Abstract: Implementations are described herein for leveraging digital media files retrieved and/or created by users to predict/determine topics of potential relevance to the users. In various implementations, digital media file(s) created and/or retrieved by a user with a client device may be applied as input across trained machine learning model(s), which in some cases are local to the client device, to generate output that indicates object(s) detected in the digital media file(s). Data indicative of the indicated object(s) may be provided to a remote computing system without providing the digital media file(s) themselves. In some implementations, information associated with the indicated object(s) may be retrieved and proactively output to the user. In some implementations, a frequency at which objects occur across a corpus of digital media files may be considered when determining a likelihood that a detected object is potentially relevant to a user.

Abstract: A process for tagging incident data captured by non-public-safety agency controlled cameras. In operation, an electronic computing device detects an occurrence of an incident and obtains an incident identifier associated with the incident. The electronic computing device determines that a non-public-safety agency controlled camera is available for capturing incident data corresponding to the incident. The electronic computing device then controls a public-safety agency controlled internet-of-things (IoT) infrastructure deployed at an incident location to emit a signal including one of an audible signal, visible signal, or a combination of audible and visible signals to enable the non-public-safety agency controlled camera to tag incident data using the emitted signal. The electronic computing device links a unique tag representing the emitted signal to the incident identifier.

Abstract: A time-series data processing apparatus according to the present invention includes: a generating unit configured to generate a generator having learned so as to generate state information representing a state of time-series data having a predetermined time width in accordance with a label given to the time-series data; a state information generating unit configured to generate, by using the generator, state information representing a state of division time-series data obtained by dividing the time-series data by a shorter time width than the predetermined time width; and a classifying unit configured to classify a plurality of division time-series data based on state information of the plurality of division time-series data.

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: Media unit retrieval methods, systems and computer program products are provided that allow a user to search for an item by iteratively presenting media units such as images representing items to the user and receiving user input consisting of selections of the presented media units (including possibly the empty selection). Features, or attributes, a user is interested in, for example semantic features, are inferred from the interaction and media units are retrieved for presentation based on similarity with user-selected media units, through sampling of a probability distribution describing the intent or interests, or combinations of approaches. Accordingly, the user-experience is akin to a conversation about what the user is looking for. Retrieval may be based on both selected and unselected media units and the selection may comprise making a selection with a single action. Further, a database of media units can capture similarity relationships for efficient media unit retrieval.

Abstract: Media unit retrieval methods, systems and computer program products are provided that allow a user to search for an item by iteratively presenting media units such as images representing items to the user and receiving user input consisting of selections of the presented media units (including possibly the empty selection). Features, or attributes, a user is interested in, for example semantic features, are inferred from the interaction and media units are retrieved for presentation based on similarity with user-selected media units, through sampling of a probability distribution describing the intent or interests, or combinations of approaches. Accordingly, the user-experience is akin to a conversation about what the user is looking for. Retrieval may be based on both selected and unselected media units and the selection may comprise making a selection with a single action. Further, a database of media units can capture similarity relationships for efficient media unit retrieval.

Abstract: Systems, computer program products, and methods are described herein for intelligent database modelling. The present invention is configured to capture, using the augmented reality application, a drawing made by the user on a medium; initiate a first machine learning model on the drawing; determine that the user is attempting to design a database model; extract at least a first entity from the drawing; determine, using the first machine learning model, one or more attributes for the first entity; initiate, via the augmented reality application, a first push notification for display on the computing device of the user; electronically receive, via the augmented reality application, a user selection of a subset of attributes from the one or more attributes; generate, using the augmented reality application, a conceptual database model based on at least the first entity and the subset of attributes.

Abstract: Systems and methods ingest machine data including logs, metadata, and cost and usage information from multiple heterogeneous cloud services. The machine data is saved as events. An application retrieves the metadata, events, metrics, and logs and causes an easy to understand visual representation of costs, resource usage, and non-compliance for each of a client's cloud services. Further, the data across the client's multiple heterogeneous cloud services is normalized to provide visual representations that compare the costs, resource usage, and non-compliance across the client's multiple heterogeneous cloud services. Further, machine learning aspects of the application can provide recommendations and trend analysis for cloud service asset usage.

Abstract: Methods, systems, and devices for sharing a physical location with online users by requiring the online users to be in physical proximity of a specific physical location before the users can experience media content associated with the physical location. A user can mark a physical location when that user is at the location by associating a virtual anchor with that location. The virtual anchor creates a data structure that contains the location's geographic coordinates as well as other details regarding the anchor. The user can also associate media content with the virtual anchor. The virtual anchor can be shared with other users and these other users are notified of the presence of the virtual anchor when they are within a physical proximity from the physical location. Only when the other users are at the location can these users access the media content associated with the virtual anchor.

Abstract: Examples include a blockchain system associated with a distributed file system, Some examples include retrieval, from the blockchain system, a latest block corresponding to a block signature of a reference block that corresponds to a file stored in the distributed file system. The latest block includes information identifying at least one of a latest location of the file, a signature of metadata of the file, or a tag associated with the file.

Abstract: Techniques for implementing artificial crowd intelligence via networking of recommendation engines are disclosed. In some example embodiments, a base recommendation engine performs a computer-implemented method comprising: receiving a recommendation request; generating a first plurality of recommendations based on the recommendation request using a base recommendation model; obtaining one or more other recommendations from one or more other recommendation engines, each one of the other recommendation engines being configured to generate the other recommendations based on the recommendation request using a corresponding recommendation model; obtaining a corresponding weight for the other recommendation engine(s); selecting at least one recommendation from the first plurality of recommendations and the other recommendation(s) using the corresponding weight(s); and causing the selected recommendation(s) to be displayed.

