Abstract: Techniques and systems for facilitating contextual information lookup and navigation are presented. Contextually relevant information can be presented without requiring a user to switch context or craft a search query. From within an application in which a user may consume or create content, a user may indicate text for further exploration. The application can receive the indication of an initial text selection, determine the intended selection and provide contextual search results to the user. In some implementations, the initial selection and a selection context (that includes at least some of the text surrounding the initial selection) is sent to a contextual search service which provides the intended selection and contextual search results based on the initial selection and selection context. In other implementations, text may be suggested by the contextual search service for selection from within the application.

Abstract: Specialized digital assistants are described that provide scoped natural language, outcome focused options. Techniques and systems for generating the specialists are also provided so that the specialists can be embedded in or employed by other software applications. A specialist can be added directly to a website or application so that visitors to the website or application can more easily interact with the website or application and obtain desired outcomes. More than one specialist may be included in a website or application. The specialist is supported by a back-end specialized digital assistant service (“service platform”) that is scoped to the particular aspects of the website or application. The service platform can provide developer assistance by, upon receiving an indication that a specialized digital assistant component is being included in an application or a webpage, setting up a specialized digital assistant for that component.

Abstract: A document lens service identifies an end-user targeted to consume a productivity document being authored in a user interface to a productivity application. The service then also identifies a document profile specific to the end-user indicative of a model document for the end-user to consume. While the productivity document is being authored, suggestions may be surfaced in the user interface for aligning the productivity document with the model document.

Abstract: A document lens service identifies a document profile specific to an end-user that is indicative of a model document for the end-user. When a productivity document is opened to be consumed by the end-user in a user interface to a productivity application, the service modifies the productivity document based at least on the document profile to align the productivity document with the model document. The productivity document is then rendered as-modified in the user interface.

Abstract: A communication application creates a team mailbox integrating messages from an email repository and documents from a content management store (CMS). The application enables an administrator to define and manage members of the team mailbox with varying privilege status for accessing the team mailbox. Content such as aggregated messages and associated documents are displayed through a unified view to standardize ease of access. The members are enabled to receive messages sent to the team mailbox by others. Policies inherited from email repository, CMS, and/or third party policy providers are enforced within the team mailbox.

Abstract: A system for reuse of content within documents can include storage media on which a document graph is stored, and indexing software that, when executed by a processing system, directs the processing system to at least: identify content component types within a source document; and represent content components in a document graph at least connected by ‘contains’ edge to source document. A search of the document graph for reusable content can then be directed to related content components including those of a particular content component type or relevancy.

Abstract: Documents can be deconstructed into component blocks for reuse in productivity applications using a document shredder. The document shredder can identify specific objects within a source document that are likely to be reused, replicate the objects, and store the objects, or component blocks, separate from the source document. The productivity application can access the objects and present these objects to a user of the productivity application for easy insertion into the user's document. The component blocks, or identified relevant objects can be from document sources of previously generated content from a same or different author than the user of a productivity application.

Abstract: A system for reuse of content within documents can include storage media on which a document graph is stored, and labeler software that, when executed by a processing system, directs the processing system to at least: aggregate, for a window of time, a plurality of actions carried out by a user on content associated with a document within a productivity application; normalize the plurality of actions into at least one predefined action class; determine a label of the user in relation to the document or a portion of the content associated with the document based on the at least one predefined action class; and store the label associated with the user and the document or the portion of the content in a document graph. A search of the document graph for reusable content can then be directed to related content associated with a user having a particular label.

Abstract: Systems and techniques for graph-driven authoring in productivity applications can include using context of a user of a productivity application, the existing content of the user's document or an expression input with a command for finding relevant documents, and a date/time to obtain a set of results of relevant documents. The relevant documents include documents associated with people known to the user that contain content relevant to the topic(s) and/or entities from the context and that are current.

Abstract: A communication application creates a team mailbox integrating messages from an email repository and documents from a content management store (CMS). The application enables an administrator to define and manage members of the team mailbox with varying privilege status for accessing the team mailbox. Content such as aggregated messages and associated documents are displayed through a unified view to standardize ease of access. The members are enabled to receive messages sent to the team mailbox by others. Policies inherited from email repository, CMS, and/or third party policy providers are enforced within the team mailbox.

Abstract: By applying a set of simple geometric rules to the connections within a connected graph of ‘topics’ it is possible to uncover hidden relationships that are otherwise inaccessible to the lay person. Interesting, potentially non-obvious threads of content, termed indirect connection clusters (ICCs), can be found from an online encyclopedia or other graph of articles that are not directly connected to a starting topic, but instead are connected via an intermediate clique of articles. A system performing a search of an information graph can receive a request for identifying relevant content, identify (in a traversal of the information graph) one or more ICCs using a starting topic associated with the request, refine the one or more ICCs (eliminating certain ICCs) and generate a refined set of ICCs, and rank ICCs within the refined set. The ranked ICCs can be provided in response to the request for identifying relevant content.

