Abstract: A method may include generating, based on text associated with an interaction between a user and an application, a query corresponding to an intent of the interaction. The query may be generated by applying one or more machine learning models trained to determine the intent of the interaction and to generate the query by mapping, to a corresponding slot in a template query, one or more named entities included in the first text. The query may be sent to an analytics engine not associated with the application to retrieve, from a database associated with the analytics engine, data relevant to the interaction between the user and the application. A data presentation providing a visual representation of the data retrieved from the database may be generated for display at a client associated with the user. Related systems and articles of manufacture are also provided.

Abstract: Embodiments provide an overview visualization of graph data by rendering a graph's structure based upon different node/edge types, rather than upon a total number of nodes and/or edges. These node/edge types are recognized based upon associated metadata. The small number of different node/edge types of a typical graph reduces components for representation in the overview, simplifying that visualization and facilitating user interaction. A user can switch between the graph overview, and a different view affording node-level details. In analysis and/or querying, switching between the overview and the detailed visualization can afford natural and intuitive control over data granularity. For example the overview visualization of graph data may allow a user to initially filter on particular node types, with the subsequent node-level view permitting exploration of finer details while maintaining the overall context of the original graph.

Abstract: A method may include generating, based on text associated with an interaction between a user and an application, a query corresponding to an intent of the interaction. The query may be generated by applying one or more machine learning models trained to determine the intent of the interaction and to generate the query by mapping, to a corresponding slot in a template query, one or more named entities included in the first text. The query may be sent to an analytics engine not associated with the application to retrieve, from a database associated with the analytics engine, data relevant to the interaction between the user and the application. A data presentation providing a visual representation of the data retrieved from the database may be generated for display at a client associated with the user. Related systems and articles of manufacture are also provided.

Abstract: A system includes reception of a selection of a point of a first data visualization associated with a first measure value, reception of an instruction from the user to create a visualization based on the first measure value, determination of a first context of the first measure value, the first context comprising one or more dimension values, generation of a first numeric point visualization of the first measure value based on the first context, and presentation of a first interface comprising the first data visualization and the first numeric point visualization, where the first numeric point visualization is presented in association with the selected first measure of the first data visualization.

Abstract: A system includes reception of a selection of a point of a first data visualization associated with a first measure value, reception of an instruction from the user to create a visualization based on the first measure value, determination of a first context of the first measure value, the first context comprising one or more dimension values, generation of a first numeric point visualization of the first measure value based on the first context, and presentation of a first interface comprising the first data visualization and the first numeric point visualization, where the first numeric point visualization is presented in association with the selected first measure of the first data visualization.

Abstract: A system, method, and computer-readable medium to receive a request to render a visualization, the visualization being defined by at least one dimension of a data set and being one of multiple visualizations belonging to a story; determine whether the at least one dimension of the visualization has a color assigned thereto; automatically assign, in an instance the at least one dimension of the visualization lacks a color assignment, a color to the at least one dimension of the visualization; store the color assigned to the at least one dimension in a color synchronization map; render, in response to the request, all visualizations in the story with the at least one dimension in the color assigned thereto in the color synchronization map.

Abstract: Embodiments provide for querying and visualization of query results of graph data. An engine processes graph data to create metadata (e.g., in JSON format) identifying at least different node types and edge types that are present in a graph database. An overview visualization simplifies presentation of graph data by depicting only various different node types and graph types. The overview visualization may form the basis for formulating queries including the metadata, which are then promulgated to the graph database. Returned query results may be visualized as an overview or on a detailed node level, promoting insight and formulation of additional queries including node/edge type metadata. The engine may convert graph data query results into tabular form for consumption by relational database analytical tools. According to particular embodiments, an engine of an in-memory database may be particularly suited to perform graph data visualization, querying, and/or tabular conversion tasks.

Abstract: A system includes reception of a selection of a point of a first data visualization associated with a first measure value, reception of an instruction from the user to create a visualization based on the first measure value, determination of a first context of the first measure value, the first context comprising one or more dimension values, generation of a first numeric point visualization of the first measure value based on the first context, and presentation of a first interface comprising the first data visualization and the first numeric point visualization, where the first numeric point visualization is presented in association with the selected first measure of the first data visualization.

Abstract: A system, method, and computer-readable medium to receive a request to render a visualization, the visualization being defined by at least one dimension of a data set and being one of multiple visualizations belonging to a story; determine whether the at least one dimension of the visualization has a color assigned thereto; automatically assign, in an instance the at least one dimension of the visualization lacks a color assignment, a color to the at least one dimension of the visualization; store the color assigned to the at least one dimension in a color synchronization map; render, in response to the request, all visualizations in the story with the at least one dimension in the color assigned thereto in the color synchronization map.

Abstract: A system includes reception of a selection of a point of a first data visualization associated with a first measure value, reception of an instruction from the user to create a visualization based on the first measure value, determination of a first context of the first measure value, the first context comprising one or more dimension values, generation of a first numeric point visualization of the first measure value based on the first context, and presentation of a first interface comprising the first data visualization and the first numeric point visualization, where the first numeric point visualization is presented in association with the selected first measure of the first data visualization.

Abstract: Embodiments provide an overview visualization of graph data by rendering a graph's structure based upon different node/edge types, rather than upon a total number of nodes and/or edges. These node/edge types are recognized based upon associated metadata. The small number of different node/edge types of a typical graph reduces components for representation in the overview, simplifying that visualization and facilitating user interaction. A user can switch between the graph overview, and a different view affording node-level details. In analysis and/or querying, switching between the overview and the detailed visualization can afford natural and intuitive control over data granularity. For example the overview visualization of graph data may allow a user to initially filter on particular node types, with the subsequent node-level view permitting exploration of finer details while maintaining the overall context of the original graph.

Abstract: Embodiments provide for querying and visualization of query results of graph data. An engine processes graph data to create metadata (e.g., in JSON format) identifying at least different node types and edge types that are present in a graph database. An overview visualization simplifies presentation of graph data by depicting only various different node types and graph types. The overview visualization may form the basis for formulating queries including the metadata, which are then promulgated to the graph database. Returned query results may be visualized as an overview or on a detailed node level, promoting insight and formulation of additional queries including node/edge type metadata. The engine may convert graph data query results into tabular form for consumption by relational database analytical tools. According to particular embodiments, an engine of an in-memory database may be particularly suited to perform graph data visualization, querying, and/or tabular conversion tasks.