Abstract: A computer may receive a request to generate a snapshot view of a virtual infrastructure. The virtual infrastructure may comprise a plurality of virtual server management applications, each managing a respective set of virtual machines. The computer may implement a multi-threaded process to contemporaneously query one or more databases and retrieve status and other information of the virtual machines from different virtual server management applications. The computer may aggregate the retrieved information to determine the summary counters and statistic information for the virtual machines. The computer may generate a snapshot view file based on the retrieved information. The snapshot view file may be in hypertext markup language (HTML) format. The computer may transmit a selectable link to the snapshot view file to multiple user devices. A user may select the link and the respective user device may display the snapshot view in an application such as a web browser.

Abstract: A user interface displays: a first column comprising non-editable input strings retrieved from a data field; a second column comprising editable output strings initialized from the data field; and an expression window displaying a transformation function ƒ. The computer iteratively processes user inputs, each user input i providing a sample row transformation to edit an ith output string ti. Some user inputs i designate a contiguous substring ssi of the corresponding input string si. The contiguous substring expresses a causal basis for transforming the input string si. into the output string ti. The computer updates the transformation function ƒ according to the provided sample row transformations so that: ƒ(s1)=t1, . . . , ƒ(si)=ti; the transformation function ƒ specifies text or string position of at least one contiguous substring; and ƒ has minimal branching among possible transformation functions that satisfy the samples.

Abstract: Techniques are disclosed for generating a database schema using trained machine learning models that, in some embodiments, may include graph neural networks (GNN). A GNN may identify source to target database schema mappings using, among other features of the graph, context data associated with each node in a graph. Context data describes relationships between a particular node and some (or all) of the other nodes in the graph. The system may use this context data (and other graph data) in combination with a trained GNN model to identify a mapping between one or more source database entities to corresponding target database entities.

Abstract: Systems and techniques are generally described for attribute-based content selection and search. In some examples, a graphical user interface (GUI) may display an image of a first product comprising a plurality of visual attributes. In some further examples, the GUI may display at least a first control button with data identifying a first visual attribute of the plurality of visual attributes. In some cases, a first selection of the first control button may be received. In some examples, a first plurality of products may be determined based at least in part on the first selection of the first control button. The first plurality of products may be determined based on a visual similarity to the first product, and a visual dissimilarity to the first product with respect to the first visual attribute. In some examples, the first plurality of products may be displayed on the GUI.

Abstract: A recipe construction system, a recipe construction method, a computer readable recording media with stored programs, and a non-transitory computer program product are provided. A dimension reduction module obtains dimension-reduced composition information according to composition information of each piece of historical recipe information and a dimension reduction algorithm. A neural network module obtains a plurality of trained neural network parameters according to the composition information and the dimension reduction composition information. The neural network module obtains dimension-reduced initial composition information according to the trained neural network parameters and initial composition information. A search module searches the historical recipe information for a first quantity of pieces of candidate recipe information.

Abstract: Provided is a sensor network capable of acquiring, from a data center, learning data composed of a pair of sensor data (input) and a classification label (output) necessary for adding/changing a classification label for updating a classifier while reducing the volume of communication between a sensor and the center. One aspect of the present invention relates to a sensor network, including one or more sensor nodes and a data center. The sensor node includes: an encoding unit for encoding sensor data by an encoder part of an autoencoder; and a transmission unit for transmitting the encoded data. The data center includes: a reception unit for receiving data encoded from sensor data by the encoder part of the autoencoder; a decoding unit for decoding the encoded data by a decoder part of the autoencoder; and a storage unit for storing the decoded data therein.

Abstract: A unified graph query system provides an abstraction layer that increases the interoperability of different graph technologies by exposing graphs stored in graph databases using a unified query language. The abstraction layer generates graph models for each of the available graph databases and extracts a graph component and other source data used to identify the source of the data requested by a query. The unified graph query system executes the query across the multiple graphs included in different graph databases by using the graph models to locate the graph component in each of the multiple graphs and extract the feature data associated with the graph component. The feature data is used to generate features that are used by a machine learning service to train machine learning models and is also used to make predictions in real time.

Abstract: The invention relates to database abstraction and data linkage. According to an embodiment of the present invention, the invention takes a variety of attributes (e.g., names, IP address, device identifiers, addresses, phone numbers, account numbers, etc.) and returns the online activity, demographic data, account data and/or other activity, events and data associated with that attribute. The tool may then iterate over each attribute and return a network of connections having multiple degrees of association. The innovative tool may be linked to known bad actor data, and perform automated searches on this data to proactively alert potentially fraudulent activity. The tool may also be developed to add attributes and apply machine learning to the associations to more intelligently describe the returned network. Further, the tool may be developed to describe larger networks having multiple degrees of connections.

Abstract: An improved analytics system dynamically generates, integrates, and deploys multiple models for use related to a particular user and/or user segment. The integrated model analytics system can deploy a set of models specifically identified for a user from a group of models related to the service. This set of models can be used to provide recommendations for the user related to content of the service. As the user continues to interact with the service, the integrated model analytics system can dynamically update multiple models from the set of models based on real-time user interactions. These updated models can be used to generate updated recommendations for the user that can be used to dynamically update content presented to the user. In addition, this updated model can be used when providing recommendations for users in other user segments.