Abstract: A text span forming either a single word or a series of two or more words that a user intended to select is predicted. A document and a location pointer that indicates a particular location in the document are received and input to different candidate text span generation methods. A ranked list of one or more scored candidate text spans is received from each of the different candidate text span generation methods. A machine-learned ensemble model is used to re-score each of the scored candidate text spans that is received from each of the different candidate text span generation methods. The ensemble model is trained using a machine learning method and features from a dataset of true intended user text span selections. A ranked list of re-scored candidate text spans is received from the ensemble model.

Abstract: An assisted content authoring productivity tool can provide a set of content related to at least one topic determined from a user's expression of intent. The content can include topics that are not obviously related to the user's topic, but identified from indirect cluster connections found in a graph model of information such as, but not limited to, an online encyclopedia like Wikipedia. A document can be populated with topic anchors so a user can return to or have generated a set of content related to a particular topic. The topic anchor can include a graphical user interface including a topic exploration object that, when selected, initiates a command to request information related to that topic.

Abstract: An assisted content authoring productivity tool can provide a set of structured content related to at least one topic determined from a user's expression of intent. Additional information can be obtained about a selected result from the set of structured content through the assisted content authoring productivity tool by using the set of structured content as context for the further search. An item from the set of structured content can be inserted into a document in a content authoring surface of a content authoring application. The insertion may be carried out cross-device. Topics for the search of structured content can be determined using the user's expression of intent and, once available, the context can include the results of previous searches for the structured content related to at least one topic determined from the user's expression of intent. Each additional search results may provide further context in a recursive manner.

Abstract: Techniques that facilitate the accomplishment of tasks within applications are presented. An intent-based user experience is available through receiving a natural language statement of intent from a user regarding use of an application, such as a productivity application. The graphical user interface for the user can be configured and reconfigured based on the user's intent; thus creating a task-oriented user interface. The user's intent can be determined through classifying and/or mapping the natural language statement of intent to particular tasks, which can then be associated with one or more tools and information that can be used to accomplish the tasks. The one or more tools and information can be surfaced to the user in the graphical user interface.

Abstract: Techniques and systems for facilitating contextual information lookup and navigation are presented. Contextually relevant information can be presented without requiring a user to switch context or craft a search query. From within an application in which a user may consume or create content, a user may indicate text for further exploration. The application can receive the indication of an initial text selection, determine the intended selection and provide contextual search results to the user. In some implementations, the initial selection and a selection context (that includes at least some of the text surrounding the initial selection) is sent to a contextual search service which provides the intended selection and contextual search results based on the initial selection and selection context. In other implementations, text may be suggested by the contextual search service for selection from within the application.

Abstract: Techniques and systems are presented for providing “contextual insights,” or information that is tailored to the context of the content a user is consuming or authoring. Given a request for information about a topic, which may be indicated by a user gesture in an application, one or more queries to search services may be formulated without requiring entry of a search query directly by a user. Moreover, techniques and systems may leverage the context of the content the user is consuming or authoring, as well as user, device, and application metadata, to construct the queries and to organize and filter the results into relevant contextual insights.

Abstract: A text span forming either a single word or a series of two or more words that a user intended to select is predicted. A document and a location pointer that indicates a particular location in the document are received and input to different candidate text span generation methods. A ranked list of one or more scored candidate text spans is received from each of the different candidate text span generation methods. A machine-learned ensemble model is used to re-score each of the scored candidate text spans that is received from each of the different candidate text span generation methods. The ensemble model is trained using a machine learning method and features from a dataset of true intended user text span selections. A ranked list of re-scored candidate text spans is received from the ensemble model.

Abstract: A method of automatically augmenting content through augmentation services can include invoking a service to receive an entity determination based on the content and an entity container comprising an object of the content and one or more attributes of the object. The entity determination and corresponding entity container can be assigned as a marker to content being clipped. The marker can be used to enable access and actions that can be taken with respect to the clipping or the structured information augmenting the clipping.

Abstract: Captured data can be transformed and augmented for a particular presentation in a document, such as a note of a notebook application, based on an identified entity for the captured data. The particular presentation of captured data can be provided based on entity detection, extraction, and knowledge base resolution and retrieval. Methods, systems, and services are provided that identify a primary entity of an item input to a notebook application and create an entity object for the primary entity of the item at least from one or more structured representations for content associated with the item. A template for presenting the entity object can be determined according to the primary entity, where the template is selected from a set of templates corresponding to different primary entities such that an arrangement and presentation for one primary entity is different than that of another primary entity.