Patents.

Data retention handling for data object stores

Methods, systems, and devices for data retention handling are described. In some data storage systems, data objects are stored in a non-relational database schema. The system may support configurable data retention policies for different tenants, users, or applications. For example, a data store may receive retention requests, where the retention requests may specify deletion or exportation actions to perform on records contained within data objects. The data store may determine retention rules based on these retention requests, and may periodically or aperiodically evaluate the rules to determine active actions to perform. To improve the efficiency of the system, the data store may aggregate the active actions (e.g., according to the dataset to perform the actions on), and may generate work items corresponding to the aggregate actions. A work processor may retrieve these work items and may efficiently perform the data retention actions on datasets stored in the data object store.

Cognitive classification-based technical support system

Disclosed embodiments provide techniques for technical support. Previously reported problem reports are analyzed and symptoms are extracted. Solutions are associated with the previously reported problem reports. A newly submitted user problem is analyzed and symptoms are extracted and compared with the symptoms of the previously reported problems. Solutions are then associated with the user problem based on relevance to symptoms, product type, and/or other factors.

Method and system for implementing circle of trust in a social network

Automatic turn delineation in multi-turn dialogue

Securely computing skill level using data anonymization

Techniques for securely computing skill level of users of an online service using data anonymization to protect data privacy are disclosed herein. In some embodiments, a computer-implemented method comprises performing a data anonymization process on a dataset.

Service-oriented data architecture for a vehicle

System, methods, and other embodiments described herein relate to a service-oriented data architecture within a vehicle. In one embodiment, a computing system for controlling electronic systems of a vehicle includes a system processing unit that executes multiple virtual machines (VMs) to isolate different services of the vehicle. The computing system includes a communication plane spanning between the multiple VMs to provide communications across the multiple VMs and with a mechatronics layer and a sensor layer of the vehicle. The multiple VMs provide the different services by executing microservices that are formed to be self-contained and standardized independent of programmed functions and to interoperate with the communication plane and the multiple VMs.

Cryptoasset custodial system with different cryptographic keys controlling access to separate groups of private keys

Methods, systems, and apparatus, including medium-encoded computer program products, for secure storage and retrieval of information, such as private keys, useable to control access to a blockchain, include, in at least one aspect, a method including: identifying for an action an associated...

Interpretable counting in visual question answering

Managing consistency models in a distributed database

Embodiments are provided for enabling a dynamic management of a multi-tenant distributed database.

Systems and methods for detecting photoplethysmographic device usage

Systems and methods for determining if a wearable photoplethysmography device is correctly positioned by using a classifier to determine if a signal is valid or invalid, with linear computational complexity and minimal memory complexity.

Term-cluster knowledge graph for support domains

Given a skeleton knowledge graph including first terms in respective nodes, wherein first terms in connected nodes have a predetermined relationship, an augmented knowledge graph is formed by a first computing device by extracting a second term from a domain corpus to form a term cluster linked with a respective node of the knowledge graph. The second term is associated with the first term of a respective node in the domain corpus while not meeting the predetermined relationship. A semantic feature between the second term and the associated first term is identified in the domain corpus and linked to the pair of the second term and the first term in the augmented knowledge graph. The augmented knowledge graph is useable by a second computing device, which may or may not be the same as the first computing device, to drive a conversation between a chatbot and user.

Machine learning systems architectures for ranking

Preventing statistical inference attacks on data that is shared among multiple stakeholders with disjoint access privileges

According to one embodiment, a method, computer system, and computer program product for preventing statistical inference attacks is provided. The present invention may include splitting records into items, and classifying these items into shared items and private items; grouping the private items according to privacy and confidentiality requirements; restricting access of the private items to stakeholders based on the confidentiality requirements using cryptographic keys; generating and encrypting one or more placeholders for both existent and non-existent stakeholders; storing private items in private storage as indicated by links; creating shared records comprising links, placeholders, and shared items; adding integrity signatures to the shared records; and publishing the shared records to a shared medium.

Rewarding affiliates

Technology for rewarding affiliates is provided. In one example, a method may include determining a physical location of a computing device. A purchase intent from the computing device while at the physical location may be identified. An affiliate associated with the physical location may be rewarded for the purchase intent.

Image augmentation and object detection

Computing systems may support image classification and image detection services, and these services may utilize object detection/image classification machine learning models. The described techniques provide for normalization of confidence scores corresponding to manipulated target images and for non-max suppression within the range of confidence scores for manipulated images. In one example, the techniques provide for generating different scales of a test image, and the system performs normalization of confidence scores corresponding to each scaled image and non-max suppression per scaled image These techniques may be used to provide more accurate image detection (e.g., object detection and/or image classification) and may be used with models that are not trained on modified image sets. The model may be trained on a standard (e.g. non-manipulated) image set but used with manipulated target images and the described techniques to provide accurate object detection.

