Abstract: Disclosed RNN-implemented methods and systems for abstractive text summarization process input token embeddings of a document through an encoder that produces encoder hidden states; applies the decoder hidden state to encoder hidden states to produce encoder attention scores for encoder hidden states; generates encoder temporal scores for the encoder hidden states by exponentially normalizing a particular encoder hidden state's encoder attention score over its previous encoder attention scores; generates normalized encoder temporal scores by unity normalizing the temporal scores; produces the intra-temporal encoder attention vector; applies the decoder hidden state to each of previous decoder hidden states to produce decoder attention scores for each of the previous decoder hidden states; generates normalized decoder attention scores for previous decoder hidden states by exponentially normalizing each of the decoder attention scores; identifies previously predicted output tokens; produces the intra-decoder at

Abstract: Disclosed technology for accurate translation of elements in a web application includes systems and methods of efficiently debugging translations for error-free localization, including using a translation debugger; comparing a source language file and a translated language file of string properties and other elements of a UI; and generating in a WYSIWYG panel of the UI, visual coding of the string properties and the other elements. The visual coding distinguishes among characteristics relating to translation states, accepting edits to string properties in a string property edit pane, displaying modified string properties in the WYSIWYG panel, and exporting an ad hoc changes list produced by the edits.

Abstract: We introduce an attentional neural machine translation model for the task of machine translation that accomplishes the longstanding goal of natural language processing to take advantage of the hierarchical structure of language without a priori annotation. The model comprises a recurrent neural network grammar (RNNG) encoder with a novel attentional RNNG decoder and applies policy gradient reinforcement learning to induce unsupervised tree structures on both the source sequence and target sequence. When trained on character-level datasets with no explicit segmentation or parse annotation, the model learns a plausible segmentation and shallow parse, obtaining performance close to an attentional baseline.

Abstract: The technology disclosed relates to a platform for ultra-fast, ad-hoc data exploration and faceted navigation on integrated, heterogeneous data sets. The disclosed method of declarative specification of visualization queries, display formats and bindings represents queries, widgets and bindings combined on a dashboard in real time, with flexible display options for analyzing data and conveying analysis results.

Abstract: To deliver up-to-date, coherent user data to applications upon request, the disclosed technology includes systems and methods for caching data and metadata after it has been synchronously loaded—for future retrieval with a page load time close to zero milliseconds. To provide this experience, data needs to be stored as locally to a user as possible, in the cache on the local device or in an edge cache located geographically nearby, for use in responding to requests. Applications which maintain caches of API results can be notified of their invalidation, and can detect the invalidation, propagate the invalidation to any further client tiers with the appropriate derivative type mapping, and refresh their cached values so that clients need not synchronously make the API requests again—ensuring that the client has access to the most up-to-date copy of data as inexpensively as possible—in terms of bandwidth and latency.

Abstract: Disclosed are methods, systems, and computer program products for providing cross-domain communication via a user interface provided in a web browser. One or more servers of a database system receive an indication of a first event message sent from a second user interface component to a first user interface component of the user interface, the first event message comprising a request to perform a first operation. The one or more servers identify a first one or more trusted data network domains for a first application displayed in the first user interface component. The one or more servers determine that the first one or more trusted data network domains comprise a second data network domain for a third party application displayed in the second user interface component. The one or more servers determine that the first operation is a safe operation and perform the first operation.

Abstract: Disclosed are systems, methods, and computer program products for suggesting actions for evaluating user performance in an enterprise social network. In some implementations, a server of the enterprise social network identifies a message configured to be published to a feed of the enterprise social network. The message may comprise textual input received from a client machine associated with a first user of the enterprise social network. The server determines that the textual input comprises information evaluating a second user associated with the enterprise social network. The server provides data for display at the client machine, the data comprising a prompt to the first user to include the information related to the second user in a performance evaluation record for the second user.

Abstract: Methods and systems are provided for regulating access to a database by a server. One exemplary method involves the server receiving one or more requests for an activity with respect to the database and determining an attempted usage for the activity over a monitoring period by a source of the requests. When the attempted usage exceeds an allowed usage of the activity for the monitoring period, the server creates an entry in a cache that includes an identifier associated with the source of the one or more requests. Thereafter, the requested activity with respect to the database on behalf of the source may be prevented based on the entry in the cache.

