Abstract: The decoder network includes multiple decoders trained to generate different types of summaries. The lower layers of the multiple decoders are shared. The upper layers of the multiple decoders do not overlap. The multiple decoders generate probability distributions. A gating mechanism combines the probability distributions of the multiple decoders into a probability distribution of the decoder network. Words in the summary are selected based on the probability distribution of the decoder network.

Abstract: The present disclosure describes an electronic document review method comprising the steps of displaying at least a first document section of said document in a first display window, displaying at least a second document section of said document in a second display window, and arranging said second window so that said second window is arranged within the border of the first window. Such a method allows for efficient review of electronic documents thereby enabling a user to reduce the amount of time the user spends for reviewing an electronic document. The disclosure also provides an electronic document review system.

Abstract: Mechanisms are provided to perform selective deep parsing of natural language content. A targeted deep parse natural language processing system is configured to recognize one or more triggers that specify elements within natural language content that indicate a portion of natural language content that is to be targeted with a deep parse operation. A portion of natural language content is received and a pre-deep parse scan operation is performed on the natural language content based on the one or more triggers to identify one or more sub-portions of the natural language content that contain at least one of the one or more triggers. A deep parse is performed on only the one or more sub-portions of the portion of natural language content that contain at least one of the one or more triggers, while other sub-portions of the portion of natural language content are not deep parsed.

Abstract: A system for generating graphical user interfaces. The system may include processors and storage devices storing instructions. The instructions may configure the one or more processors to perform operations including identifying a plurality of attributes from an image captured with a client device, identifying a plurality of first results based on the attributes, generating a first graphical user interface for display in the client device. The first graphical user interface may include a plurality of result icons corresponding to a subset of the first results having confidence scores above a threshold, a plurality of filter icons displaying options, and a search button. The operations may also include receiving a selection of at least one of the result icons or at least one of the filter icons, performing a search, based on the selection, and generating a second graphical user interface.

Abstract: A data processing system implements obtaining a first textual content, segmenting the first textual content into a plurality of first segments, and providing each segment of the plurality of first segments to a first natural language processing (NLP) model to obtain a set of first readability scores for the plurality of first segments. The first NLP model is configured to analyze a textual input and to output a readability score representing a measurement of readability of the textual input. The system further implements aggregating the set of first segment readability scores to determine a first readability score for the first textual content, and perform at least one of causing the first readability score to be presented to a user or performing one or more actions on the first textual content based on the readability score.

Abstract: Computing systems, computing apparatuses, computing methods, and computer program products are disclosed for machine learning ranking. An example computing method includes receiving a search query and determining a plurality of machine learning model execution engines based on the search query and a plurality of search result types. The example computing method further includes generating a plurality of subsets of search results based on the search query and the plurality of machine learning model execution engines. The example computing method further includes generating a set of search results comprising at least one search result from each of the plurality of subsets of search results.

Abstract: Embodiments described herein provide document summarization systems and methods that utilize fine-tuning of pre-trained abstractive summarization models to produce summaries that more faithfully track the content of the documents. Such abstractive summarization models may be pre-trained using a corpus consisting of pairs of articles and associated summaries. For each article-summary pair, a pseudo label or control code is generated and represents a faithfulness of the summary with respect to the article. The pre-trained model is then fine-tuned based on the article-summary pairs and the corresponding control codes. The resulting fine-tuned models then provide improved faithfulness in document summarization tasks.

Abstract: The present application relates to defining acronyms in messages between a sender device and a receiver device. The sender device receives a text content of a message from a user and parses the text content to identify one or more acronyms within the message. The sender device presents, to the user, a secondary task pane including the one or more acronyms present within the message and corresponding acronym definitions in editable text fields and receives, from the user, an update to one or more of the corresponding acronym definitions. The sender device sends the message and metadata including the corresponding acronym definitions to one or more other users. A receiver device receives the message and converts each of the acronyms to a respective hyperlink. The receiver device detects a user interaction with the respective hyperlink for a respective acronym and presents at least the respective first acronym definition.

Abstract: The present disclosure provides a method and apparatus of pre-rending a page, an electronic device, and a storage medium. The method of pre-rendering the page may include: acquiring a static resource of a page to be processed; acquiring a hypertext markup language structure information compiled of the page; and generating a static hypertext markup language file of the page by using a native server-side rendering function of a web frame according to the static resource and the hypertext markup language structure information.

