to the first live metric and the second live metric to display the interactive node graph.
G06F 3/0484 - Techniques d’interaction fondées sur les interfaces utilisateur graphiques [GUI] pour la commande de fonctions ou d’opérations spécifiques, p. ex. sélection ou transformation d’un objet, d’une image ou d’un élément de texte affiché, détermination d’une valeur de paramètre ou sélection d’une plage de valeurs
Various embodiments of the present technology generally relate to systems and methods for optimizing resources within a customer data platform (CDP). In certain embodiments, a method may comprise operating a resource optimizer system to implement a CDP resource optimization process to remove obsolete metadata, the CDP resource optimization process including monitoring metadata usage within a CDP, generating metrics for a metadata element based on the metadata usage, defining a rule set for selecting the obsolete metadata for removal based on the metrics, and applying the rule set to remove the obsolete metadata.
G06F 16/11 - Administration des systèmes de fichiers, p. ex. détails de l’archivage ou d’instantanés
G06F 11/14 - Détection ou correction d'erreur dans les données par redondance dans les opérations, p. ex. en utilisant différentes séquences d'opérations aboutissant au même résultat
