1. USING METADATA TO FACILITATE UNDERSTANDING AND CERTIFICATION ABOUT THE PRESERVATION PROPERTIES OF A PRESERVATION SYSTEM Jewel H. Ward, Hao Xu, Mike C. Conway, Terrell G. Russell, and Antoine de Torcy CRADLE Seminar, 20 September 2013 University of North Carolina at Chapel Hill School of Information and Library Science

2. I. Introduction II. Background III. Method (Mapping a Data Grid to the OAIS Functional Model) IV. Conclusion V. Future Work

3. Introduction Need to provide internal audit mechanisms to verify assertions about preservation capabilities OAIS RM  TRAC  TDR Certification  Certifying the Certifiers This project grew out of Moore’s work, “Towards a Theory of Digital Preservation” & work for SAA in 2010: “A preservation environment manages communication from the past while communicating with the future. Information generated in the past is sent into the future by the current preservation environment. The proof that the preservation environment preserves authenticity and integrity while performing the communication constitutes a theory of digital preservation. We examine the representation information that is needed about the preservation environment for a theory of digital preservation. The representation information includes descriptions of the preservation management policies, the preservation processes, and the state information that is needed to verify the correct working behavior of the system.” IJDC, 2008, Vol. 3, No. 1, pp. 63-75

4. Introduction We propose to do this by: Metadata-as-a-preservation task: mapping computer functions and procedures to the OAIS model and ISO 16363 (TRAC) Metadata-as-state: use state transition systems to connect computer code & data management policies Our method bridges the gap between computer code & human-readable policies Facilitates certifying preservation repository properties Not a funded project per se. Melding of dissertation (Jewel Ward), Master’s (Mike Conway), and post-Doc (Hao Xu) work.

5. Background Metadata Bibliographic, administrative IP numbers, date, time Event data, state The OAIS Composite of Functional Entities Common Services, Administration, Ingest, Data Management, Archival Storage, Preservation Planning, Access Human- vs Machine-readable Policies “The repository shall comply with Access Policies” (ISO 16363) Vs. acAclPolicy { msiAclPolicy(“STRICT”); } #computer code “Bottom Up” vs. “Top Down”

6. Background

10. Method 1

12. Method 2 State transition system captures State of data objects Their transitions Example: SIP  AIP  DIP Data management policies determine which sequences of states are compliant or not These refinements yield a formal structure for data management policies The state transition system thus obtained specifies which metadata are necessary if the preservation properties are going to be certified.

13. Method 2

14. Method 3 Data system consists of: Actions change states Events that trigger rules Rules that consist of actions Need to ID all states and actions, and the non- determinism What events trigger rules? Need a set of metadata attributes == states of the state transition system

15. Method 3 Mapping States State Transitions Non-determinism Interpreting the components of the OAIS Reference Model as a state transition system allows us to map the high level recommendations to machine code in a rigorous manner State transitions provide log information; log information == metadata; these attributes are stored in the metadata catalog

16. Method 4 If not a 1:1 mapping, need additional metadata Example: SIP  AIP; AIP  AIP; AIP  DIP Use the OAIS Composite of Functional Entities to create “policy domains” in the preservation system. E.g., the OAIS Reference Model as a “state transition system”. State transitions provide logging information about each transition that may be preserved as metadata Example, can store whether or not an object is part of a SIP or an AIP as a metadata attribute attached to that object No change = compliant Change = not compliant

17. Method 4 Attach metadata for each area of the data grid Similar to the OAIS Composite of Functional Entities Combined with Customized Policies = a “policy domain”

18. Method 4

19. Conclusion 1 “Basing policy domains on [the OAIS Reference Model Composite of Functional Entities] enables the development of assessment criteria that evaluate the recorded state of each object and allow automated validation of the properties such as authenticity, integrity, chain of custody, and trustworthiness. Our approach makes the assessment of the preservation state of the archive amenable to formal, automated validation via state transition events metadata.”

20. Summary We examined the “state of the state” We detailed a method for abstracting state transition systems from data management policies “…this approach bridged the gap between concrete computer code and human-readable, abstract standards such as the OAIS Reference Model Functional Model by using metadata. We posit that this method may be applied to any preservation system in order to facilitate the certification of a trustworthy digital repository. “

21. Future Work I finish my dissertation, which examines the “state of the state”. Capture how information objects in the OAIS RM may be transformed, combined, or divided Capture the static and dynamic aspects of the OAIS RM Map iRODS policy enforcement points to the OAIS RM semantics (static vs. dynamic) Classify the policies required for repositories to create a library of policies by domain or function Define a formalism to verify these states/transitions/actions

22. Acknowledgements This research is partially supported by NSF grant #0940841 “DataNet Federation Consortium” and NSF grant #1032732 “SDCI Data Improvement: Improvement and Sustainability of iRODS Data Grid Software for Multi-Disciplinary Community Driv- en Application”.

23. Questions? Comments?