A Multi-Relational Adjacency Matrix Dataset of Java Class Dependencies

This dataset provides a structural representation of class relationships extracted from Java-based applications developed using Object-Oriented Programming (OOP) principles. The data is modeled as a graph and presented in the form of adjacency matrices. The edges within the graph represent four fundamental types of inter-class relationships: aggregation, association, composition, and dependency. Furthermore, each node (class) is enriched with a comprehensive set of features, including quantitative metrics such as method count, attribute count, visibility proportions, and Lines of Code (LOC), as well as qualitative metadata including class stereotypes and semantic features derived from identifiers and comments.