Abstract: In one embodiment, a processor of a computing device receives a query. The computing device may compare a centroid of each of a plurality of clusters to the query such that a subset of the plurality of clusters is selected, each of the plurality of clusters having a set of data points. An assignment of the subset of the plurality of clusters may be communicated to a hardware accelerator of the computing device. A plurality of threads of the hardware accelerator of the computing device may generate one or more distance tables that store results of intermediate computations corresponding to the query and the subset of the plurality of clusters. The distance tables may be stored in shared memory of the hardware accelerator. A plurality of threads of the hardware accelerator may determine a plurality of data points using the distance tables. The processor may provide query results pertaining to at least a portion of the plurality of data points.

Abstract: A database management engine provides a user interface that allows users to access and modify employee information in a database. The database includes entries for employees, and each database entry includes identifying information about the associated employee. A user can request to modify data within database entries, for instance in order to update information associated with an employee. Responsive to the request, the database management engine identifies liabilities associated with the database modification stemming from associated tax laws. Based on the identified tax liabilities, the engine computes the aggregate tax liability owed by the employer and/or employee. Before modifying a database entry, the engine modifies the user interface to include interface elements detailing the computed aggregate tax liability. The user explicitly can be required to confirm the database modification in view of the aggregate tax liability.

Abstract: A server system obtains a collection of images, each image in the collection of images being associated with a first set of text descriptors. The server system obtains a media item being associated with a second set of text descriptors. The server system selects a subset of the collection of images, including: selecting an initial subset of the collection of images, wherein the initial subset of the collection of images consists of images that share a text descriptor with the media item; obtaining a set of preferences for a user of the media-providing service; and selecting the subset of the collection of images from the initial subset of the collection of images based on the set of preferences for the user of the media-providing service. The server system concurrently presents: a respective image of the subset of the collection of images; and the media item.

Abstract: Provided are a method and device for video search, a terminal and a storage medium. The method includes: receiving a first event generated by triggering a first control in a video playback page; acquiring, in response to the first event, a current video image frame played in the video playback page when the first event is triggered; acquiring a first to-be-searched target positioned by a second control in the current video image frame and a first display position of the first to-be-searched target in the current video image frame, and displaying the second control on the first display position; and acquiring a first recommendation result corresponding to the first to-be-searched target, and displaying the first recommendation result in a search result page.

Abstract: Embodiments of the present disclosure relate to methods, devices, and computer readable medium for restoring a file in a virtual machine disk. The method comprises: receiving, from a client, a user's request regarding restoring a file in a virtual machine disk. The method further comprises: determining, based on the request and from a backup disk of the virtual machine disk, files accessible to the user. In addition, the method further comprises providing the client with information related to the files accessible to the user.

Abstract: Low-dimensional feature values with which semantic factors of content are ascertained are generated from relevance between sets of two types of content. Based on a relation indicator indicating a pair of groups indicating which groups are related to first types of content groups among second types of content groups, an initial feature value extracting unit 11 extracts initial feature values of the first type of content and the second type of content. A content pair selecting unit 12 selects a content pair by selecting one first type of content and one second type of content from each pair of groups indicated by the relation indicator. A feature value conversion function generating unit 13 generates feature conversion functions 31 of converting the initial feature values into low-dimensional feature values based on the content pair selected from each pair of groups.

Abstract: The present invention is a system and user interface for a persistent notification that sits within a notification or control center of an operating system (OS) and looks native to the device. The persistent notification provides access to frequently used applications, utility applications, a news feed as well as a search function by sending requests and receiving data from a platform that provides the persistent notification with data related to the device being used by the user to allow the persistent notification to appear native to the device.

Abstract: The present invention relates to dimensionality reduction for machine learning (ML) models. Herein are techniques that individually rank features and combine features based on their rank to achieve an optimal combination of features that may accelerate training and/or inferencing, prevent overfitting, and/or provide insights into somewhat mysterious datasets. In an embodiment, a computer ranks features of datasets of a training corpus. For each dataset and for each landmark percentage, a target ML model is configured to receive only a highest ranking landmark percentage of features, and a landmark accuracy achieved by training the ML model with the dataset is measured. Based on the landmark accuracies and meta-features values of the dataset, a respective training tuple is generated for each dataset. Based on all of the training tuples, a regressor is trained to predict an optimal amount of features for training the target ML model.

Abstract: Methods, non-transitory computer readable media, and computing devices for three-dimensional searching that precisely targets retrieval within diverse types of content includes retrieving web content in response to a received request from a client device. A destination page, an address within the page, and an offset distance from a starting point at the address to a correlated responsive portion of the web content to the received request is identified. The responsive web content with the identified destination page, the address, and the offset distance is provided to the requesting client device.

Abstract: Methods are presented for providing dynamic search filter suggestions that are updated and ranked based on the user filter selections. One method includes detecting a query received in a user interface (UI), calculating, by a search-candidate model, first search results, and calculating, by a suggestions model, first filter suggestions for filter categories to filter responses to the query. The suggestions model is obtained by training a machine-learning algorithm utilizing pairwise learning-to-rank modeling. The first search results and the first filter suggestions are presented in the UI. When a selection in the UI of a filter suggestion is detected, the search-candidate model calculates second search results for the filter categories based on the query and the selected filter suggestion, and the suggestions model calculates second first filter suggestions based on the query and the selected filter suggestion. The second search results and the second filter suggestions are presented in the UI.

