Abstract: Techniques for user behavior anomaly detection. At least one low-variance characteristic is compared to an expected result for the corresponding low-variance characteristics to determine if the low-variance characteristic(s) is/are within a pre-selected range of the expected results. A security response action is taken in response to the low-variance characteristic not being within the first pre-selected range of the expected results. At least one high-variance characteristic is compared to an expected result for the corresponding high-variance characteristics to determine if the high-variance characteristic(s) is/are within a pre-selected range of the expected results. A security response action is taken in response to the high-variance characteristic not being within the first pre-selected range of the expected results. Access is provided if the low-variance and the high-variance characteristics are within the respective expected ranges.

Abstract: Trigger-free asynchronous maintenance of custom indexes and skinny performance meta-structures is described. A database system receives a request to manipulate data in an object in a database table, and determines whether the object in the database table is associated with a user-based data structure. If the object in the database table is associated with the user-based data structure, the database system manipulates the data in the object in the database table and data in an object in an intermediate data table. The database system receives a query, and determines whether the query requests any data that is in the user-based data structure. If the query requests any data that is in the user-based data structure, the database system creates a query result by combining any data that is in the object in the intermediate data table with any requested data that is in the user-based data structure, and outputs the query result.

Abstract: A system for text summarization includes an encoder for encoding input tokens of a document and a decoder for emitting summary tokens which summarize the document based on the encoded input tokens. At each iteration the decoder generates attention scores between a current hidden state of the decoder and previous hidden states of the decoder, generates a current decoder context from the attention scores and the previous hidden states of the decoder, and selects a next summary token based on the current decoder context. The selection of the next summary token prevents emission of repeated summary phrases in a summary of the document.

Abstract: A system is provided for object localization in image data. The system includes an object localization framework comprising a plurality of object localization processes. The system is configured to receive an image comprising unannotated image data having at least one object in the image, access a first object localization process of the plurality of object localization processes, determine first bounding box information for the image using the first object localization process, wherein the first bounding box information comprises at least one first bounding box annotating at least a first portion of the at least one object in the image, and receive first feedback regarding the first bounding box information determined by the first object localization process. The system is further configured to persist the image with the first bounding box information or access a second object localization process based on the first feedback.

Abstract: A method for managing a user's email subscriptions has been developed. The method provides a user interface (UI) to the user that allows the user to view and select multiple stand-alone components across multiple dashboard displays for an email subscription to an email address. The user's selections are received via the UI and include one or more stand-alone components from a first dashboard display and one or more stand-alone components from a second dashboard display in the email subscription. The first dashboard display is not the same was the second dashboard display. A display of each of the subscribed stand-alone components is rendered periodically based on the user's preference. The rendered displays are combined into a single email message that is sent to the email address upon completion of the rendering of the displays.

Abstract: Systems and methods for generating and testing formulas written using a formula language may include receiving, by a server computing system, in a formula input area a formula written using a formula language, the formula associated with a record in a database; displaying, by the server computing system, one or more fields in a field area based on the one or more fields being included in the formula, the one or more fields associated with the record; enabling, by the server computing system, a value associated with the one or more fields to be modified based on the one or more fields being included in the formula; and displaying, by the server computing system, in a result area a result generated based on execution of the formula, wherein the result area is to be visible concurrently with the formula input area and the field area.

Abstract: Implementations store an event associated with a topic in a database to allow for the event to be transmitted in order with respect to other events associated with the topic which are stored in the database. A commit of a transaction is performed that includes inserting the event's payload in a record in a database object. After the commit, a field for the record is updated with an identifier of the commit. The identifier and identifiers of other commits of other transactions on the database indicate ordering of the commit with respect to the other commits Another field of another record is updated based on the identifier and a last stored value such that the payloads of the event and of the other events are retrievable to be transmitted in an order based on the value of the other field and values of other fields of other records.

Abstract: Implementations are described for selectively publishing an event responsive to an operation on a database in a transaction. Publishing the event is delayed or not delayed respectively, responsive to determining that a value of an attribute for the event indicates that publishing the event is to be delayed, or not to be delayed, until the transaction is rolled back or committed. If publishing the event is delayed, the event is published if the transaction is committed and discarded if the transaction is rolled back. If publishing the event is not delayed, the event is published. Optionally, if the event is published, the event is published in a database or a datastore respectively, responsive to determining whether another value of another attribute for the event indicates that the event is to be published in a database or a datastore.