Language profiling service

Methods, systems, and computer programs are presented for estimating the language used in a user communication. One method includes an operation for utilizing counters to track use of languages by a user of an online service, the counters being updated based on interactions of the user in the online service.

Two-stage online detection of action start in untrimmed videos

Facilitating dynamically unified system of record in an on-demand services environment

Distributed processing in a messaging platform

Dialog-style application programming interface for a machine learning solution

An approach is provided for using a dialog-style API for a machine learning solution. A server receives from a client a first call to a dialog-style application programming interface (API) as a first request having an incomplete set of data items. The API determines that one or more data items required by the first request are missing... The server sends to the client a reply having values of data items in a response to the second request and second metadata that describes the data items in the response.

Distributed processing in a messaging platform

Transposable behavior data

Aspects of the present invention disclose a method, computer program product, and system for capturing and transposing user behavioral data. The method includes determining that a user enters a first venue based on tracking a computing device of the user. The method further includes capturing behavioral data of the user in the first venue as the user interacts with a first product, where the captured behavioral data is saved to a database. Responsive to determining that the user enters a second venue, a set of recommendations is determined for the user based on the behavioral data in the user profile and a product available to the user in the second venue.

Video analysis system for optimizing an activity protocol

A method implements an activity protocol that has been established for a living entity. The method directs an artificial intelligence system to compare a video of an activity by a living entity to an activity protocol that has been established for the living entity, and then receives an indication from the artificial intelligence system that the activity by the living entity does not match the activity protocol that has been established for the living entity. In response to receiving the indication from the artificial intelligence system that the activity by the living entity does not match the activity protocol that has been established for the living entity, the method directs a monitoring entity to implement the activity protocol that has been established for the living entity.

Real-time sensor based balance gamification and feedback

A real-time performance management may be used for cycling. A first computing device retrieves profile data related to a cyclist. The first computing device receives sensor readings from one or more sensors. A first sensor of the one or more sensors detects cyclist tilt. Profile data and the sensor readings from the one or more sensors determine cyclist status. A notification based on the determined cyclist status communicates to a user.

User action collection for sensitive customer data

A secure tenant activity data system for a deployment service includes a processor configured to decrypt a set of encrypted stream processed monitoring data, determine and match a transaction identifier, aggregate the stream processed monitoring data with previously decrypted data, re-encrypt the aggregate and provide the encrypted aggregated data.

Methods and systems for determination of a routing policy for an autonomous vehicle

Methods and systems for determining an optimal routing policy for an autonomous vehicle. A set of states representing portions of lanes is determined, with a set of costs for actions (stay in lane, lane change, forced lane change) associated with each state; a lane change action has a non-deterministic outcome with a probability based on the length of the state, and an optimal routing policy is computed over the resulting model.

Ambiguity resolution with dialogue search history

Expandable dialogue system

Differentially private budget tracking using Renyi divergence

A differentially private security system communicatively coupled to a database storing restricted data receives a database query from a client. The database query includes a relation indicative of data to perform the query upon and at least one privacy parameter indicative of a level of differential privacy.

Automatic spoken dialogue script discovery

Hazardous chemical detection in a microwave oven

In an apparatus for hazardous chemical detection in a microwave oven, a microwave oven comprises an electromagnetic wave shielding material, at least one chemiresistor, a microcontroller, and a digital display, wherein the at least one chemiresistor is overlaid by the electromagnetic wave shielding material, and the microcontroller is electronically coupled to the digital display. In an approach to detect the release of a hazardous chemical during a cooking cycle in a microwave oven, a processor receives a reading from at least one chemiresistor, wherein the at least one chemiresistor is configured to detect a chemical. A processor determines whether a pre-set threshold is met. In response to the pre-set threshold being met, a processor sends an alert.

Outgoing communication scam prevention

One embodiment provides a method comprising intercepting an incoming communication for a user, and determining whether to classify the incoming communication as a suspicious incoming communication based on content of the incoming communication and a learned classification model or learned signatures. The method further comprises monitoring one or more outgoing communications from the user, and invoking an action relating to scam prevention in response to determining an outgoing communication from the user is linked to a suspicious incoming communication.

Data resolution among disparate data sources

Cross-matching of match indexes assigned to respective data sources can be performed. Each match index can store at least a portion of data from a subset of data records in the data source to which the match index is assigned. Such data can be stored in the match index as match index records. Responsive to identifying matches between match records of different match indexes, the data records in the data sources to which the match records correspond can be updated to add links between the respective data records.

