Abstract: Computer systems and methods are provided for determining authenticity of a document. A computer system receives first image data from a remote third party. The first image data includes a first image of a document. The computer system generates an image hash based on at least a portion of the received image data. The computer system compares the generated image hash with a stored image hash to determine whether the generated image hash meets image matching criteria. The stored image hash is associated with a first predetermined authentication decision. In accordance with a determination that the generated image hash meets the image matching criteria, the computer system transmits the first predetermined authentication decision to the remote third party.

Abstract: Techniques, methods and/or apparatuses are disclosed that enable prediction of cyber risks of assets of networks. Through the disclosed techniques, a cyber risk prediction model, which may be a form of a machine learning model, may be trained to predict cyber risks. The cyber risk model may be provided to a cyber risk predictor two predict cyber risks of an asset, without the need to scan the asset at a very deep scan level.

Abstract: Techniques are disclosed relating to training a machine learning model to handle adversarial attacks. In some embodiments, a computer system perturbs, using a set of adversarial attack methods, a set of training examples used to train a machine learning model. In some embodiments, the computer system identifies, from among the perturbed set of training examples, a set of sparse perturbed training examples that are usable to train machine learning models to identify adversarial attacks, where the set of sparse perturbed training examples includes examples whose perturbations are below a perturbation threshold and whose classifications satisfy a classification difference threshold. In some embodiments, the computer system retrains, using the set of sparse perturbed training examples, the machine learning model. The disclosed techniques may advantageously enable a machine learning model to correctly classify data associated with adversarial attacks.

Abstract: A mobile malicious code classification method based on feature selection includes extracting Application Programming Interface (API) feature information including a package name, a class name, a method name and a description from a malicious application of a predefined category, vectorizing a training dataset generated using the package name, the class name and the method name in the API feature information for deep learning, learning the vectorized training dataset to generate a classifier, probabilistically classifying to fit a target malicious application into a category, and defining the category of the target malicious application using a result of the classification and outputting a classification important API. Accordingly, it is possible to deal with malicious behaviors of malicious applications quickly and prevent damage caused by the malicious behaviors.

Abstract: An interface accessing method, a display apparatus, and an electronic device, which relates to the technical field of displaying. The method includes: by the on-screen display board card, when the on-screen display board card receives a viewing instruction of a microphone-volume interface, accessing the interface periodically according to a predetermined accessing mode, wherein the predetermined accessing mode includes, at each of periods, after each time of performance of a first preset time quantity of interface accessing operations to microphone-volume data, performing a second preset time quantity of interface accessing operations to video-channel-access-state data; and by the on-screen display board card, via the interface, reading alternately the microphone-volume data and the video-channel-access-state data from the field-programmable-gate-array board card.

Abstract: During a conversation the participants of the conversation may use catchphrases that are associated with particular characters, media assets, or media personalities. The catchphrases may be associated with information such as the media asset where the catchphrase was originally spoken, the speaker, and keywords relating to the media asset. Based on this associated information and the participant's use of the catchphrase, the system may provide a media asset recommendation to the participant for a media asset that does not include the catchphrase.

Abstract: Apparatus and methods for continuous real-time authentication are provided. The apparatus and methods may include using a biometric sensor on an Internet-of-things (“IoT”) device to measure a biometric attribute of a user. The measured biometric attribute may be transmitted to a server to compare to a stored biometric attribute and authenticate the user. The IoT device may continue to measure the biometric attribute of the user at a pre-determined interval. Each measurement may be used to authenticate the user. This continuous authentication process may take place without the user's conscious interaction with the IoT device. The apparatus and methods may include using the continuous authentication to process and approve a transaction the user requests.

Abstract: A method of determining the security condition of a network includes executing an agent program on one or more computerized devices coupled to the network. Each executing agent program executes one or more security tests and reports the results of such tests to a network assessment engine, and the network assessment engine determines an authoritative security test score and a configurable security test score for the network based on a weighted combination of the security test results.

Abstract: A broadcast receiving device is provided with: a broadcast receiving unit which receives a digital broadcast signal including broadcast program video and application-related information; a storage unit which stores the received broadcast program video and application-related information; a video decoding unit which decodes the video; an application acquisition unit which acquires an application on the basis of location information included in the application-related information; an application execution unit which executes the acquired application; an output unit which is able to output the video; and a control unit for reproducing and decoding the broadcast program video from the storage unit, reproducing the application-related information from the storage unit, acquiring the application on the basis of the location information included in the reproduced application-related information, and executing the acquired application.

Abstract: A contextual navigational control for digital television is described. An apparatus, comprises a contextual navigation control interface (CNCI). The CNCI includes a first area that represents a plurality of cable television programs having a first level of relevance. A second area represents a first group of the plurality of programs having a second level of relevance. A third area represents a first subgroup of the first group having a third level of relevance. A fourth area represents a second subgroup of the first group having a fourth level of relevance. There may be additional areas representing additional sub-groups of relevance to the current viewing context.

