Abstract: A contextual navigation and action user experience framework that facilitates workflows across multiple contexts and levels of object hierarchy is disclosed. Exemplary features include a swapper interface, an action toolbar with contextual buttons and contextual tabs, and a toolkit that provides an overview portal to view alerts, cross reference information, and perform actions on objects and insights that are important to the user across an entire product suite.

Abstract: A method for facilitating automated data curation in real-time is disclosed. The method includes retrieving electronic documents from a source; converting the electronic documents into data sets, the data sets corresponding to a predetermined format; preprocessing the data sets to identify linguistic units, the linguistic units relating to paragraphs and sentences; extracting, by using a model, attributes based on the linguistic units, the attributes relating to a key detail in the electronic documents; and generating, in real-time, messages based on the extracted attributes. Additionally, the electronic documents include a corporate action prospectus that provides information for a corresponding corporate event, the information including term and condition information, date information, and restriction information.

Abstract: The disclosed embodiments include a method performed by computer system. The method can include retrieving a layout object including layout information, receiving an indication of a request for a formatted document, and determining that the formatted document should have multiple pages with a common layout feature. The method can further include dynamically generating a template for the formatted document based on the layout information of the layout object and generating the formatted document by using the template to create the multiple pages of the formatted document with the common layout feature.

Abstract: Systems, methods, and computer-readable storage media for aggregating media (and commentary on that media) into a topology. To do so, the system receives first content (and associated metadata) as well as second content (and associated metadata). The system then generates a topology based on a relationship between the first content and the second content, where the topology has a number of dimensions based on the metadata of the different pieces of content. The system then compares the topology and the metadata to previously stored topologies and/or metadata and, based on that comparison, executes a machine learning algorithm. The output of that machine learning algorithm includes predicted future changes to the topology, which the system uses to reduce the number of dimensions within the topology.

Abstract: A system and method are provided for examining data from a source. The method is executed by a device having a processor and includes receiving a set of historical data and a set of current data to be examined, from the source. The method also includes generating multiple statistical models based on the historical data and a forecast for each model. The method also includes selecting one of the multiple statistical models based on at least one criterion, and generating a new forecast using the selected model. The method also includes comparing the set of current data against the new forecast to identify any data points in the set of current data with unexpected values. The method also includes outputting a result of the comparison, the result comprising any data points with unexpected values.

Abstract: Dynamic web content personalization may be used to alter a web page such that two different people who view that web page do not see the exact same content on that page. The present disclosure includes techniques that allow a web page to be structured, via underlying source code, so that content can be dynamically generated based on a user transaction history. Further, dynamic content customization can effectively change the layout and functionality of a web page in ways not previously contemplated. For example, different control elements may become accessible based on a particular contextual user profile assigned to a user based on a transaction history. Modular content modules may also have a different appearance for different users.

Abstract: Various embodiments, methods and systems for implementing a form generation system are provided. Generating forms includes dynamic generation, personalization, and optimization of the forms based on automation objects that instruct on how to construct, structure and present forms for personalized data capture experiences. In operation, a form generator engine receives a request from a computing device to access a form. The form generator engine accesses form generation automation rules that are based on form generation parameters and automation objects. Using form generation automation rules, form generation parameters are used to generate automation objects including an annotated schema, a machine learning model, and a layout. Based on the form generation automation rules the automation objects are used to generate the form such that at least a field or a section of the form is selected based on a relevance score associated with field or section. The form is communicated for display.

Abstract: The embodiments provide a system and method for detecting sensitive data in one or more documents, which may be housed in a data storage system. More specifically, the embodiments provide a system that can not only identify potentially sensitive data based on pattern matching, but which can also infer context in unstructured domains. Context may be inferred using both part of speech analysis and by analyzing the similarity of potentially sensitive data with key terms using a vector space model.

Abstract: In general, the technology relates to navigation control in a web application that includes receiving an event of a navigation trigger in a first page of a web application, and obtaining, in response to the event, an expression for the navigation trigger. The technology further includes gathering data from a first data source specified by the expression. The technology further includes evaluating the expression using the data to select a second page from a set of possible target pages, where each of the possible target pages is separately pre-stored in a data repository. The technology further includes presenting the second page.

Abstract: There are provided systems and methods for automatic data entry for form data structures using application programming interfaces. An entity, such as company or business, may want to onboard with a service provider that provides services to the entity, such as those associated with business credit, expense management, and the like. When onboarding, data entry to one or more digital form fields and structures may be required. In order to automate form filling based on data from disparate sources, the service provider may provide an external application programming interface (API) that receives API calls and converts and/or formats data for the fields and data structures of the digital form. The data entered to the fields and/or corresponding data structure may be stored in a specific database and tables as data structures retrievable for the digital form. When a referral token is received, the data may be retrieved.

