Abstract: A fabrication method comprising: forming a mask of an amorphous material over a crystalline surface of a substrate, the mask having a pattern of openings defining areas of an active region in which one or more components of one or more active devices are to be formed, the mask further defining a non-active region in which no active devices are to be formed; and forming a deposition material through the mask by an epitaxial growth process. The deposition material thus forms in the openings of the active region. The pattern of openings through the mask further comprises one or more reservoirs formed in the non-active region, each of the reservoirs being connected by the pattern of openings in the mask to at least one of the areas in the active region, and the deposition material forming in the reservoirs as part of the epitaxial growth.

Abstract: This document generally relates to techniques for adaptively triggering charging notifications for a battery-powered user device. One example includes a method or technique that can be performed on a computing device. The method or technique can include obtaining a battery charge level of a battery of a battery-powered user device and accessing user data for a user of the battery-powered user device. The user data can reflect previous usage sessions of the user with the battery-powered user device. The method or technique can also include predicting a future usage session of the user with the battery-powered user device based on the user data. The method or technique can also include estimating confidence that the battery will last through the future usage session based on the battery charge level and the predicted future usage session. The method or technique can also include triggering a charging notification prior to the future usage session based on the estimated confidence.

Abstract: Systems and methods for conducting a videoconference including receiving multimedia streams of a plurality of participants in a multimedia conference, the multimedia streams including audio components and video components and displaying video tiles of the participants on a display screen. The audio components and/or the video components of the multimedia streams are analyzed to detect characteristics indicative of a first participant and a second participant having a first conversation with each other. Camera positions on the computing devices of the participants are identified. In response to identifying that the first participant and the second participant are having the first conversation with each other, a video tile for the first participant and a video tile for the second participant are moved to edges of the respective display screens toward the camera positions.

Abstract: Methods and systems for generating and using a semantic index are provided. In some examples, content data is received. The content data includes a plurality of subsets of content data. Each of the plurality of subsets of content data are labelled, based on a semantic context corresponding to the content data. The plurality of subsets of content data and their corresponding labels are stored. The plurality of subsets of content data are grouped, based on their labels, thereby generating one or more groups of subsets of content data. Further, a computing device is adapted to perform an action, based on the one or more groups of subsets of content data.

Abstract: Systems and methods for inferring and notifying an end user about reduced availability of a target user or group of target users in a time range of interest. For instance, the reduced availability service includes components for collecting calendar event information and calendar settings information corresponding to a calendar of a target user, generating an interval graph data structure based on the collected calendar information, determining working hours for the target user, identifying periods of time where reduced availability is determined in the target user's calendar, and generating a notification of the target user's reduced availability for alerting the end user.

Abstract: Systems and methods are provided for interactively highlighting a region as pixel data on a screen and automatically retrieving context data associated with content of the highlighted region for contextual notetaking. The highlighted region includes at least a part of one or more windows and one or more applications associated with the one or more windows. The disclosed technology determines a context associated with content of the highlighted region and automatically retrieves context data that are contextually relevant to the content. Notes data are generated based on an aggregate of the highlighted content, window-specific context data, application-specific context data, and user-specific context data. A notetaking application retrieves stored the notes data from a notes database and displays the notes data for recall and for use. The contextual notetaking enables the user reducing a burden of performing manual operations for notetaking and utilizing notes that are enriched relevant data by context.

Abstract: In examples, a schema augmentation system for exploratory research leverages intelligence from a machine learning model to augment such tasks by leveraging intelligence derived from machine learning capabilities. Augmenting tasks include schematization of content, such as information units and groupings of information units. Based on the schematization of such content, semantic proximities for information units are determined. The semantic proximities may be used to identify and present potentially relevant information units, for example to accelerate the exploratory research task at hand. As such, users engaged in consumption of heterogenous content (e.g., across client applications and/or content sources), may receive machine-augmented support to find potential information units.

Abstract: A system and method automatically generating a title for a cluster of documents includes accessing a plurality of documents that have been categorized as belonging to a document cluster and providing the plurality of documents as an input to a trained title generating machine-learning (ML) model. The trained title generating ML model is trained for generating a title for a document and provides a titles for each of the plurality of documents. An embedding is created for the generated titles and then an embedding is generated for the document cluster. A similarity between the embeddings for the titles and embedding for the document cluster is measured to identify titles that are more similar to the embedding for the document cluster and based on the similarity one or more titles are selected as title candidates for the document cluster and provided as an output.

