Ugrás a fő tartalomhoz

OMI-SPEC-001: Document Model

Status​

Draft

Version: 0.1

Purpose

The Open Manuscript Initiative (OMI) defines a semantic document model designed specifically for scholarly communication.

Unlike traditional word processors, OMI separates scholarly meaning from visual presentation. Authors describe what information represents rather than how it should appear.

The document model serves as the common foundation for authoring, peer review, publishing, preservation, and machine processing.

Design Principles

The document model is built upon the following principles:

  • Semantic first
  • Platform independent
  • Human readable
  • Machine processable
  • Extensible
  • Interoperable
  • Accessible
  • Long-term preservable

Layered Architecture

The OMI document model consists of multiple layers.

┌──────────────────────────────────────────────┐
│               Discipline Profiles            │
├──────────────────────────────────────────────┤
│               Scholarly Objects              │
├──────────────────────────────────────────────┤
│              Core Document Model             │
└──────────────────────────────────────────────┘

The Core Document Model remains stable across all disciplines.

Higher layers extend, but never replace, the core specification.

Core Document Model

Every OMI manuscript contains a common set of semantic objects.

Metadata​

  • Title
  • Subtitle
  • Authors
  • Affiliations
  • ORCID
  • Abstract
  • Keywords
  • Language
  • License
  • Funding
  • Persistent Identifiers

Document Structure​

  • Section
  • Subsection
  • Paragraph
  • Quotation
  • List
  • Figure
  • Table
  • Caption
  • Equation
  • Appendix

Scholarly Objects​

  • Citation
  • Bibliography
  • Footnote
  • Endnote
  • Cross-reference
  • Dataset
  • Software
  • Supplementary Material
  • Review Comment
  • Annotation

Discipline Profiles

Different scientific disciplines require specialised semantic objects.

OMI extends the Core Document Model through Discipline Profiles.

A profile adds new object types, validation rules, metadata, and export behaviour without modifying the underlying document model.

Examples include:

  • History
  • Theology
  • Linguistics
  • Mathematics
  • Physics
  • Chemistry
  • Biology
  • Medicine
  • Computer Science
  • Engineering
  • Economics
  • Law

Example: History Profile

The History profile introduces domain-specific semantic objects such as:

  • Archival Source
  • Historical Person
  • Historical Place
  • Historical Event
  • Manuscript Source
  • Regest
  • Critical Apparatus
  • Diplomatic Transcription

Example: Mathematics Profile

The Mathematics profile extends the model with:

  • Definition
  • Axiom
  • Theorem
  • Lemma
  • Corollary
  • Proof
  • Formula
  • Symbol Registry

Example: Physics Profile

Additional semantic objects include:

  • Equation
  • Physical Constant
  • Measurement
  • Unit
  • Instrument
  • Experiment
  • Simulation

Example: Chemistry Profile

Additional semantic objects include:

  • Molecule
  • Compound
  • Reaction
  • Crystal Structure
  • Spectrum
  • Chemical Formula

Example: Biology Profile

Additional semantic objects include:

  • Species
  • Taxonomic Classification
  • Gene
  • Protein
  • DNA Sequence
  • Specimen

Example: Medicine Profile

Additional semantic objects include:

  • Clinical Trial
  • Patient Cohort
  • Intervention
  • Outcome
  • Ethics Approval
  • CONSORT Checklist

Multiple Profiles

Scientific research is increasingly interdisciplinary.

An OMI manuscript may therefore use multiple Discipline Profiles simultaneously.

Example:

History
+ Computer Science
+ Linguistics

or

Biology
+ Computer Science
+ Statistics

Each profile contributes additional semantic objects while remaining fully compatible with the Core Document Model.

Plugin Architecture

Discipline Profiles are implemented as plugins.

OMI Core

    │

    ├── History Profile

    ├── Mathematics Profile

    ├── Physics Profile

    ├── Medicine Profile

    ├── Law Profile

    └── Custom Profile

Each plugin may define:

  • additional semantic objects
  • metadata
  • validation rules
  • export mappings
  • user interface extensions

without modifying the OMI Core.

Object-Oriented Model

An OMI manuscript is composed of semantic objects.

Document
│
├── Metadata
├── Sections
├── Paragraphs
├── Figures
├── Tables
├── Citations
├── Bibliography
├── Annotations
├── Review Layer
└── Discipline Objects

Each object has:

  • unique identifier
  • semantic type
  • metadata
  • relationships
  • version history

Semantic Relationships

Objects may reference one another.

Examples include:

  • citation → bibliography entry
  • figure → caption
  • theorem → proof
  • archival source → repository
  • person → ORCID
  • dataset → DOI

Relationships are explicit and machine-readable.

Extensibility

The document model is designed for long-term evolution.

Future versions may introduce additional object types without breaking existing manuscripts.

Backward compatibility is a fundamental design requirement.

Future Work

Future specifications will define:

  • Annotation Model
  • Review Model
  • Citation Model
  • Metadata Model
  • Publishing Model
  • Plugin API
  • File Format Specification

Summary

The OMI Document Model is designed to represent scholarly knowledge rather than formatted documents.

A stable Core Document Model combined with extensible Discipline Profiles enables OMI to support every scientific discipline while maintaining interoperability, portability, and long-term sustainability.