Abstract: A computer implemented method processes alerts. A computer system creates a representation of an alert received for processing. The computer system determines a similarity of the alert with previously processed alerts using the representation of the alert and representations of the previously processed alerts. A first evaluator in the computer system evaluates an alert level for the alert based on previously processed similar alerts in response to the similarity being above a similarity threshold for similar alerts. A second evaluator in the computer system evaluates, the alert level for the alert using a machine learning model in response to the similarity not being above the similarity threshold.

Abstract: A method for automatically providing instructional content based on wearable data from a wearable device is provided. The method may include automatically tracking and receiving, via the wearable device, real-time wearable data associated with a user and a detected user activity including automatically detecting body sensor data associated with the user and corresponding to the detected user activity. The method may also include automatically determining whether the tracked and received real-time wearable data associated with the user and the user activity indicates a user frustration event. The method may further include in response to automatically determining that the tracked and received real-time wearable data associated with the user and the user activity indicates the user frustration event, automatically retrieving and presenting the instructional content from instructional material related to the user activity and the user frustration event.

Abstract: An embodiment senses a thermal event in a data center, responsive to the sensed thermal event, identifies an affected server, and decides to migrate a workload from the affected server to a target server. The embodiment senses a performance metric, a temperature and a proximity from the thermal event for each of a plurality of servers, responsive to the sensed performance metric, the temperature and the proximity, computes a weighted score. The embodiment generates a ranking based on the weighted score of each of the plurality of servers. The embodiment selects the target server from the plurality of servers according to the ranking.

Abstract: The technologies described herein are generally directed toward migrating wireless coverage among access points. In an embodiment, a method can include identifying a first network function of a network provided by first network equipment, with the first network function serving a resource grid of a geographic area, and with the resource grid being designated to serve a selected user equipment in the geographic area. The method can further include instantiating a second network function of the network at second network equipment, with the second network function serving a first subset of the resource grid. Further, the method can include designating the first network function to serve a second subset of the resource grid, different from the first subset of the resource grid. The method can further include designating the second network function to serve, after connection of the selected user equipment, the second subset of the resource grid.

Abstract: Enhanced CU-UP change associated with a device can be performed and managed. CU-CP can trigger transition from source CU-UP to target CU-UP. Before bearer context modification request is communicated from CU-CP to source CU-UP, CU-CP can effectuate communication path update to transition from first uplink path, associated with source CU-UP, DU, and UPF, to second uplink path, associated with target CU-UP, DU, and UPF. Target CU-UP can buffer uplink data until CU-CP receives uplink PDCP count from source CU-UP and forwards that count to target CU-CP. CU-CP or UPF can transition or facilitate transitioning from first downlink path associated with source CU-UP, DU, and UPF, to second downlink path, associated with target CU-UP, DU, and UPF. Target CU-UP can buffer downlink data until CU-CP receives downlink PDCP count from source CU-UP and forwards that count to target CU-CP. Source CU-UP can transfer remaining downlink data it has to DU.

Abstract: Requests are received from viewing users for rendering a virtual world in a computing environment that mimics real world, where at least one user appears in the virtual world. For each of the viewing users, a processor determines a group to which a viewing user belongs, determines a virtual representation of the at least one user that is mapped to the group, and renders the virtual representation of the at least one user in the virtual world to a device associated with the viewing user. Based on the determined mapping of the group to the virtual representation, the virtual representation of the at least one user rendered on the device associated with the viewing user appears at the same time differently from the virtual representation of the at least one user rendered on another device associated with another viewing user among the viewing users.

Abstract: A disclosed method facilitates an increase in utilization with respect to a resource quota allocated to a tenant from a shared resource pool. The method includes transmitting a lease request to a quota service on behalf of the tenant, where the lease request identifies a processing task and specifies quantity of cloud-based resources requested from the shared resource pool for execution of the processing task. The method further provides for determining, based on a feedback signal received from the quota service, whether grant of the lease request would cause the tenant to exceed a resource quota allocated to the tenant and dynamically decreasing parallelism of active tasks being processed by the cloud-based resources on behalf of the tenant in response to determining that grant of the lease request would cause the tenant to exceed the resource quota limit.

Abstract: A system provides capacity-based load balancing across model endpoints of a cloud-based artificial intelligence (AI) model. The system includes a consumption determination engine executable to determine a net resource consumption for processing tasks in a workload generated by a client application for input to the trained machine learning model. The system also includes a load balancer that determines a distribution of available resource capacity in a shared resource pool comprising compute resources at each of the multiple model endpoints. The load balancer allocates parallelizable tasks of the workload among the compute resources at the multiple model endpoints based on the net resource consumption of the tasks and on the distribution of available resource capacity in the shared resource pool.

Abstract: Methods and apparatuses for identifying groups of electronic messages, generating conversation documents using the groups of electronic messages, and indexing the conversation documents to improve search ranking and retrieval for content contained within the electronic messages are described. In some cases, understanding the contents of a group of electronic messages may require context from outside the group of electronic messages, such as context provided by electronic messages outside of the group of electronic messages or context provided by electronic documents linked to by messages of the group of electronic messages. The identification of a group of electronic messages includes detecting a conversation boundary that separates a first set of messages from a second set of messages, which may be determined using machine learning approaches or various heuristics.

Abstract: A disclosed method reduces memory consumption of a trained sequential model. The method includes receiving, from a client application, an initial processing request identifying an input sequence to be processed by the trained sequential model and an initial value for an output size parameter specifying a requested size of output from the trained sequential model. The method further includes sequentially transmitting, to the trained sequential model, multiple partial processing requests based on the initial processing request that each specify a fraction of the initial value as the output size parameter and receiving a sequence of output responses from the trained sequential model generated in response to processing the multiple partial processing requests. The method further provides for returning, to the client application, a final merged response that includes the sequence of output responses.

