MuSimA: Multi-Modal Simulator for ABAC Systems - Output Bundle
===========================================================

This folder contains the generated datasets and configurations for Attribute-Based Access Control (ABAC).

Files Overview:
--------------

1. output.json
   - A complete JSON representation of the generated ABAC system.
   - Contains:
     - "S", "O", "E": Lists of Subject, Object, and Environment identifiers.
     - "SA", "OA", "EA": Lists of attribute names for Subjects, Objects, and Environments.
     - "SAV", "OAV", "EAV": Dictionaries mapping attribute names to their possible values.
     - "SV", "OV", "EV": Dictionaries mapping specific entities (e.g., "S_1") to their assigned attribute values.
     - "rules": A list of generated ABAC policy rules. Each rule specifies a condition that grants access.
       Example: "SA_1 = SA_1_1, OA_2 = OA_2_3" means access is allowed if Subject Attribute 1 is SA_1_1 AND Object Attribute 2 is OA_2_3.

2. ACM.txt (Access Control Matrix)
   - A 3D matrix representation of access decisions.
   - Format: Space-separated 0s and 1s.
   - 1 = Access Allowed, 0 = Access Denied.
   - The matrix dimensions correspond to Subjects x Objects x Environments.

3. access_data.txt
   - A flattened dataset of access logs, useful for machine learning tasks.
   - Each line represents a single access request.
   - Format: [Subject Attributes] [Object Attributes] [Environment Attributes] [Decision]
     - The first set of columns are Subject Attribute values.
     - The next set are Object Attribute values.
     - The next set are Environment Attribute values.
     - The last column is the decision (0 or 1).

4. distribution_attestations/plots/ (Directory)
   - Visualizations of the attribute distributions.
   - Shows "Expected" counts (based on the configured distribution) vs "Actual" counts (realized in the generated data).

5. distribution_attestations/comparisons/ (Directory - Multimodal Input Only)
   - Side-by-side images comparing the original hand-drawn input sketches with the realized distribution plots.
   - Useful for verifying that the generated data matches the visual intent.

Generated by MuSimA.
