Abstract: The disclosed systems and methods provide a framework for a proactive prediction of the toxic propensity of an article. Prior to the publication and/or reception of comments to online content, the disclosed framework determines the toxic propensity of the content's context and/or specific words, sentences, sentiments, tone or other messages receivable from consumption of the content. Thus, disclosed framework performs proactive forecasting of the content's toxicity propensity”, which quantifies how likely the content is prone to incur or attract toxic comments. The framework can function and/or be configured to operate in a manner that can perform specifically adherent moderation actions that correspond to the content and control how the content can be interacted with, based on the toxic propensity determination, prior to the content's publication in an effort to thwart, prevent or stop toxic environments surrounding or stemming from the content from coming into existence.

Abstract: One or more computing devices, systems, and/or methods for content recommendations using historical future data are provided. A model serving delay time is computed as an average of training delays of events. A historical data time interval is determined based upon the model serving delay time. A model is trained for predicting user content preferences using historic user distribution data and historic content distribution data associated with the historic data time interval. The model is utilized to generate and provide content recommendations to users.

Abstract: The present teaching relates to method, system, medium, and implementations for machine learning. Machine learning is performed based on training data via a dual loop learning process that includes a first loop for data decoding learning and a second loop for label decoding learning. In the first loop, first parameters associated with decoding are updated to generate updated first parameters based on a first label, estimated via the decoding using the first parameters, and a second label, predicted via the label decoding using second parameters. In the second loop, the second parameters associated with the label decoding are updated to generate updated second parameters based on a third label, obtained via the decoding using the updated first parameters, and a ground truth label.

Abstract: Techniques for electronic document content bookmarking are disclosed. In one embodiment, a method is disclosed comprising receiving user input indicating selection of a portion of the electronic document, receiving a request to bookmark the selected portion of the electronic document, generating a bookmark card for the selected portion of the electronic document, the bookmark card comprising information identifying the electronic document, the selected portion of the electronic document as bookmarked content and the at least one user-defined information item, receiving, via the computing device, a bookmark view request via the GUI, and causing display of the bookmark view via the GUI in response to the bookmark view request, the bookmark view comprising an entry corresponding to the generated bookmark card and providing the bookmarked content from the generated bookmark card, the bookmark view providing a number of actions selectable by the user in connection with the entry from the bookmark view.

Abstract: Disclosed are systems and methods for improving interactions with and between computers in content generating, searching, hosting and/or providing systems supported by or configured with personal computing devices, servers and/or platforms. The systems interact to identify and retrieve data within or across platforms, which can be used to improve the quality of data used in processing interactions between or among processors in such systems. The disclosed systems and methods provide systems and methods for privacy-aware machine learning using an improved data encoding that withholds more information. The disclosed systems and methods encode a corpus of data and encode each query used in searching and generating query results from the corpus of encoded data.

Abstract: In some aspects, the techniques described herein relate to a method including: receiving a set of messages; filtering the set of messages to identify a set of newsletters in response to a selection of a newsletter control within a messaging application; rendering a newsletter view within the messaging application, the newsletter view displaying a plurality of tiles corresponding to the set of newsletters; receiving a section of a given tile in the plurality of tiles from a user; and rendering a newsletter reader view, the newsletter reader view including a subject and body of a selected newsletter corresponding to the given tile.

Abstract: One or more techniques and/or systems for sending push notifications of content items to client devices are provided herein. For example, an input received from a user can be expanded to obtain an expanded user interest. Content items from a content source can be filtered based upon the expanded user interest to obtain a set of filtered content items. A push notification can be constructed to comprise one or more of the filtered content items from the set of filtered content items. The push notification can be sent to a client device of the user for display as a device alert notification. In an example, the filtered content items, within the push notification, may be ranked based upon a ranking metric.

Abstract: One or more computing devices, systems, and/or methods for evaluating email activity and/or controlling, based upon the email activity, transmission of instructions associated with quality are provided. A first email, transmitted by an email account associated with an entity to a plurality of email accounts, may be identified. First activity associated with the first email may be detected. A first set of activity information associated with the first activity may be stored in an entity profile associated with the entity. The entity profile may comprise a plurality of sets of activity information associated with a plurality of emails transmitted by one or more email accounts associated with the entity. A quality score corresponding to the first entity may be generated based upon the entity profile. A notification may be generated based upon the quality score. The notification may be transmitted to the first client device.

Abstract: The present teaching relates to communication related to predictions. A communication chain is created with multiple chain units, each operating independently to provide prediction services related to a prompt with embedded information therein associated with predictions for the prompt to enable prediction services. A request is received for a requested operation related to a query prompt on one of prediction solicitation, prediction entry, verification entry, and prediction access. If a chain unit for the query prompt, the request is directed to the chain unit to perform the requested operation. If not, a new chain unit is created for the query prompt and the request is forwarded to the new chain unit to carry out the requested operation.