Abstract: A connector includes a housing including a plug opening and a cable opening. A cable extends through the cable opening away from the housing. An electronic plug is connected to the cable within the housing and extends through the plug opening away from the housing. The electronic plug is selectively moveable relative to the housing between an extended position and a retracted position when a pulling force is applied to the cable. A bias mechanism biases the electronic plug to the extended position.

Abstract: Systems and methods for monitoring health of virtual machines (VMs) include determining a leader virtual machine (VM) count for a group of VM nodes hosted on a plurality of computing devices; selecting a number of the VM nodes of the group to serve as leader VMs for the group, the number of the VM nodes selected corresponding to the leader VM count; and periodically performing a peer VM monitoring process. The peer VM monitoring process includes periodically storing health information for each of the VM nodes of the group in a data store; periodically accessing the health information of each of the VM nodes to identify sick VMs using each of the leader VMs, respectively; and automatically performing a healing process on the sick VMs to improve a performance of the sick VMs.

Abstract: Systems and methods are provided for performing privacy transformation of data to protect privacy in data analytics under the multi-access edge computing environment. In particular, a policy receiver in an edge server receives privacy instructions. Inference determiner in the edge server in a data analytics pipeline receives data from an IoT device and evaluates the data to recognize data associated with personally identifiable information. Privacy data transformer transforms the received data with inference for protecting data privacy by preventing exposure of private information from the edge server. In particular, the privacy data transformer dynamically selects a technique among techniques for removing information that is subject to privacy protection and transforms the received data using the technique.

Abstract: Examples are disclosed that relate to display devices having a common light path region. One example provides a display device comprising a light source configured to emit illumination light along an illumination path, and a spatial light modulator configured to modulate the illumination light and emit the modulated illumination light as image light along an imaging path, wherein at least a portion of the illumination path and at least a portion of the imaging path extend through a common light path region. The display device further comprises one or more optical elements positioned within the common light path region, at least one optical element being configured to guide the illumination light as the illumination light travels through the common light path region toward the spatial light modulator, and shape the image light as the image light travels through the common light path region.

Abstract: In an example embodiment, a deep learning model is used to learn embedding representations of a heterogeneous information network, where the embedding represents entity-specific properties and network environment properties. Position-aware embeddings specific to the heterogeneous information network may be used as input features of the deep learning model. Furthermore, meta-path embedding specific to the heterogeneous information network may also be used as input features of the deep learning model. Modified embedding propagation methods are further designed to explore better ways to capture network meta-path properties.

Abstract: The description relates to hinged devices, such as hinged computing devices. One example can include first and second portions that rotate around a hinge shaft that is fixedly secured to the first portion and rotationally secured to the second portion. The second portion defining a first contact surface spaced apart from a second contact surface. Multiple friction clips friction fit around the hinge shaft and rotating with the hinge shaft between the first contact surface and the second contact surface.

Abstract: Various embodiments enable a web site to expose various search providers that it promotes to browsers that navigate to that site. In at least some embodiments, the search providers are exposed using the HTML of a web page. When a browser navigates to that site, it processes the HTML of the associated web page and exposes the promoted search providers to the user. In at least some embodiments, the browser can permit use of one or more search providers without requiring the user to install the search provider. Alternately or additionally, in at least some other embodiments, the browser can enable the user to quickly and efficiently install the search provider so that the search provider can be available whenever the user uses their browser application.

Abstract: Methods and systems are provided for converting a loop (e.g., a cursor loop) to a declarative Structured Query Language (SQL) query that invokes a custom aggregate function. The loop includes a select query and a loop body that includes a program fragment that can be evaluated over a result set of the select query one row at a time. The system verifies that the loop body does not modify a persistent state of the database. A custom aggregate function that expresses the loop body is automatically constructed according to a contract. An aggregate class comprising aggregation methods of the contract are used to construct the aggregate function based on results of static analysis. The select query is automatically rewritten to form a declarative SQL query that invokes the custom aggregate function. The declarative SQL query may be executed by a database management system (DBMS) SQL server.

Abstract: Systems and methods for managing a power budget are provided. The method includes designating, by a power budget manager implemented on at least one processor, each of one or more applications with an individual quality of service (QoS) designation, the one or more applications executable by the at least one processor, assigning, by the power budget manager, a throttling priority to each of the one or more applications based on the individual QoS designations, determining, by the power budget manager, whether a platform mitigation threshold is exceeded, and responsive to determining that the platform mitigation threshold is exceeded, throttling, by the power budget manager, processing power allocated to at least one application of the one or more applications based on the throttling prioritization.

