A Meta-Model for Automated Black-Box Testing of Visualization Based Software Applications

Abstract

Visualization of data is a new field that is increasingly gaining popularity in both scholastic and modern conditions. The need of data visualization in market is increasing day by day for the management of big companies which have massive data related to their business operations. For effective and properly analyzed data visualizations involved the formulas to calculate the results or for analyzing and filtering the data on any required basis. The more data, formulas and complex information is involved in the creation of a visualization, the more it is difficult to test or evaluate the correctness of that visualization for and software tester. It is difficult for testers to look into different modules for every single test, as it is a time consuming process. In past, most of the research work is done only on creation of test cases for visuals or interactive diagrams (Visualizations). Therefore, in this paper a hybrid approach (combination of both manual and automation testing) is proposed to provide automated black-box testing of Visualizations on the basis of configurations (settings), rules (business logics), used library, transformation and information provided for the transformation of data into visuals. The output of proposed approach provides results of testing for each test-case as a report. For the validation of this approach, a case study of Supply Chain domain problem is taken into consideration.