Abstract: An application server sends a public key from an asynchronous key-pair to a user system to encrypt a user encryption secret that forms part of a first encryption key. The application server uses a second encryption key provided by a key derivation server to encrypt a private key from the asynchronous key-pair. The application server then deletes the second encryption key to prevent decryption of the user encryption secret received from the user system. The application server receives the encrypted user encryption secret from the user system and sends a request to the key derivation server to re-encrypt the user encryption secret. The key derivation server uses a key encryption secret to generate the second encryption key and decrypt the private key. The key derivation server uses the decrypted private key to decrypt the user encryption secret and then re-encrypts the first encryption secret to prevent decryption by the application server.

Abstract: Systems and methods are provided for classifying support-related messages from users in a support-related group. A method includes receiving a support-related message containing a support-related problem. The received support-related message is classified by using a processor-implemented machine learning model to identify a support-related category. The identified support-related category is provided for user display.

Abstract: Computing systems, database systems, and related methods are provided for suggesting logical relationships between database objects. One exemplary method involves a server identifying a database action with respect to a contact object associated with an event object, and in response to the action, identifying a second contact object in the database likely to be related to the event object based on one or more attributes of the event object. The server provides an indication of the second contact object in association with the event object on a client device and updates the database to maintain an association between the event object and the second contact object in response to an input at the client device.

Abstract: Computing systems, database systems, and related methods are provided for providing visualizations of distributed package components. One exemplary method involves a server providing a graphical representation of a plurality of classes of a package, receiving a request to view a global class of the plurality of classes, obtaining previously compiled code associated with the global class of the from a database, filtering the code to include global source components of the global class while removing non-global source code components, and providing a graphical representation of the filtered code on the client device requesting presentation of the global class.

Abstract: Systems and methods are disclosed for managing transmissions of notifications to user devices. A method can include determining that an event has occurred that requires a user notification. User and device characteristics are analyzed to determine that a single user device is currently in active use. The user notification is transmitted to the user device that has been determined to be in active use.

Abstract: A method for providing query responses to a user via online chat establishes a first communication connection for online chat between a user interface and an artificial intelligence (AI) entity comprising a processor and a memory element configured to store a database of query answers; receives a user input query transmitted via the first communication connection; performs a lookup in the database of query answers, to locate a query answer corresponding to the user input query; when unable to locate a query answer, establishes a second communication connection for online chat between the user interface and a live agent interface that transmits responses dynamically provided by a human operator; evaluates a chat between the user interface and the live agent interface; identifies an answer to the user input query, based on evaluating the chat; and stores the answer to be provided by the AI entity in the future.

Abstract: The technology disclosed supports content collaboration between any number of users and systems by constructing WYSIWYG (What You See Is What You Get) outputs of changes proposed by the users to a shared live application or to a document such as code, text, or graphics. The technology disclosed allows authors to update consumer facing content instantaneously based on walkthrough feedback from other contributors or co-authors. Further, the authors, without requiring technical expertise of a developer, can immediately see the consequences of incorporating changes suggested by the contributing authors, based on the WYSIWYG outputs of the changes proposed by the authors.

Abstract: Outbound processing with application firewalls. An outbound message is generated with an application. The outbound message includes at least a trustworthiness indicator and/or marking information for the one or more portions of the outbound message. The outbound message is received by an application firewall. The outbound message is analyzed based on the trustworthiness indicator and/or marking information, and context information. An action is performed on the outbound message based on the trustworthiness indicator and/or marking information, and the context information.

Abstract: In accordance with disclosed embodiments, there are provided systems, methods, and apparatuses for implementing analytics on-demand functionality.

Abstract: In accordance with disclosed embodiments, there are provided methods, systems, and apparatuses for implementing cross organizational data sharing including, for example, means for storing customer organization data in a database of the host organization; allocating at least a sub-set of the customer organization data to be shared as shared data; configuring a hub to expose the shared data to a proxy user and configuring the proxy user at the hub with access rights to the shared data; configuring one or more spokes with access rights to the shared data of the hub via the proxy user; receiving a request from one of the hubs for access to the shared data of the customer organization via the proxy user at the hub; and returning a response to the hub having made the request. Other related embodiments are disclosed.

Abstract: Systems and methods are provided for viewing a contact network feed in a business directory environment. A system searches, based on a user specified search, a business directory to determine contacts, identifies network feeds associated with the contacts, and outputs the network feeds. For example, the system outputs network feeds from data center managers identified by the business directory. The system filters, based on a user specified filter, the network feeds to determine filtered network feeds, and outputs the filtered network feeds to a display device. For example, the system outputs the network feeds from data center managers that mention a specific product. The system can enable communication with a contact using contact information from the business directory. The system can subscribe a user to the network feed of a contact, wherein the user receives updates for the network feed of the contact via a feed to a display device.