This repository contains the source code and other artificts for "ProProv: A Language and Graphical Tool for Specifying Data Provenance Policies", published in the proceedings of The Fourth IEEE International Conference on Trust, Privacy and Security in Intelligent Systems, and Applications.
The following folders are included:
-
additional-images-and-tables contains figures and tables omitted from the paper to meet the submission page limit
- These are described at the end of this readme
-
experiment-graphs contains all of the information about the experiment tasks, including the task prompt, test cases, and solutions
-
experiment-website-policy-only contains a simplified version of the experiment website without the CodeIgniter framework that only includes the policy tasks, allowing for easy testing of the tasks (for example, questions such as the exit survey are ommitted and data collection is not performed).
- Can be run simply by hosting on a webserver with PHP (i.e., a local Apache server) and opening the rego or proprov pages (for example, run task 2 with proprov by opening
localhost/proprov.php?task=2)
- Can be run simply by hosting on a webserver with PHP (i.e., a local Apache server) and opening the rego or proprov pages (for example, run task 2 with proprov by opening
-
experiment-website-source contains the CodeIgniter source code for the experiment website as students viewed it, including all sections and data collection
- Requires the usual CodeIgniter framework installation and database migrations
-
proprov-standalone-gui contains the original ProProv GUI that was to be used for the experiment before it was integrated directly into the experiment website
-
proprov-website-integration contains the ProProv backend that is used to evaulate ProProv policies with the experiment website
- The source code for the
experiments-proprov.jarfiles found in the experiment-website-policy-only and experiment-website-source folders
- The source code for the
-
training-slides contains the training slides presented in the training videos
-
training-videos contains the training videos in mp4 and webm format, with vtt formatted captions
The following files are included:
-
exit-survey.md contains the exit survey questions
-
experiment-scenario.md contains the experiment scenario as provided to participants
-
consent-form.pdf is the consent form participants signed to participate in the study
Additional Table 1: The meaning of each provenance edge relation and their expected source and destination node types. The possible node types are entity, activity, agent.
| Relation | Source Node Type | Destination Node Type | Meaning |
|---|---|---|---|
| wasAttributedTo | entity | agent | The agent is responsible for the state of the entity |
| wasDerivedFrom | entity | entity | The destination entity is dependent on the source entity |
| used | activity | entity | The entity was input to the activity |
| actedOnBehalfOf | agent | agent | The source agent delegated responsibility to the destination agent |
| wasAssociatedWith | activity | agent | The activity describing the computation was performed by the agent |
| wasGeneratedBy | entity | activity | The entity is the result of the activities computation |
| wasInformedBy | activity | activity | The source activity input data generated by the destination activity |
 |
|---|
| Additional Figure 1: Illustration of the W3C Povenance Data Model |
 |
|---|
| Additional Figure 2: Illustration of the ACDC Povenance Data Model |
| Step-by-step illustration of the selection process for the policy example in Section 3 |
|---|
 |
| Additional Figure 3: Selection of an existential (exists) policy after universal policy |
 |
| Additional Figure 4: Selection of an edge wasAttributedTo policy after existential policy |
 |
| Additional Figure 5: Selection of nodes for edge policy |
 |
|---|
| Additional Figure 6: Rego Interface Main Window with the solution for Task 1 |