Abstract: Disclosed herein are system, method, and device embodiments for implementing visual context in dataflows. A computing device may display a dataflow representation associated with a transformative process, receive selection of an individual node of the dataflow representation, and determine a visual context associated with the individual node based on a data transformation step associated with the individual node. Further, the computing device may determine a plurality of nodes associated with the visual context, and apply a graphical effect to the individual node and the plurality of nodes, the graphical effect modifying an attribute of the individual node and the plurality of nodes.

Abstract: Implementations are described which provide for recipient-based filtering of an event that relates to a topic to which consumers are subscribed. Responsive to determining that an attribute of the event includes a set of one or more identifiers for intended recipients for the event, the event is delivered to consumers that correspond to the intended recipients. Alternatively, responsive to determining that the attribute of the event does not include a set of one or more identifiers for intended recipients for the event, the event is delivered to all of the consumers subscribed to the topic to which the event relates.

Abstract: Disclosed herein are system, method, and device embodiments for implementing dynamic customizable animations. A multi-tenant service may configure a visual component of an application to present an animation based on a rule, generate a multi-tenant entity record defining the rule, and generate application code associated with the application. Further, the multi-tenant service may receive a request for animation information associated with the multi-tenant entity record, and send the animation information to the client device. Further, in some embodiments, a client device executing the application may present the animation based on evaluating the animation information.

Abstract: Disclosed herein are system, method, and device embodiments for executing an application program interface (API) plugin for use in conjunction with a vector editing design tool. For example, a method may include: retrieving contextual information related to a plurality of features of a design tool upon initialization of the design tool, the contextual information being stored on a remote server from one or more computing devices; detecting a placement of a first feature from among the plurality of features in a design environment of the design tool; and displaying the contextual information in the design environment in response to detecting the placement of the first feature.

Abstract: A method is provided for generating a ranked list of candidate responders. In some embodiments, the method includes receiving a question from a user and generating a question feature vector representing an intent of the question and a first skill set inferred from the question. The method also includes for one or more candidate responders, generating a candidate feature vector representing a skill set and questions associated with the respective candidate responder; computing a reputation score based on questions and user feedback associated with the respective candidate responder; and computing, based on the question feature vector, candidate feature vector, and reputation score, a probability score representing a prediction of the quality of an answer that would be provided by the respective candidate responder if the input question were routed to the respective candidate responder. The method further includes generating a ranked list of candidate responders using the computed probability scores.

Abstract: A geo-analytical program is integrated into a customer relationship management (CRM) solution. Via the geo-analytical program, users are able to define layer configuration settings for a layer for plotting on a map-based GUI. Layer configuration settings for a respective layer comprise an indication of a data object type serving as a base object type for the respective layer. A method involves receiving from a user user-defined configuration settings for a first layer, electronically receiving, at a geospatial computer system, geospatial data for a plurality of connected devices; electronically receiving, at the geo-analytical program from the geospatial computer system, real-time geospatial data for the plurality of connected devices; and utilizing, by the geo-analytical program, the user-defined layer configuration settings for the first layer to plot the first layer on the map-based GUI based on the received real-time geospatial data for the plurality of connected devices.

Abstract: A moderation framework monitors content posted in a database network and uses a declarative configuration scheme that defines moderation rules without having to write new software. A user interface operated by the moderation platform allows selection of different user criteria and different content criteria for triggering the moderation rules. The user interface also provides selectable actions for the moderation rules to apply when the content matches the selected user and content criteria. The user interface also allows selection of different entities for associating with different moderation rules, such as accounts, cases, opportunities, Chatter® feed posts, or custom objects defined by customers.

Abstract: Disclosed embodiments are related to send time optimization technologies for sending messages to users. The send time optimization technologies provide personalized recommendations for sending messages to individual subscribers taking into account the delay and/or lag between the send time and the time when a subscriber engages with a sent message. A machine learning (ML) approach is used to predict the optimal send time to send messages to individual subscribers for improving message engagement. The personalized recommendations are based on unique characteristics of each user's engagement preferences and patterns, and deals with historical feedback that is generally incomplete and skewed towards a small set of send hours. The ML approach automatically discovers hidden factors underneath message and send time engagements. The ML model may be a two-layer non-linear matrix factorization model. Other embodiments may be described and/or claimed.