Abstract: Performing set operations using sparse matrix operations offered by a multi-core processing unit (such as a graphics processing unit). The set operation is converted into operand matrices, and sparse matrix operations, foregoing the use of hash tables. The input set is converted into a matrix, a matrix operation corresponding to the set operation is identified, and one or more operands of the set operation are also represented within a matrix. The matrix operation is then performed on these matrices to obtain an output matrix, which is then converted to an output set.

Abstract: Techniques of implementing software filtered non-volatile memory in a computing device are disclosed herein. In one embodiment, a method includes detecting an entry being written to a guest admin submission queue (gASQ) by a memory driver of a virtual machine hosted on the computing device. Upon detecting the entry written to the gASQ by the memory driver, the command in the entry is analyzed to determine whether the command is allowed based on a list of allowed or disallowed commands. In response to determining that the command in the entry is not allowed, without sending the command to the non-volatile memory, generating an execution result of the command in response to the entry being written to the gASQ by the memory driver. As such, potentially harmful commands from the memory driver are prevented from being executed by the non-volatile memory.

Abstract: A method of docking an accessory device includes, at the accessory device, receiving (324) a first transmission energy for wireless charging from an electronic device; in accordance with receiving the first transmission energy: setting (326) a timer, and entering (328) a docked mode; and exiting (330) the docked mode upon expiration of the timer.

Abstract: Edit automation functionality generalizes edits performed by a user in a document, locates similar text, and recommends or applies transforms while staying within a current workflow. Source code edits such as refactoring are automated. The functionality uses or provides anchor target lists, temporal edit patterns, edit graphs, automatable edit sequence libraries, and other data structures and computational techniques for identifying locations appropriate for particular edits, for getting transforms, for selecting optimal transforms, for leveraging transforms in an editing session or later, and for displaying transform recommendations and results. The edit automation functionality enhances automation subtool generation, discoverability, and flexibility, for refactoring, snippet insertion, quick actions in an integrated development environment, and other automatable edit sequences.

Abstract: A sensor device includes a first antenna configured to receive an interrogation radio frequency (RF) signal; a first sensor configured to detect a first state of a first sensed signal, to output a first sensor signal responsive to the first state meeting a first threshold condition; a first circuit coupled to the first antenna, the first circuit configured to receive the interrogation RF signal from the first antenna, modulate a second RF signal, and transmit the modulated second RF signal as a RF response signal; and a first tag controller configured to receive the first sensor signal and to selectively connect or disconnect the first antenna to or from ground responsive to the first sensor signal, wherein connecting the first antenna to ground prevents the first antenna from transmitting the RF response signal.

Abstract: Examples are provided relating to recovering depth data from noisy phase data of low-signal pixels. One example provides a computing system, comprising a logic machine, and a storage machine holding instructions executable by the logic machine to process depth data by obtaining depth image data and active brightness image data for a plurality of pixels, the depth image data comprising phase data for a plurality of frequencies, and identifying low-signal pixels based at least on the active brightness image data. The instructions are further executable to apply a denoising filter to phase data of the low-signal pixels to obtain denoised phase data and not applying the denoising filter to phase data of other pixels. The instructions are further executable to, after applying the denoising filter, perform phase unwrapping on the phase data for the plurality of frequencies to obtain a depth image, and output the depth image.

Abstract: Techniques for aligning images generated by two cameras are disclosed. This alignment is performed by computing a relative 3D orientation between the two cameras. A first gravity vector for a first camera and a second gravity vector for a second camera are determined. A first camera image is obtained from the first camera, and a second camera image is obtained from the second camera. A first alignment process is performed to partially align the first camera's orientation with the second camera's orientation. This process is performed by aligning the gravity vectors, thereby resulting in two degrees of freedom of the relative 3D orientation being eliminated. Visual correspondences between the two images are identified. A second alignment process is performed to fully align the orientations. This process is performed by using the identified visual correspondences to identify and eliminate a third degree of freedom of the relative 3D orientation.

Abstract: Technologies for grouping images, and ranking the images and the groupings, based on entities shown in the images. Images may be grouped based on faces shown in the images. Different images with faces that indicate the same entity (e.g., Adam) may be automatically grouped together. Different images with faces that indicate the same multiple entities (e.g., the people in my family) may also be automatically grouped together. Such automatic grouping may be based on facial recognition technologies. Further, images and groups of images may be automatically ranked based on the faces shown and entities represented. Such rankings may also be influenced by adjacent data that indicates family and friends and the like, and that can be used to identify such entities in the images.

