?
An Approach to Developing Data Visualization Tools Based on Domain Specific Modeling
An approach to the development of data visualization tools is described that provides the ability to customize to the needs of users and the specifics of the domains in which they work, based on domain-specific modeling. The results of the analysis of data visualization tools and the possibility of customizing them to subject area based on the needs of users and the tasks they solve are briefly presented. It is shown that existing tools require programming skills from users to customize the data visualization format or to develop new visualization models. It is proposed to use domain-specific modeling tools (language toolkits) to create domain-specific languages (DSL) designed to develop new data visualization models that reflect the specifics of the tasks solved by users. The use of the developed tools does not require users to have professional knowledge of programming languages. The architecture of a knowledge-driven software system is described. The core of the system is a multifaceted ontology, which includes descriptions of languages and domains, as well as rules for generating new languages and transforming constructed models. Languages are designed to describe different classes of diagrams. The system includes tools for automating the creation of new DSLs by mapping the domain ontology onto the metamodel of the base language according to user-specified rules, which are also stored in the ontology. The classification of different types of diagrams forms the basis for creating an ontology of data visualization languages. An example of a basic language for creating diagrams is described. The ability to customize the DSL and transform visualization models developed with its help and generate code that implements the model is demonstrated. It is shown how interactive visualization can be created on the basis of the developed metamodels, extending the possibilities of visualizing large structured data.