Dynamic data visualized analysis tool based on workflows

Abstract

The invention discloses a dynamic data visualization analysis tool based on workflow. The invention extends the static multi-view display to the metaphorical interface of the bubble, so that the data analysis task can be supported by free layout. Each bubble is a functional unit, which can be used for programming calculations, and can also be used to create a visualization instance. It supports the operation of multiple bubbles forming a group, and the operation of bubbles in the same group is unified, which enhances the user experience. The invention can effectively connect different stages of the data analysis pipeline, reduce the efficiency loss of creating visualization instances and program transfer, and improve the efficiency of dynamic data visualization. Transforming causality-based data exploration into a pipelined approach can support transfer between procedures, automation of repetitive tasks, data capture records, and reuse of complex data analysis processes at different levels of detail. Provides more convenience for program pipelines in a broader framework.

Description

technical field [0001] The invention relates to the field of big data analysis, in particular to a workflow-based dynamic data visualization analysis tool. Background technique [0002] We found that some existing workflow tools do not well support the loop of the sense-making process. The meaning construction process of general scientists starts from the completion of data collection and ends with obtaining some meaningful knowledge from the data. The following three issues are mainly considered in the system design: [0003] The first problem is the independent separation of program implementation and data analysis, which forces scientists to switch between multiple working environments. Scientists use Matlab to perform analyzes and then visualize the calculated results to test hypotheses or generate new hypotheses. If they need to do more analysis, they need to switch back to the Matlab environment to make some changes. Different applications use different data format...

AI Technical Summary

Problems solved by technology

Different applications use different data formats, and scientists will have to adapt to these changes, which will seriously disturb the data analysis process and make it difficult to reproduce some scientific results

Switching back and forth between Matlab and the data analysis process directly introduces the overhead of context switching

[0004] The second problem is that some current visualization implementations require users to choose some encoding methods to establish a mapping between data and tags

But choosing an appropriate mapping from data to visual components is generally difficult, especially for diverse data

Some anomalies in the data that could have been found in the visualization were detected later in the analysis phase, resulting in a significant loss of efficiency

[0005] The third problem is that the visual interface fails to provide sufficient support for the dynamic data analysis process

The path of data exploration and analysis is unpredictable, and the existing tools cannot implement cyclic and iterative reasoning to continuously improve some assumptions

Images