Abstract: A method for signal request caching is described. Signal requests are received at a signal processor from a plurality of computing devices. The received signal requests are routed to a signal data store. An ingestion rate of the received signal requests is monitored by the signal processor. When the ingestion rate meets a signal request rate threshold of the signal data store, overflow signal requests of the received signal requests are automatically routed to an intermediate cache instead of the signal data store. The overflow signal requests within the intermediate cache are aggregated into one or more signal packages, each of the one or more signal packages containing a plurality of overflow signal requests. The one or more signal packages are stored at the signal data store.

Abstract: Described herein is electromagnetic shielding that is configured to attenuate electromagnetic interference (EMI) by at least a threshold amount when the EMI has a frequency within a predefined frequency range. The electromagnetic shielding includes a layer of metal, such as aluminum foil, and a layer of thermoplastic polymer fabric (such as woven polyethylene fabric), where the electromagnetic shielding has several apertures that extend therethrough. The electromagnetic shielding is at least partially draped over electronic equipment that is to be shielded from EMI.

Abstract: The disclosed technology controls a digital inking device by communicating electrostatic inking signals between the digital inking device and an ink-receiving computing device in an inking mode enabling the digital inking device to render digital ink in a display of the ink-receiving computing device via the electrostatic inking signals, detecting proximity of a peripheral communication device relative to the digital inking device, transitioning the digital inking device from the inking mode to a non-inking mode that terminates communication of the electrostatic inking signals between the digital inking device and the ink-receiving computing device, based at least in part on the detecting operation, and communicating electrostatic data signals in the non-inking mode between the digital inking device and the peripheral communication device in the non-inking mode, based at least in part on the transitioning to the non-inking mode.

Abstract: Cloud services intelligently provision new VMs for a VM scale set when the original label that included the OS or other software images used to provision existing VMs of the VM scale set is no longer available or has been changed. Metadata of the existing VMs are analyzed to identify an OS image or other software image used to provision the existing VMs. The metadata also reveals updates that are running on the new VMs. These updates include software that was not part of the original label used to provision the existing VMs and are used to find the second label in a label database. The second labels include the OS or software of the existing VMs—or a later version thereof—and some portion of the updates. A VM-provisioning service uses the second label to provision the new VM.

Abstract: Examples are disclosed herein that relate to driving a resonant scanning mirror system using a linear LC resonant driving scheme. In one example, a resonant scanning mirror system includes a scanning mirror, first and second mirror drive elements, and a drive circuit to drive the scanning mirror at a resonant frequency. The drive circuit includes one or more signal sources configured to create a first source signal and a second source signal that is 180 degrees out of phase with the first source signal. The drive circuit further includes a buffer stage configured to receive the first and second source signals and output first and second drive signals, a first resonant LC stage configured to amplify the first drive signal for provision to the first mirror drive element, and a second resonant LC stage configured to amplify the second drive signal for provision to the second mirror drive element.

Abstract: Disclosed herein are innovations for bitstreams having clean random access (CRA) pictures and/or other types of random access point (RAP) pictures. New type definitions and strategic constraints on types of RAP pictures can simplify mapping of units of elementary video stream data to a container format. Such innovations can help improve the ability for video coding systems to more flexibly perform adaptive video delivery, production editing, commercial insertion, and the like.

Abstract: An EMI attenuation device includes a housing stator, a fan rotor, and an electrical bridge therebetween. The housing stator has an aperture therethrough, and at least a portion of the housing stator is electrically conductive. The fan rotor is adjacent to the aperture and has a rotational axis relative to the housing stator and a proximate surface proximate the housing stator. The fan rotor is electrically conductive, and the proximate surface is continuous around a rotational direction of the fan rotor. The electrical bridge is between the proximate surface of the fan rotor and a contact surface of the housing stator.

Abstract: Video decoding innovations for multithreading implementations and graphics processor unit (“GPU”) implementations are described. For example, for multithreaded decoding, a decoder uses innovations in the areas of layered data structures, picture extent discovery, a picture command queue, and/or task scheduling for multithreading. Or, for a GPU implementation, a decoder uses innovations in the areas of inverse transforms, inverse quantization, fractional interpolation, intra prediction using waves, loop filtering using waves, memory usage and/or performance-adaptive loop filtering. Innovations are also described in the areas of error handling and recovery, determination of neighbor availability for operations such as context modeling and intra prediction, CABAC decoding, computation of collocated information for direct mode macroblocks in B slices, reduction of memory consumption, implementation of trick play modes, and picture dropping for quality adjustment.

