Preservation and Archiving Policy
Insights in Biology and Medicine (IBM) maintains a multi-layered digital preservation and archiving program that safeguards the Version of Record (VoR), associated metadata, supplementary materials, and machine-readable renditions (e.g., JATS XML) across independent, geographically distributed systems. Preservation is aligned with community best practices for scholarly publishing, enabling verification, citation, and reuse over decades.
- Scope Articles, issues, supplements, editorials, corrections, retractions, datasets (where hosted), and supplementary files.
- Formats PDF/PDF-A, HTML, XML/JATS, image assets (SVG/PNG/JPEG), audio/video, and ancillary data packages.
- Redundancy At least three preserved copies in distinct networks with regular fixity checks and integrity reporting.
Preservation Goals
- Integrity: Ensure bit-level and content-level integrity through checksums, version control, and documented change logs.
- Authenticity: Maintain provenance and unbroken DOI resolution to the canonical landing page.
- Accessibility: Provide reliable access with reasonable accommodations (accessible PDFs/HTML, alternative text, captions).
- Interoperability: Offer standards-based exports (JATS, Crossref, DOAJ XML, OAI-PMH) for libraries and aggregators.
- Continuity: Guarantee content availability via independent preservation networks in case of platform changes or cessation.
Preservation Systems & Layers
Articles are hosted on the journal platform with routine backups and off-site replicas. Storage utilizes RAID redundancy and snapshotting with scheduled exports of PDFs, HTML, and JATS XML to cold storage.
- Nightly database and file-store backups retained with a tiered retention policy (daily/weekly/monthly).
- Immutable “vault” snapshots to protect against accidental deletion and ransomware.
- Automated DOI link checking and landing-page audits to validate persistence.
IBM participates in community archiving networks (e.g., LOCKSS/CLOCKSS–style programs and PKP PN equivalents) to ensure long-term survivability independent of the publisher’s infrastructure.
- Distributed, library-operated nodes with consensus repair using peer-to-peer polling.
- Triggered access policies that make preserved content accessible if the journal becomes unavailable.
- Regular content audits and fixity verification.
Where applicable, VoR or AAM copies are deposited in trusted repositories and index-linked services to increase resilience and discovery.
- Institutional and subject repositories with DOI-based canonical linking (see Repository Policy).
- Abstracting and indexing services that maintain metadata mirrors and archival thumbnails.
- Open metadata exposure via OAI-PMH and Crossref deposits for reference continuity.
Datasets and code associated with articles should be preserved in domain-relevant repositories (e.g., Dryad, Zenodo, SRA, GEO, PRIDE), with persistent identifiers cited in the article’s Data Availability statement.
- Preferred: repository-assigned DOIs/accessions; immutable releases for code.
- Checksums and version labels included; metadata aligns with FAIR principles.
- Licensing consistent with dataset terms (e.g., CC0 for metadata, CC BY or ODC for data).
Preservation Formats & Normalization
IBM maintains the VoR in the original publisher formats and, where feasible, normalized archival variants for long-term readability.
| Content Type | Preferred Format(s) | Archival Variant | Notes |
|---|---|---|---|
| Article text | HTML, PDF | PDF/A-2b; JATS XML | Semantic tagging in JATS for machine readability. |
| Figures | PNG/JPEG | TIFF (for raster), SVG (for vector) | ≥300 dpi for raster; embed fonts for vector. |
| Tables | HTML | JATS XML tables | Preserve captions and footnotes. |
| Supplementary files | PDF/CSV/ZIP | Checksumed ZIP; README.txt | Document file origins and software versions. |
| Audio/Video | MP3/MP4 (H.264) | Lossless masters where supplied | Include captions/transcripts where available. |
| Code | Repository release (ZIP/TAR) | Immutable release + DOI | Dependency manifest preserved. |
Disaster Recovery & Continuity
Our disaster-recovery plan ensures continuity of access and editorial operations in adverse events (infrastructure failures, data loss, legal transitions, force majeure).