Abstract: Methods, apparatus, systems and articles of manufacture are disclosed to use station identification to enable confirmation of exposure to live media. Example apparatus disclosed herein include a watermark data segmenter to determine a media identifier conveyed by a watermark, a detection time determiner to determine a time at which a media meter detected the watermark from a media presentation by a media device, a reference identifier to query a library of reference signatures for a reference signature associated with the media identifier of the watermark and associated with a reference timestamp corresponding to the time at which the media meter detected the watermark, and a viewing type determiner to determine whether the media presentation is live or time-shifted based on whether a result of the query indicates the reference signature does not exist in the library of reference signatures.

Abstract: An illustrative computing system for a dynamic user access control management system classifies users and data resources according to their risk and importance by a user management engine with artificial intelligence, machine learning characteristics. The dynamic user access control management system analyzes the log files of data resources to measure system performance characteristics and user access behavior. This system monitors the device and network by which a data access request to a data resource is made. The dynamic user access control management system validates the leave status of a user initiating a data access request. The dynamic user access control management system automatically determines a user access level for a data resource through intelligent analysis of collected information and defers to a user's manager for an access level determination when the determination to grant an access level is outside of the knowledge base of the user management engine.

Abstract: A method includes receiving a data stream comprising content generated by an application executing on a user device. The data stream is received from a guidance service that is separate from the application. The data stream is processed using a set of machine learning models to identify a first set of artifacts within the content. A first state of the application is identified based on the first set of artifacts. First transition data is identified in a logic flow of the application. The first transition data corresponds to transitioning from the first state to a second state of the application. Based on the first transition data, first guidance data is generated that describes user input for transitioning the application from the first state to the second state. The first guidance data is sent to the user device, where it is separately presented from the application by the guidance service.

Abstract: A recommender system can include a defender computing device that is configured to obtain customer interaction data characterizing customer interactions with an ecommerce marketplace. The defender computing device can also be configured to determine an item recommendation based on the customer interaction data using a trained differentially private recommendation model and send the item recommendation to the customer. The trained differentially private recommendation model is more likely to determine the same item recommendation after poisoned data is injected into the customer interaction data than a recommendation model that is not privately trained.

Abstract: A transmitting method according to one aspect of the present disclosure includes transmitting video control information for reproduction of video data having an individually reproducible unit, transmitting the video data, determining a transmission order of audio control information for reproduction of audio data in a reproduction interval corresponding to a reproduction interval of the video data in such a way that the audio control information is transmitted after the video control information, and transmitting the audio control information according to the determined transmission order, and transmitting the audio data. Thus, in the receiving apparatus, a delay time from reception of data to reproduction of video and audio is shortened.

Abstract: A resource database which stores structured data describing resources from a diverse array of origins (e.g., an application or cloud environment) is built and maintained to support querying, policy enforcement, and remediation of resources from any origin. Structured data representing resources are obtained from any origin for insertion and categorized based on their type and/or origin. Resources within a category have a shared set of potential object paths as defined by the hierarchical tree structure of their structured data. Resources may be correlated across categories based on having values at different object paths in common. Queries and rules/policies can thus reference resources of any category and also resources across different categories based on correlations between the resources, thereby extending rule/policy enforcement and incident remediation across multiple different origins of resources.

Abstract: Systems and methods are provided for user interface deployment that include a server with a cloud application and a client device with a client application. The cloud application is a fully functional application, such as a headless application, and transmits scene graph data including presentation data and behavior data for a first set of user interface elements viewable at the client device in a first operational state. The client application locally renders the first set of user interface elements on a client device display to locally render the portion of the scene graph related to the first operational state. In response to a user input, the client application transmits event or state data to the cloud application and may perform an action associated with the user input while awaiting receipt of updated scene graph data, possibly related to a subsequent operational state, reducing or eliminating the perception of latency.

Abstract: The present disclosure provides an intelligent commentary generation method. The method includes: obtaining a match data stream; parsing the match data stream, to obtain candidate events from the match data stream; determining events from the candidate events, to generate a sequence of events; and generating commentary scripts corresponding to the match data stream according to the sequence of events.

Abstract: A system for protecting public cloud-hosted virtual resources features cloud visibility logic. According to one embodiment, the cloud visibility logic includes credential evaluation logic, data collection logic, correlation logic, and reporting logic. The credential evaluation logic is configured to gain authorized access to a cloud account within a first public cloud network. The data collection logic is configured to retrieve account data from the cloud account, while the correlation logic is configured to conduct analytics on the account data to determine whether the cloud account is subject to a cybersecurity threat or misconfiguration. The reporting logic is configured to generate an alert when the cloud account is determined by the correlation logic to be subject to the cybersecurity threat or misconfiguration.

Abstract: A system and method are provided implementing user searching. The method includes receiving information from a potential user by an onboarding system including one or more attributes for attaching the user to one or more communities with similar attributes, and receiving one or more search terms from the user by a search system. A processor then performs a search based upon the received search terms and the one or more communities to which the user is attached, and a response to the search is displayed to the user on a display. A user preference to the presented response is confirmed by a preference system, and one or more communities to which the user is attached are dynamically updated, along with updating the attributes of the communities to which the user is attached in response to the user preference.