Abstract: Innovations in use of chroma quantization parameter (“QP”) offsets when determining a control parameter for deblock filtering. For example, as part of encoding, an encoder sets a picture-level chroma QP offset and slice-level chroma QP offset for encoding of a slice of a picture. The encoder also performs deblock filtering of at least part of the slice, where derivation of a control parameter considers only the picture-level chroma QP offset. The encoder outputs at least part of a bitstream including the encoded content. As part of decoding, a corresponding decoder sets a picture-level chroma QP offset and a slice-level chroma QP offset for decoding of a slice of a picture, but derivation of a control parameter for deblock filtering considers only the picture-level chroma QP offset.

Abstract: Innovations in syntax and semantics of coded picture buffer removal delay (“CPBRD”) values potentially simplify splicing operations. For example, a video encoder sets a CPBRD value for a current picture that indicates an increment value relative to a nominal coded picture buffer removal time of a preceding picture in decoding order, regardless of whether the preceding picture has a buffering period SEI message. The encoder can signal the CPBRD value according to a single-value approach in which a flag indicates how to interpret the CPBRD value, according to a two-value approach in which another CPBRD value (having a different interpretation) is also signaled, or according to a two-value approach that uses a flag and a delta value. A corresponding video decoder receives and parses the CPBRD value for the current picture. A splicing tool can perform simple concatenation operations to splice bitstreams using the CPBRD value for the current picture.

Abstract: Embodiments of the present disclosure include techniques for compressing data using a tree encoded bit mask that may result in higher compression ratios. In one embodiment, an input vector having a plurality of values is received by a first plurality of switch circuits. Selection of the input values is controlled by sets of bits from the bit mask. The sets of bits specify locations of portions of the input vector where particular value of interest reside. The switch circuits output multiple values of the input vector, which include the particular value of interest. A second stage of switch circuits is controlled by logic circuit that detects values on the outputs of the first stage of switch circuits and outputs the values of interest. In some embodiments, the values of interest may be non-zero values of a sparse input vector, and the switch circuits may be multiplexers.

Abstract: Disclosed in some examples are methods, systems, devices, and machine-readable mediums which provide for sidebar communication threads forked from, or related to, a principal thread. Messages in the sidebar communication thread may include a history of the principal thread, including one or more messages from the principal thread, and may include a proposed principal thread message that is the subject of the sidebar thread discussion. The sidebar thread may also include sidebar thread messages that carries the conversation of the sidebar thread participants. Once a termination condition is reached for the sidebar thread, the sidebar thread terminates and either the proposed principal thread message (as potentially modified by participants of the sidebar thread) becomes an accepted principal thread message and it is posted to the principal thread as if it was sent by the sidebar initiator or no message is posted (e.g., the proposed principal thread message is rejected).

Abstract: An imaging system is configured to use an array of time-of-flight (ToF) pixels to determine depth information using the ToF imaging method and/or the stereo imaging method. A light emitting component emits light to illuminate a scene and a light detecting component detects reflected light via the array of ToF pixels. A ToF pixel is configured to determine phase shift data based on a phase shift between the emitted light and the reflected light, as well as intensity data based on an amplitude of the reflected light. Multiple ToF pixels are shared by a single micro-lens. This enables multiple offset images to be generated using the intensity data measured by each ToF pixel. Accordingly, via a configuration in which multiple ToF pixels share a single micro-lens, depth information can be determined using both the ToF imaging method and the stereo imaging method.

Abstract: Innovations in adaptive encoding and decoding for units of a video sequence can improve coding efficiency when switching between color spaces during encoding and decoding. For example, some of the innovations relate to adjustment of quantization or scaling when an encoder switches color spaces between units within a video sequence during encoding. Other innovations relate to adjustment of inverse quantization or scaling when a decoder switches color spaces between units within a video sequence during decoding.

Abstract: A deep learning model trained to learn to predict source code is tuned for a target source code generation task through reinforcement learning using a reward score that considers the quality of the source code predicted during the tuning process. The reward score is adjusted to consider code-quality factors and source code metrics. The code-quality factors account for the predicted source code having syntactic correctness, successful compilation, successful execution, successful invocation, readability, functional correctness, and coverage. The source code metrics generate a score based on how close the predicted source code is to a ground truth code.

