Low-Modeling of Software Systems
Jordi Cabot
TL;DR
The paper tackles the escalating complexity of modern software systems by introducing low-modeling as a set of strategies to rapidly bootstrap and refine high-level specifications with reduced hand-modeling. It categorizes approaches into heuristic-based generation, knowledge-based enrichment, and ML-based inference, detailing how each can seed initial models and be refined by experts. Through a case study on conversational interfaces for open tabular data, it demonstrates end-to-end automation from data analysis to a deployed chatbot, illustrating the data-to-conversation pipeline. The BESSER platform is proposed as an open-source implementation to scale low-modeling across smart front-ends and back-ends, emphasizing cross-component consistency. Overall, the work positions low-modeling as a means to democratize software development, reduce boilerplate, and adapt systems to uncertain, evolving environments.
Abstract
There is a growing need for better development methods and tools to keep up with the increasing complexity of new software systems. New types of user interfaces, the need for intelligent components, sustainability concerns, ... bring new challenges that we need to handle. In the last years, model-driven engineering has been key to improving the quality and productivity of software development, but models themselves are becoming increasingly complex to specify and manage. In this paper, we present the concept of low-modeling as a solution to enhance current model-driven engineering techniques and get them ready for this new generation of software systems.