Abstract: A system and method are provided for generating and publishing electronic content using predetermined templates. The method comprises providing a predetermined template specifying a predetermined layout for at least one slide to contain content; enabling the content to be added to the template by a plurality of contributors, each contributor having access to a project comprising the template over a network; receiving the content; and compiling a published output according to the received content and parameters of the predetermined template.

Abstract: An automated content annotation workflow is disclosed. An example embodiment is configured for: registering a plurality of labelers to which annotation tasks are assigned; populating a labeling queue with content data to be annotated; assigning annotation tasks from the labeling queue to the plurality of labelers; enabling the plurality of labelers in an annotation review queue to modify or delete annotations applied by prior labelers; and evaluating a level of performance of the plurality of labelers in applying the annotations.

Abstract: A method for displaying media information includes: receiving a media information request sent by an originating client, where the media information request carries media information and destination client information; parsing the received media information to obtain text information corresponding to the media information; synthesizing information related to the media information with information related to the text information to obtain composite information; and sending the composite information to a destination client according to the destination client information, so that the destination client obtains the media information and the text information according to the composite information and displays the media information and the text information.

Abstract: Embodiments described herein provide methods and systems for presenting a document and generating a human-AI summary. A system provides a user with a selection of an amount of time to spend reading the document, or a list of questions from which the user may select which questions they would like answered by reading the document. The system highlights sections of the document according to the user selection. Implicit and explicit user data such as dwell times, user highlights, and user notes, are collected while displaying the document. A human-AI summary is generated based on the document and the user data.

Abstract: A sensory content production system generates sensory associated with concepts identified in an electronic document. Definitions of each concept are identified, and the concepts are stored in connection with their associated definitions. Sensory content is generated for the concepts and concepts definitions and stored in connection with each other in a sensory content layer. Availability of sensory content is indicated to users by interactive markers.

Abstract: Systems, methods, and computer-readable media for automatically updating an electronic word processing document based on a change in a linked file and vice versa are disclosed. The systems and methods may involve accessing the electronic word processing document; identifying in the electronic word processing document a variable data element; accessing the external file identified in the link; pulling, from the external file, first replacement data corresponding to the current data; replacing the current data in the electronic word processing document with the first replacement data; identifying a change to the variable data element in the electronic word processing document; upon identification of the change, accessing the external file via the link; and updating the external file to reflect the change to the variable data element in the electronic word processing document.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for evaluating interactions with a user interface are disclosed. In one aspect, a method includes identifying a set of mutation events specifying changes to the structure of a user interface that occurred during the user session, and generating user interface states specifying different structures of the user interface throughout the given user session. Based at least in part on the user interface states, playback data that present visual changes of the user interface corresponding to the set of mutation events that occurred during the user session are generated. Session activity data describing user interactions that occurred during the user session are also generated. At least a portion of the playback data and the session activity data are output to a requesting device.

Abstract: The present disclosure relates to UI systems and processes including methods for controlling text position in a computer display. A target word in a body of text may be maintained in position by forward rendering and backward rendering, iteratively as the text is modified by the addition or deletion of words or by modifications affecting height or width of a word.

Abstract: The present disclosure relates to UI systems and processes including rendering of text in a combination of styles and sizes to a computer graphics display. Text may be rendered with one or more emphases to modify style, size, and weight according to emphases weightings and or relative positions of text to be emphasized.

Abstract: Generating a table of contents from a computer document is disclosed. The computer document is converted into a markup language, from which a list of grouped textblocks is generated. Headings are detected from among the list of grouped textblocks. For a grouped textblock, a first vector corresponding to a semantic representation of the grouped textblock and a second vector based on evaluation of pre-defined features in the grouped textblock are generated. Based on the first and second vectors, the grouped textblock is classified as a heading or a plain-text using a trained classifier.

Abstract: In some implementations, a system dynamically adjusts an electronic form of an application based on user data and form data specified for the electronic form. Form data that specifies characteristics of the electronic form is initially received by a computing device. The form data specifies one or more rules configured to vary the content of the electronic form that is presented for different users. User data indicating characteristics of a user is then obtained by the computing device. An interactive element is then selected from among a set of multiple interactive elements based on the received user data and the rules in the received form data. A view of the electronic form is customized for the user by including, in the customized view, the interactive element. The view of the customized electronic form is then displayed by the computing device.