Abstract: Disclosed herein is a system and method to synchronize files from a hosted file system where the client endpoint devices have direct access to the files stored on a hosted file store. When the client desires to synchronize with the files stored on the hosted file store the process identifies a sync role that can process the request. The particular sync role does not need to maintain state with the client device or the file store. The state necessary for the processing of the synchronization process is maintained separate from the sync role such that any sync role in the system can pick up the requests and process the requests. Once the role has been selected the synchronization process proceeds as normal. As state is not maintained by the sync role multiple roles can process the request or two roles can handle different portions of the process.

Abstract: A computing device includes a first portion and a second portion rotatable relative to each other about a hinge. A bumper is located between the hinge and the first portion. The bumper includes an impact resistant material. The impact resistant material absorbs some or all of the force from an impact between a hinge body and the first portion.

Abstract: Innovations in syntax and semantics of coded picture buffer removal delay (“CPBRD”) values potentially simplify splicing operations. For example, a video encoder sets a CPBRD value for a current picture that indicates an increment value relative to a nominal coded picture buffer removal time of a preceding picture in decoding order, regardless of whether the preceding picture has a buffering period SEI message. The encoder can signal the CPBRD value according to a single-value approach in which a flag indicates how to interpret the CPBRD value, according to a two-value approach in which another CPBRD value (having a different interpretation) is also signaled, or according to a two-value approach that uses a flag and a delta value. A corresponding video decoder receives and parses the CPBRD value for the current picture. A splicing tool can perform simple concatenation operations to splice bitstreams using the CPBRD value for the current picture.

Abstract: A data processing system implements techniques for restricting which notifications and/or conversations are presented on a plurality of user devices associated with a user. Each user device is associated with a device category. The device categories are associated with time category information that associates each of a plurality of time periods with permitted device category information that indicates which categories of user devices associated with the user are permitted to present notifications that messages have been received for a particular time period and the categories of messages for which the notifications may be presented for that time period and/or which categories of conversations may be presented or hidden. The message category may be determined based on user input or by analyzing the message content with a machine learning model configured to predict the message category.

Abstract: Techniques are described herein that are capable of provisioning a trusted execution environment (TEE) based on (e.g., based at least in part on) a chain of trust that includes a platform on which the TEE executes. Any suitable number of TEEs may be provisioned. For instance, a chain of trust may be established from each TEE to the platform on which an operating system that launched the TEE runs. Any two or more TEEs may be launched by operating system(s) running on the same platform or by different operating systems running on respective platforms. Once the chain of trust is established for a TEE, the TEE can be provisioned with information, including but not limited to policies, secret keys, secret data, and/or secret code. Accordingly, the TEE can be customized with the information without other parties, such as a cloud provider, being able to know or manipulate the information.

Abstract: A computing system is configured to perform zero-trust domain name resolution. The computing system includes applications coupled to a zero-trust client. The zero-trust client is configured to receive requests for IP addresses corresponding to endpoint identifiers for internet connected endpoints. The zero-trust client includes a synthetic DNS service configured to identify synthetic IP addresses for the endpoint identifiers. The zero-trust client provides the synthetic IP addresses for the endpoint identifiers to the applications. The zero-trust client sends data traffic from the applications to a zero-trust service with the synthetic IP addresses and sends corresponding endpoint identifiers to the zero-trust service in a fashion that allows the synthetic IP addresses to be correlated to the endpoint identifiers at the zero-trust service.

Abstract: Methods, systems, apparatuses, and computer program products are provided for reconstructing network activity. A network activity monitor is configured to monitor network activity for various network entities. Based on the monitoring, a set of features may be obtained for each network entity. A determination may be made for a number of vertices suitable for describing the sets of features in a multidimensional space. In some implementations, the vertices may define a convex hull in the multidimensional space. Each of the vertices may be assigned a different usage pattern that represents a certain type of network usage types. Reconstructed network activity for a particular network entity may be represented as a weighted combination of the usage patterns. Based on the reconstruction, a network anomaly may be detected, a network may be modified, and/or an alert may be generated.

Abstract: Computerized systems are provided for detecting one or more tables and performing information extraction and analysis on any given table. Information can be extracted from one or more cells or fields of a table and feature vectors representing individual cells, rows, and/or columns of the table can be derived and concatenated together. In this way, embodiments can use some or all of the “context” or values contained in various feature vectors representing some or all of a single table as signals or factors to consider when generating a decision statistic, such as a classification prediction, for a particular cell.