Abstract: According to an aspect, a method for ranking messages of conversation graphs in a messaging platform using machine-learning signals includes receiving a conversation view request to retrieve messages of a conversation graph stored on a messaging platform and determining a plurality of predictive outcomes for each of a plurality of messages of the conversation graph using predictive models. The determining step may include obtaining a plurality of signals relevant to the predictive models, the plurality of signals including data structure-related signals relating to the conversation graph and inputting the plurality of signals to the predictive models. The method includes ranking the plurality of messages based on the predictive outcomes and transmitting at least a subset of the plurality of messages to be rendered on a client application according to the rank.

Abstract: According to an aspect, a method for identifying relevant messages in a conversation graph on a messaging platform includes transmitting, over a network, messages between a plurality of users on the messaging platform, generating a conversation graph based on relationships between the messages, where the conversation graph includes a plurality of messages related to a conversation, and the plurality of messages of the conversation graph include a root message and one or more reply messages connected to the root message. The method includes marking a subset of the plurality of messages of the conversation graph as relevant to the conversation, including marking a message from a user account having a credibility rating over a threshold level, and transmitting, over the network, digital information to render the subset of the plurality of messages on a user interface of a computing device.

Abstract: A system and method for user content presentation, including: receiving, in association with a request to establish a graph relationship between a first account and a second account, a strength of relationship from the first account to the second account, the strength of relationship selected based on input from a user of the first account; selecting, by a computer processor, a delivery attribute associated with the second account based on the strength of relationship, where the delivery attribute is positively correlated with a measure of content authored by the second account to be provided to the first account; selecting content authored by the second account for inclusion in a stream of the first account based on the delivery attribute; and providing the stream of the first account for display on a client computing device.

Abstract: Disclosed is a system for delegating authentication of an untrusted application executing on a client device. For delegated authentication, an untrusted application relies on a trusted application executing in the same environment for authentication purposes. The delegated authentication process avoids requiring the user of the untrusted application to provide authentication credentials. The disclosed system for delegating authentication enables any trusted application executing in the same computing environment to authenticate the untrusted application.

Abstract: A neural network is trained to process received visual data to estimate a high-resolution version of the visual data using a training dataset and reference dataset. A set of training data is generated and a generator convolutional neural network parameterized by first weights and biases is trained by comparing characteristics of the training data to characteristics of the reference dataset. The first network is trained to generate super-resolved image data from low-resolution image data and the training includes modifying first weights and biases to optimize processed visual data based on the comparison between the characteristics of the training data and the characteristics of the reference dataset.

Abstract: A system and method for trend detection in a messaging platform. A trend detection model is selected and a time series having a plurality of instances of social data is received, wherein the instances of social data share a countable parameter. A count is made of occurrences of countable parameters in each instance of social data assigned to that bin and a trend detected based at least in part on the trend detection model and on the count for each bin.

Abstract: Methods, computer readable media, and apparatuses for providing enhanced user interface mechanics are presented. In one arrangement, a scrollable list of content items may be displayed. Input associated with a scroll command may be received. Then, based on the scroll command, a scrollable refresh trigger may be displayed. Subsequently, the scrollable list of content items may be refreshed in response to determining, based on the scroll command, that the scrollable refresh trigger has been activated. In at least one instance, it may be determined that the scrollable refresh trigger has been activated in response to determining that the scroll command was completed while the scrollable refresh trigger was fully displayed.

Abstract: Example systems and methods are described for implementing a swipe-to-like feature. In an example implementation, a list of content items is displayed on a touchscreen display, and based on detecting input of a first gesture, such as, for example, a swipe gesture, for a first one of the content items in the list, associating a predetermined first sentiment with the first content item.

Abstract: A system for online conversion attribution. The system includes a short uniform resource locator (URL) service programmed to, in response to receiving a short URL from a device: provide a cookie that includes a short URL ID to the device, and provide a short URL descriptor including the short URL ID to a user mapping service. The system further includes the user mapping service programmed to receive the short URL descriptor, receive a social media descriptor including a social media ID, map the short URL ID to the social media ID using the short URL descriptor and the social media descriptor, and attribute, using the mapping, a conversion on a website accessed using the device based upon receipt of the short URL ID from the device.

Abstract: Disclosed is a system for recommending content of a predefined category to an account holder, detecting spam applications, or account holders based on the account holder application graphs. The system receives information corresponding to applications executing on the client device of the account holders and generates an application graph for each account holder that includes a list of predefined application categories that are preferred by the account holder. For each predefined category, a list of account holders preferring content relevant to that category is predicted based on the set of generated application graphs. Some application graphs may be detected as spam application graphs by comparing the generated application graphs with a set of predefined spam application graphs. Alternatively, if the generated application graph does not match the predefined spam application graphs, they are compared to a set of application graphs from a database to find similar application graphs.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for information retrieval. One of the methods includes receiving a search query from a user device at a messaging system; determining one or more relevant messages responsive to the search query, the determining comprising: processing the search query using a first stage of a ranking system, the first stage using a first set of features to identify a first set of relevant message, processing the search query using a second stage of the ranking system, the second stage using a second set of features and the first set of relevant messages to determine a second set of relevant messages, and ranking the second set of relevant messages according to respective scores; and providing one or more results to the user device for display as search results.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for providing to user devices identifiers of edge servers from which to obtain content. One example system includes a traffic control service. The traffic control service is configured to receive polling requests from user devices. In response to a particular polling request, the traffic control service obtains aggregated data about the user devices and edge servers. The traffic control service selects, based on the aggregated data, a best edge server for the user device that sent the particular polling request. The traffic control service sends an identifier of the selected best edge server to the user device.

Abstract: A system and method for feature switching in software applications. A feature switching action (FSA) is received and at least one user device is identified based on the received FSA. At least one feature switching instruction (FSI) is generated based on the received FSA and the FSI is sent to at least one of the identified user devices. A feature switching instruction response (FSIR) is received from each user device that received a FSI and a feature switching action response (FSAR) is generated based on the received FSIRs.

Abstract: A real-time messaging platform and method are disclosed which suggests messages and accounts from the real-time messaging platform.

Abstract: In general, embodiments of the invention relate to servicing fill request using ad networks and demand-side partners.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for data retention and modification. One of the methods includes dividing partitions into a set of generations according to a retention policy; accumulating modification and deletion events that define changes to be applied to data of the distributed dataset; and when a triggering event occurs for a triggered generation in the set of generations, rolling an oldest partition out of the triggered generation, the rolling comprising: if the oldest partition has reached the end of a retention period for the dataset, marking the oldest partition for deletion in the triggered generation; otherwise: creating a new partition corresponding to the data of the oldest partition, wherein the data is cleaned using a scrubbing process; adding the new partition to a next generation in the set of generations; and marking the oldest partition for deletion in the triggered generation.

Abstract: A system and method for event-based content presentation, including: receiving a content presentation request to present a content item at an event where presentation of the content item can be captured. The content presentation request can include an event identifier identifying the event at which the content item is to be presented, and can further identify the content item to be displayed. The submitted content item can be presented on a content presentation device located at the event. Presentation of the content item can be captured and the resulting captured content presentation can be stored, broadcasted on a messaging platform, provided to the requesting user, etc.