?
Преодоление ограничений современных инструментов визуализации данных: новая методология
An analysis of existing data visualization methods and tools was conducted, revealing their limitations associated with the need for advanced programming skills or insufficient configuration flexibility. A methodology for developing data visualization tools based on expert knowledge is proposed, enabling the effective creation of customizable visual representations. The methodology comprises three key approaches: (1) the development of Domain-Specific Languages (DSLs) for data visualization; (2) the automated generation of visualization scripts using Python libraries and large language models (LLMs); and (3) the extension of existing visualization libraries with packages that incorporate custom elements. It is expected that the proposed methodology will improve the accessibility and efficiency of data visualization tools by reducing programming requirements while maintaining high accuracy in data representation across various domains.
