User interface for visualizing high-dimensional datasets

Inventors

Savary, Simon

Assignees

Bentley Systems Inc

Abstract

In example embodiments, a user interface of a software application is provided for visualizing high-dimensional datasets, which simultaneously displays marginal distributions and joint distributions of variables that represent different attributes (e.g., properties) of entities (e.g., elements of infrastructure). The marginal distributions and joint distributions are combined into a single visualization that may be shown in a single window of the application. The visualization may include a graph (e.g., a bar chart) for each of the variables showing the marginal distribution of the variable, wherein each graph is displayed along a different portion of a perimeter of a closed shape (e.g., a circle). The visualization may also include graphics (e.g., lines) connecting portions of the bar charts showing the joint distribution for possible pairs of variables, wherein each graphic (e.g., line) is displayed with visual properties (e.g., a thickness) that indicates co-occurrence frequency of values of the variables.

Core Innovation

A method is disclosed for visualizing a high-dimensional dataset in a user interface of a software application executing on a computing device. The high-dimensional dataset includes a plurality of values for each of three or more variables, where each variable represents a different attribute of entities. The method accesses the high-dimensional dataset, calculates a marginal distribution for each variable, and displays a bar chart for each variable showing the marginal distribution along a different part of a perimeter of a closed shape.

The method further calculates a joint distribution for possible pairs of variables in the variables. In the user interface, lines are displayed connecting portions of the bar charts that correspond to the joint distribution for each possible pair of variables. Each line is displayed with one or more visual properties that indicate co-occurrence frequency of values of variables. This arrangement enables a single window visualization that combines marginal distributions with pairwise joint distributions.

A computing device implementation is also described in which a processor and memory store and execute a software application configured to visualize a dataset. The software application accesses the dataset, receives a selection of variables to visualize in a user interface, calculates marginal distributions and joint distributions for possible pairs of selected variables, and displays the results using bar charts and connecting lines. The description also includes categorical variables ordinally encoded and continuous variables organized by binning ranges before calculating the marginal and joint distributions, and normalization of co-occurrence frequencies to support the visual encoding.

Claims Coverage

The independent claims cover a method, a computing device, and non-transitory electronic-device readable media instructions to compute and display marginal and joint distributions of a high-dimensional dataset using a closed-shape perimeter layout with connector graphics whose visual properties indicate co-occurrence frequency. Across these independent claims, the core inventive features are the marginal distribution bar charts positioned along different parts of a perimeter of a closed shape, and the joint distribution shown as connecting lines whose visual properties encode co-occurrence frequency of variable values for each possible pair of variables.

The independent claims are directed to a visualization approach that computes marginal distributions per variable and displays them as bar charts or graphs along different parts of a closed-shape perimeter, while computing pairwise joint distributions and displaying them as connector lines or graphics with visual properties indicating co-occurrence frequency, with at least one claim emphasizing crossing connectors over the interior of the closed shape.

Stated Advantages

Documented Applications