- RPO/RTO: Recovery Point Objective ≤ 24 hours; Recovery Time Objective ≤ 72 hours for read access to VoR.
- Failover: Read-only mirror service to expose preserved VoR and metadata while primary systems recover.
- Runbooks: Documented steps for DNS failover, cache invalidation, and DOI update verification.
- Incident Logs: Post-incident reports shared with stakeholders and preservation partners.
Corrections, Retractions & Takedowns
IBM follows the principle of preserving the scholarly record while transparently signaling amendments.
- Corrections: Linked from the VoR with bidirectional metadata; preserved as separate citable items.
- Retractions: VoR retained with clear watermark and explanatory notice; metadata flags (retraction) propagated to Crossref and repositories.
- Takedowns: In rare legal/ethical cases, access may be restricted; tombstone pages remain with reason codes and DOI continuity.
Roles & Responsibilities
- Maintain preservation SLAs and audit schedules.
- Deposit Crossref metadata, updates, and relationships (corrections/retractions).
- Coordinate with preservation networks and repositories for content verification.
- Monitor DOI resolution, OAI-PMH exposure, and JSON-LD correctness on article pages.
- Provide final files in preferred archival formats where possible.
- Use recognized data/code repositories with persistent identifiers.
- Ensure third-party permissions are documented for long-term availability.
- Harvest metadata via OAI-PMH; maintain DOI backlinks to the canonical VoR.
- Honor license terms; expose dc.rights.uri and licenseRef fields.
- Propagate amendments to maintain record integrity.
Security & Integrity Controls
- Checksums & Manifests: Per-object checksums; release manifests signed by the system.
- Access Controls: Principle of least privilege; immutable logs for preservation actions.
- Versioning: Every file replacement generates a new version with diff notes; prior versions preserved.
- Monitoring: Uptime, DOI resolve rates, OAI-PMH response health, and fixity dashboards.
Metadata, Identifiers & Interoperability
Rich metadata underpins preservation. IBM publishes machine-readable metadata on article pages and through aggregators.
- Identifiers: DOI for articles and components; ORCID for contributors; funder IDs; grant numbers.
- Exchanges: Crossref deposits (relations for updates), DOAJ export, OAI-PMH (oai_dc, jats), and JSON-LD/Schema.org on the VoR.
- Rights: CC BY 4.0 license exposed in machine-readable fields; third-party exceptions labeled.
Testing, Audits & Reporting
IBM runs scheduled tests to validate recoverability and content completeness.
- Quarterly restore drills from off-site snapshots to a sandbox environment.
- Monthly fixity verification reports with exception handling and remediation notes.
- Annual preservation statement with metrics: object counts, failed fixity events, restore test outcomes.
| Metric | Target | Status Signal |
|---|---|---|
| DOI resolution uptime | ≥ 99.95% | OK if ≥ target; investigate if ↓ |
| Fixity mismatches per month | 0 | Any > 0 triggers incident workflow |
| Restore drill success | 100% | Report steps and timings |
Accessibility & Inclusive Preservation
Preserved content aims to meet accessibility expectations over time. PDFs include tagged structures and bookmarks; HTML includes ARIA roles, alt text for figures, and captions/transcripts for media when provided. Preservation packages retain these features to avoid regressions during migrations.
Policy Updates & Review Cycle
This policy is reviewed annually or upon significant platform changes. Amendments are versioned and time-stamped; historic policy versions remain accessible for auditability.
Frequently Asked Questions
What happens if the journal changes platforms?
Content and metadata are migrated with preservation copies validated against checksums; DOIs continue to resolve to the canonical pages.
Can authors retain personal copies?
Yes. Authors are encouraged to keep personal copies and deposit AAM/VoR in institutional repositories per our Repository Policy.
How are data supplements preserved?
Where hosted by IBM, supplements are packaged with README, checksums, and versioning; otherwise, we reference the external repository DOI/accession.