Auto-selection of hierarchically-related near-term forecasting models

Processing forms using artificial intelligence models

An application server may receive an input document including a set of input text fields and an input key phrase querying a value for a key-value pair that corresponds to one or more of the set of input text fields. The application server may extract, using an optical character recognition model, a set of character strings and a set of two-dimensional locations of the set of character strings on a layout of the input document. After extraction, the application server may input the extracted set of character strings and the set of two-dimensional locations into a machine learned model that is trained to compute a probability that a character string corresponds to the value for the key-value pair. The application server may then identify the value for the key-value pair corresponding to the input key phrase and may out the identified value.

On-demand activity feature generation for machine learning models

Technologies for generating user activity features for machine learning models on demand. The system receives a request for a user activity feature and a timestamp from a machine learning model, determines a data access mechanism, time window and feature computation algorithm, retrieves event and attribute data from a real-time data store, computes the feature and returns it to the model.

Systems and methods for reducing virality of content from domains with questionable reputation in a social networking system

Systems and methods are disclosed herein for receiving, by a social networking system, from a domain, a request to publish a content item. The social networking system publishes the content item and receives a request to reshare the content item from a user. The social networking system responsively determines a reputability score for the domain based on a number of other domains that point to the domain in the context of the social network interface, and determines whether the reputability score exceeds a threshold score. If the reputability score exceeds the threshold, the content item is published using a default ranking; otherwise a discounted ranking is used.

Generating a table of recommendations

In one example implementation according to aspects of the present disclosure, a computer-implemented method includes receiving a list of items to be purchased by a user. The method further includes performing an analysis of the list of items based at least in part on a cost for each item of the ...

Image analysis based document processing for inference of key-value pairs in non-fixed digital documents

An online system extracts information from non-fixed form documents. The online system receives an image of a form document and obtains a set of phrases and locations of the set of phrases on the form image. For at least one field, the online system determines key scores for the set of phrases. The online system identifies a set of candidate values for the field from the set of identified phrases and identifies a set of neighbors for each candidate value from the set of identified phrases. The online system determines neighbor scores, where a neighbor score for a candidate value and a respective neighbor is determined based on the key score for the neighbor and a spatial relationship of the neighbor to the candidate value. The online system selects a candidate value and a respective neighbor based on the neighbor score as the value and key for the field.

Training natural language system with generated dialogues

Multi-tenant, metadata-driven recommendation system

A method for generating a model for recommendations from an item data set for a target data set includes embedding a set of targets from the target data set in a shared coordinate space using a first embedding function.

System user attribute relevance based on activity

Techniques for using online user activity in determining relevance of attributes to improve computer functionality in generating recommendations of online content are disclosed herein. In some embodiments, a computer system calculates a corresponding relevance score for each attribute of a user.

Generating dialogue events for natural language system

Robust expandable dialogue system

Sensitive data service access

A system for accessing data includes and interface and a processor. The interface is configured to receive a request to move stored data to a new location and requestor information. The stored data comprises sensitive data. The processor is configured to determine whether read access, storage access, and deletion access are allowed based at least in part on the requestor information; in the event that read access, storage access, and deletion access are allowed: read the sensitive data at an original location using an original token, store the sensitive data to the new location and receive a new token associated with the new location, and delete the original token and the sensitive data at the original location.

Load balancing deterministically-subsetted processing resources using fractional loads

Continuation of US10579432B1: load balancing between a set of servers where subsets are assigned to clients via deterministic subsetting using fractional loads.

Server-based game activity suggestions by readiness score

Dynamic Memory Network

A novel unified neural network framework, the dynamic memory network, is disclosed. This unified framework reduces every task in natural language processing to a question answering problem over an input sequence. Inputs and questions are used to create and connect deep memory sequences. Answers are then generated based on dynamically retrieved memories.

Selective enforcement of privacy and confidentiality for optimization of voice applications

A computer-implemented method includes identifying a plurality of protected pieces from a conversation. The computer-implemented method further includes generating one or more confidence scores for each protected piece, wherein a confidence score is a degree of associativity between a protected piece and a type of sensitive information. The computer-implemented method further includes determining that the protected piece is associated with the type of sensitive information. The computer-implemented method further includes determining a type of protection action for each protected piece in the plurality of protected pieces. The computer-implemented method further includes performing the type of protection action for each protected piece in the plurality of protected pieces to form a modified conversation that is devoid of the sensitive information. A corresponding computer system and computer program product are also disclosed.

Communicating trace information between security zones

An application runs in a first security zone of a computer system. Trace information generated from running the application is stored in a first security zone. Filtered trace information is generated by removing specified information from the trace information... The adapter application transmits, from the first security zone to a datastore within the second security zone, the filtered trace information.