Abstract: Enhanced metadata with optimized output. In an embodiment, a content object is received. First metadata to be associated with the content object is determined. At least one metadata field to be acquired is determined based on an association, within a stored knowledge structure, of the at least one metadata field with the first metadata. Second metadata to be associated with the content object is acquired based on the at least one metadata field. A metadata structure and markup format for the content object are determined. The metadata structure incorporates both the first metadata and the second metadata. The content object is output with the metadata structure and in the markup format.

Abstract: Each of a plurality of wire web components predicted to be included in a graphical user interface (GUI) for presentation at a client machine. Each data object instance may be associated with a respective data object instance identifier and a respective one or more data object fields. A wire web component graph that includes a plurality of nodes may be constructed. A node may correspond to a wire web component, a data object field, or an application procedure interface (API). One or more of the data values may be retrieved from the respective APIs based on the wire web component graph. A GUI message including the retrieved data values and the wire web component graph may be transmitted to the client machine.

Abstract: The present disclosure relates to processing operations that generate and present personalized content suggestions to assist a user with document creation. Machine learning modeling may be trained and implemented to evolve pre-canned suggestions for document creation into highly personalized content suggestions, thereby improving the document creation process and user interface experience for users of applications/services that are utilized to create digital documents. As an example, signal data may be detected and analyzed, identifying a specific user's intent to create a digital document. Machine learning modeling may be implemented to evaluate different aspects of collected signal data and identify content from previously created documents, associated with a user account, that may be most relevant to the real-time document creation experience of the user. Personalized contextual suggestions may be presented to a user through a user interface.

Abstract: Computerized systems and methods are disclosed to generate a document by providing a document structure having one or more seed landmark texts therein, each landmark text including a milestone overview text and a plurality of component texts; from the milestone overview text, generating one or more computer-generated text suggestions to supplement the milestone overview text; combining the milestone overview text with each component text and generating one or more computer-generated component text suggestions; and creating the document by combining the milestone overview, the one or more computer-generated text suggestions, and each component text with corresponding one or more computer-generated component text suggestions.

Abstract: As described herein, various embodiments of the present invention provide methods, apparatus, systems, computing devices, computing entities, and/or the like for performing natural language processing operations using a combination of a cross-token attention machine learning, a cross-utterance attention machine learning model, and an integer linear programming joint keyword-utterance optimization model to select an extractive keyword summarization of a multi-party communication transcript data object that comprises a selected utterance subset of U utterances (e.g., U sentences) of a document data object and a selected keyword subset of K candidate keywords of the document data object.

Abstract: A computer-implemented method for displaying at least a portion of content being displayed on a first device to also be displayed on a second device. The method includes causing content to be displayed on a first device. The method then includes detecting a second device to concurrently display at least a portion of the content being displayed on the first device. A capability of the first device is compared with a capability of the second device. Based on the comparing, the at least the portion of the content is automatically provided to be displayed on the second device.

Abstract: Techniques are described including displaying an interface associated with a computing device, the interface being configured to present a form including a text entry box, receiving data from an activity information pipeline associated with a self-reporting service, the activity information pipeline having a source of activity information associated with a tool configured to geotag the activity information and another source of associated with a mobile sensor configured to capture other activity information geotagged by the tool, the other activity information being processed by the self-reporting service to deprioritize a suggestion over another suggestion, displaying, in response to selection of the text entry box, a suggestion indicative of a journey and identified as prioritized and presented proximate to the center of the interface of the display device, and dynamically adjusting the suggestion while prioritizing display of the suggestion over a generic suggestion.

Abstract: A machine learning system executed by a processor may generate predictions for a variety of natural language processing (NLP) tasks. The machine learning system may include a single deployment implementing a parameter efficient transfer learning architecture. The machine learning system may use adapter layers to dynamically modify a base model to generate a plurality of fine-tuned models. Each fine-tuned model may generate predictions for a specific NLP task. By transferring knowledge from the base model to each fine-tuned model, the ML system achieves a significant reduction in the number of tunable parameters required to generate a fine-tuned NLP model and decreases the fine-tuned model artifact size. Additionally, the ML system reduces training times for fine-tuned NLP models, promotes transfer learning across NLP tasks with lower labeled data volumes, and enables easier and more computationally efficient deployments for multi-task NLP.

Abstract: A computer-implemented method of providing specialized media on a website is disclosed. The method includes producing campaign media which is delivered to a user through a third-party provider. After the user has received the campaign media, and upon indication of a request from the user to access a website controlled by a content provider, specialized media on a website controlled by the content provider may be provided. Providing the specialized media may be done based on the campaign media previously delivered to the user through the third-party provider.

Abstract: An example method is provided for encoding text for language processing. The method may be executed by a processing system, and the method includes receiving text comprising a plurality of alphanumeric characters or symbols and converting the text into a numerical vector comprising a plurality of numerical values, by mapping each alphanumeric character or symbol of the text to a vertex coordinate of one of a plurality of vertices of a hypercube, wherein a number of the plurality of vertices is equal to or greater than a number of the plurality of alphanumeric characters or symbols, wherein the numerical vector consumes less space in memory than the text. An amount of time consumed by language processing of the numerical vector may be less than an amount of time consumed by language processing of the text.