Abstract: Garbage collection (GC) states are stored within references stored on a heap memory to track a progress of GC operations with respect to the references. GC state may be stored in a non-addressable portion of references. Based on the GC state of a particular reference, a set of GC operations are selected and performed for the reference. However, references stored on a call stack do not include any indication of GC state. Hence, loading a reference from heap to call stack involves removing the indication of GC state. Writing a reference to heap involves adding the indication of GC state. References embedded within a compiled method also do not indicate any GC state. Metadata of the compiled method indicate a GC state, which is implicated to the embedded references. GC operations are selected and performed for each embedded reference based on the GC state of the compiled method.

Abstract: The present teaching relates to a method, system, and programming for providing content. A plurality of content items and publication information related thereto are obtained. For each of the plurality of content items, one or more topics are determined in accordance with a model. The related publication information associated with each content item is analyzed to identify at least one source of a plurality of sources that published the content item. A distribution is generated of each of the plurality of content items with respect to the plurality of sources and the one or more topics of the content item, and a bias of a source with respect to publishing content is identified based on the distributions of the plurality of content items.

Abstract: A computer system is provided that is programmed to select feature sets from a large number of features. Features for a set are selected based on metagradient information returned from a machine learning process that has been performed on an earlier selected feature set. The process can iterate until a selected feature set converges or otherwise meets or exceeds a given threshold.

Abstract: Methods, systems and computer program products are provided for spoiler prevention. Media consumption applications may be placed in “spoiler-free” mode, for example, to prevent media content from spoiling first-hand user experience. A user may provide and/or authorize access to and use of spoiler prevention information. A user may request media content (e.g., while surfing the Internet). Digital media content to be presented to a user may be searched in real-time and/or pre-searched for spoiler content and/or associated spoiler indications relative to spoiler prevention information. Identified spoiler content may be concealed from users. A procedure may be provided for users to determine one or more reasons why content is concealed, to selectively reveal concealed content, and to provide feedback whether concealed content was or was not spoiler content for a user. Feedback may be used to improve spoiler prevention, for example, by retraining a machine learning model, which may be user-specific.

Abstract: Systems and methods for querying a database are disclosed. A server computing device includes a processor and a memory storing instructions that, when executed by the processor, cause the server computing device to: receive a query input from a client device over a network connection; establish a non-blocking socket between the client computing device and the server computing device; generate a database query based on the query input; cause the database query to begin executing against one or more databases; perform a read request on the non-blocking socket; receive a code in response to the read request on the non-blocking socket; determine whether the network connection exists between the client device and the server device based on the received code; and cause the database query to be cancelled when the server determines that the network connection does not exist between the client device and the server device.

Abstract: Systems and methods for training a machine learning (ML) ranking model to rank genealogy hints are described herein. One method includes retrieving a plurality of genealogy hints for a target person, where each of the plurality of genealogy hints corresponds to a genealogy item and has a hint type of a plurality of hint types. The method includes generating, for each of the plurality of genealogy hints, a feature vector having a plurality of feature values, the feature vector being included in a plurality of feature vectors. The method includes extending each of the plurality of feature vectors by at least one additional feature value based on the number of features of one or more other hint types of the plurality of hint types. The method includes training the ML ranking model using the extended plurality of feature vectors and user-provided labels.

Abstract: A learning automaton can be trained to merge data from input data streams, optionally with different data rates, into a single output data stream. The learning automaton can learn over time from the input data streams. The input data streams can be low-pass filtered to suppress data having frequencies greater than a time-varying cutoff frequency. Initially, the cutoff frequency can be relatively low, so that the effective data rates of the input data streams are all equal. This can ensure that initially, high data-rate data does not overwhelm low data-rate data. As the learning automaton learns, an entropy of the learning automaton changes more slowly, and the cutoff frequency is increased over time. When the entropy of the learning automaton has stabilized, the training is completed, and the cutoff frequency can be large enough to pass all the input data streams, unfiltered, to the learning automaton.

Abstract: An automatic multi-threshold feature filtering method and an apparatus thereof are provided. In an iterative process of training a machine learning model, the feature filtering method calculates a feature filtering threshold and feature correlation values of a current round of iteration based on a result of a previous iteration, and performs feature filtering on samples based on the calculated feature filtering threshold and the calculated feature correlation values. The feature filtering apparatus of the present disclosure includes a calculation module and a feature filtering module. The method and apparatus of the present disclosure can automatically generate different feature filtering thresholds at each iteration, which greatly improves an accuracy of a filtering threshold, and can greatly increase the training speed of automatic machine learning and an accuracy of a machine learning model compared with fixed and single thresholds nowadays.

Abstract: Systems and methods for responding to a subscriber's text-based request for content items are presented. In response to a request from a subscriber, word pieces are generated from the text-based terms of the request. A request embedding vector of the word pieces is obtained from a trained machine learning model. Using the request embedding vector, a set of content items, from a corpus of content items, is identified. At least some content items of the set of content items are returned to the subscriber in response to the text-based request for content items.

Abstract: Embodiments are directed to managing event information. A plurality of events associated with entities may be provided. A plurality of state types may be determined based on the plurality of events such that each state type is associated with a state. State keys associated with each entity may be determined based on events associated with each entity and the state types. A state stream for each entity may be provided based on their state keys such that each state stream may be an ordered sequence of the keys associated with each entity. In response to a query that includes a pattern filter, the pattern filter may be employed to determine a portion of the entities based on the state stream for each entity such that the pattern filter matches the state stream for each of the portion of the entities.