Abstract: Indoor localization, which estimates the location of a wireless device using Wi-Fi/Zigbee/Bluetooth, is increasingly important for Industry 4.0 applications, such as tracking of robots and large-scale inventory management. Existing approaches have certain practical issues related to enterprise hardware as specific data required by them is not always available due to security and other reasons. Present disclosure provides system and method that implement indoor Location Aggregator Model, which combines multiple estimation location models to provide localization to wireless/user devices while being compliant to enterprise environments. More specifically, channel state information is used for estimating position and identification of candidate sub-region within a region. Further, at least one Location Estimation Model is identified based on the identified candidate sub-region by all LEMS and an accuracy map. The identified LEM is then used for determining an actual location of a user device.

Abstract: A method, computer system, and a computer program product for environment monitoring. Exemplary embodiments may include emitting one or more soundwaves within an environment, detecting the emitted one or more soundwaves, and identifying motion of an object within the environment based on an analysis of the detected one or more soundwaves.

Abstract: A computer-implemented method for displaying advertisements. The method includes associating feature metadata with a product video advertisement of a product. The method includes identifying, based on the feature metadata, a feature of the product corresponding with a section of the product video advertisement when a triggering event is detected during playing of the product video advertisement. The method includes displaying a second product video advertisement of a second product that includes the feature of the product.

Abstract: A method for dynamically rearranging one or more packages for delivery on a delivery vehicle in response to unforeseen changes in a pre-planned delivery route. The method detects, via one or more sensor-enabled shelves in a delivery vehicle, an address and package details of a respective delivery package. The method further integrates the one or more sensor-enabled shelves with a route map that depicts the pre-planned delivery route. The method further assigns new values to the one or more sensor-enabled shelves, in response to a change in the pre-planned delivery route, indicating a new sequence of delivery for each respective delivery package. The method further notifies a delivery person which delivery package is to be delivered at a current stop by illuminating the one or more sensor-enabled shelves containing the respective delivery package.

Abstract: The technology described herein is directed towards near-instantaneous optimization of radio resource control (RRC) rejection wait times per user equipment, which can be per user equipment of a group of multiple user equipment. In inference mode, a trained wait time prediction model (e.g., an artificial intelligence/deep reinforcement learning (AI/DRL) model in a base station) adaptively determines the per-UE rejection wait times based on current (measured) network state data, including current centralized unit and distributed unit utilization levels. The technology described herein operates to decreases the overall RRC request rate, while maintaining very high utilization of gNodeB resources, which results in increased per-UE energy efficiency and decreased signaling traffic loads.

Abstract: A method, system, and non-transitory computer readable medium storing instructions executable by a computer processor, the instructions comprising functionality for determining remediation categories for a plurality of wells include: obtaining, by a computer processor, a plurality of attributions for the plurality of wells; obtaining, by the computer processor, a set of weightages for each of the plurality of attributes for the plurality of wells; generating, by the computer processor, a first Well Health Index (WHI) set for a well among the plurality of wells; generating, by the computer processor, a second WHI set for the well; determining, by the computer processor, a sum of weightages of the second WHI set for the well; determining, by the computer processor, a remediation category among a plurality of remediation categories for the well based on the sum of weightages; and generating, by the computer processor, tracking notifications and action reports for the well.

Abstract: Described are techniques for bandwidth allocation in container orchestration systems. The techniques include monitoring storage bandwidth designated for one or more storage operations in association with one or more storage volumes attached to a server that hosts containerized applications, where a fixed amount of bandwidth provided to the server is divided between the storage bandwidth available for the one or more storage operations and network bandwidth available for network operations. The techniques further include determining that increasing the storage bandwidth will improve performance of at least one of the one or more storage operations. The techniques further include identifying an amount of the network bandwidth as being available for reallocation to the storage bandwidth, and reallocating the amount of the network bandwidth to the storage bandwidth for use in performing the one or more storage operations.

Abstract: Embodiments of the present invention provide an approach for authenticating video data utilizing a contextual identifier. Specifically, a set of video data including contextual identifiers is captured. A respective cryptographic signature for each frame of the set of video data is generated utilizing metadata of the respective frame. Each frame of the set of video data and the respective cryptographic signature are stored on a blockchain, and the set of video data is authenticated utilizing the respective cryptographic signatures stored on the blockchain.

Abstract: Embodiments of the present disclosure provide systems and methods for synchronizing an unaligned audio stream and a corresponding unaligned video stream of real time streaming media. A non-limiting disclosed method comprises performing, using a video classifier, video reference point classification of a video stream based on an audio-video dataset; performing, using an audio classifier, audio reference point classification of the audio stream based on the audio-video dataset. The system correlates object related reference points in video segments of the video stream and in audio segments of the audio stream to identify a set of audio-video synchronization candidates. The system compares context of the set of audio-video synchronization candidates to identify an audio-video synchronization candidate to synchronize the audio stream and video stream based on reference point alignment.

Abstract: Aspects of the present disclosure relate generally to robotic process automation (RPA) and, more particularly, to systems, computer program products, and methods of accelerating RPA tasks that interact with web pages with machine learning generation of network requests. For example, a computer-implemented method includes receiving, by a processor, plural runtime variables of a bot executing client-side web application code; generating, by the processor, a network request from a machine learning model given input of the plural runtime variables of the bot; and sending, by the processor, the network request generated from the machine learning model to the bot.