Abstract: Disclosed are systems and methods for improving interactions with and between computers in content providing, searching and/or hosting systems supported by or configured with devices, servers and/or platforms. The disclosed systems and methods provide a novel framework for analyzing messages shared over a network and identifying information in them referencing entities based on a determined context of the messages. Once the entity information is identified, the disclosed framework segregates them based on the context, and then presents them in contextual views as content cards within or as part of generated, content-specific interfaces. The disclosed framework enables users to continue messaging while pivoting between different views of the generated interfaces, which enables the retrieval of additional information related to the entity information included on the content cards.

Abstract: One or more computing devices, systems, and/or methods for content item selection for goal achievement are provided herein. A goal of a user is identified. A model is utilized to evaluate the goal, user information, and a set of content items to generate predictions for the content items of how likely each content item will be a causation factor of the user making progress towards the goal in response to the user being provided with each content item. A target content item is selected from the set of content items based upon the target content item having a predicted likelihood of being the causation factor above a threshold. The target content item is provided through a registered media channel accessible through a device of the user.

Abstract: One or more computing devices, systems, and/or methods for providing a system with access to resources associated with a user account using a disposable email address (DEA) are provided. A request to provide a system with access to a first set of resources may be received. The first set of resources may be associated with a first email account associated with a first email address. A first DEA associated with the first email account may be generated. The first DEA may be transmitted to the system. A first set of modified emails may be generated based upon a first set of emails of the first email account. A first modified email of the first set of modified emails may comprise an indication of the first DEA. Access to the first set of modified emails may be provided to the system. The first DEA may be deactivated.

Abstract: One or more computing devices, systems, and/or methods for debiasing training data based upon information seeking behaviors are provided. Users associated with a set of training data are segmented into information seeking behavior groups corresponding to varying degrees of information seeking behaviors of the users. Biases for the information seeking behavior groups may be estimated based upon information seeking behaviors of users within the information seeking behavior groups. The training set of data is debiased using the biases to generate a debiased training set of data. A model may be trained to perform a task based upon the debiased training set of data.

Abstract: One or more computing devices, systems, and/or methods are provided. Event information associated with a plurality of events may be identified. The plurality of events may be associated with first entities corresponding to a first entity type and second entities associated with a second entity type. A first network profile associated with the first entities and the second entities may be generated based upon the event information. An arrangement of particles corresponding to the first entities and the second entities may be generated. Charges associated with the particles may be determined based upon the first network profile. The particles may be rearranged to a second arrangement of particles based upon the charges. One or more clusters of particles in the second arrangement of particles may be identified. One or more coalition networks associated with fraudulent activity may be identified based upon the one or more clusters of particles.

Abstract: The present teaching generally relates to detecting providing pre-validated data buckets for online experiments. In a non-limiting embodiment, user activity data representing user activity for a first plurality of user identifiers may be obtained. A first set of values and a second values, representing first and second user engagement parameters, respectively, may be generated for each user identifier based on the user activity data. A first ranking and a second ranking may be determined for the first and second sets, respectively. A first exclusion range including a first number of values to be removed from the first and second sets may be determined. A homogenous value set may be generated by removing the first number of values from the first and second sets, where each value from the homogenous value set corresponds to a user identifier available to be placed in a data bucket for an online experiment.

Abstract: The present teaching relates to method, system, medium, and implementations for machine learning. A bid is received, from an expert during training, for a training sample with an amount within a level of available bidding currency associated with the expert. The training sample is used for training a model associated with the expert. It is determined whether the expert is among at least one winner selected based on bids from one or more experts. If the expert is among the at least one winner, the training sample is sent to the expert. The at least one winner is selected based on one or more criteria aiming at expert diversification.

Abstract: In some aspects, the techniques described herein relate to a method including: receiving a request from a client device to join a room, the room associate with a plurality of other client devices streaming multimedia content; adding the client device to the room in response to the request; computing a synchronization time for the client based on time values associated with the client device and the other client devices; and transmitting the synchronization time to the client device, causing the client device to adjust playback of the multimedia content based on the synchronization time.

Abstract: A method for presenting personalized content to a user includes receiving user data corresponding to a user having a user profile, the user data including at least one or more messages in a user mailbox and a user web browser history within a network, extracting one or more data tags from the received user data, associated at least one data tag with a message, finding information in the network that corresponds to the associated data tag, generating a notification for the user, the notification including the found information in the network, and outputting the generated notification to a user interface of a device of the user.

Abstract: Systems and methods are disclosed for vertical integration of a multi-dimensional machine learning algorithm with a digital content management system. A variety of machine learning algorithms are implemented to aid publishers in the discovery of digital content. In contrast to prior methods, which perform retroactive assessments on digital content performance, the instant methods train machine learning algorithms to predict future performance of digital content items. As digital content becomes available to publishers, machine learning algorithms analyze the digital content items in order to present editorial actions publishers should take via a digital content management dashboard.

Abstract: A method of adaptive online user profiles includes displaying a website for a particular topic of interest based on a user request, receiving a request from the user to display user profile data, obtaining the user profile data, the user profile data including topic-area-specific user-specific profile information for multiple topic areas, comparing a topic area of the displayed website to a topic area for each topic-area-specific user-specific profile information, and displaying the user profile data such that the topic-area-specific user-specific profile information most closely matching the topic area of the displayed website is displayed most prominently.