Universe of Knowledge: Nature and Modes of Formation of Subjects
The Universe of Knowledge is the totality of all subjects, ever expanding through the continuous formation of new subjects. Ranganathan identified several modes by which new subjects form, including dissection (splitting a broad subject into narrower parts), lamination (combining common elements across subjects), and loose assemblage (subjects that combine without deep fusion, as in interdisciplinary fields).
Classification: Theories, Canons and Principles
Library classification arranges knowledge (and the documents embodying it) into a logical, helpful sequence. Ranganathan's Canons of Classification guide the choice and arrangement of characteristics (e.g., canon of relevance, canon of ascertainability), and his Postulates and Principles (like the Postulate of five fundamental categories: Personality, Matter, Energy, Space, Time — PMEST) underpin faceted classification.
| Scheme | Full name / Nature |
|---|---|
| DDC | Dewey Decimal Classification — enumerative, hierarchical, notation in Arabic numerals, most widely used worldwide |
| UDC | Universal Decimal Classification — based on DDC but adds auxiliary tables and synthetic notation for greater specificity |
| CC | Colon Classification (Ranganathan) — a fully faceted/analytico-synthetic scheme using the PMEST formula, notation includes the colon (:) as a facet separator |
Modern Knowledge Organisation Systems
- SKOS (Simple Knowledge Organization System) — a W3C standard for representing thesauri, taxonomies, and classification schemes in a machine-readable, linked-data-friendly format.
- Taxonomies — hierarchical classification of concepts, typically simpler and less faceted than full classification schemes.
- Folksonomy — a bottom-up, user-generated system of tagging content, without a controlled vocabulary (e.g., hashtags).
Cataloguing: Canons, Principles and Codes
Cataloguing principles (e.g., the user's convenience, uniformity, and consistent entry) guide how bibliographic records are constructed and arranged, whether in a centralised catalogue (one agency catalogues for many libraries) or a cooperative catalogue (multiple libraries jointly contribute records, as in shared/union catalogues).
| Code | Description |
|---|---|
| CCC | Classified Catalogue Code (Ranganathan) — an Indian cataloguing code built around faceted classification principles |
| AACR-II | Anglo-American Cataloguing Rules, 2nd edition — the internationally dominant cataloguing code before RDA, structured around ISBD-based description |
Standards of Bibliographic Record Formats and Description
| Standard | Purpose |
|---|---|
| ISBD | International Standard Bibliographic Description — standardises the punctuation and order of elements in a bibliographic description |
| MARC 21 | MAchine-Readable Cataloguing — the dominant format for encoding bibliographic records for computer systems |
| CCF | Common Communication Format — UNESCO format for exchanging bibliographic data between systems |
| RDA | Resource Description and Access — successor to AACR-II, built on FRBR's conceptual model, designed for the digital environment |
| FRBR | Functional Requirements for Bibliographic Records — a conceptual entity-relationship model (Work → Expression → Manifestation → Item) |
| BIBFRAME | Bibliographic Framework — the Library of Congress's linked-data-based replacement initiative for MARC |
Standards for Bibliographic Information Interchange
- ISO 2709 — the structural standard defining how MARC records are physically formatted for exchange between systems.
- Z39.50 — a protocol enabling one system to search and retrieve records from another (cross-catalogue search).
- Z39.71 — a related NISO standard concerning holdings information format.
Metadata Standards
| Standard | Use case |
|---|---|
| Dublin Core | A simple, 15-element generic metadata set for describing any digital resource |
| MARC21 | Rich, detailed bibliographic metadata for library catalogues |
| METS | Metadata Encoding and Transmission Standard — wraps descriptive, administrative, and structural metadata for complex digital objects |
| MODS | Metadata Object Description Schema — a MARC-compatible XML schema, richer than Dublin Core |
| EAD | Encoded Archival Description — standard for encoding archival finding aids |
Indexing Systems, Techniques and Vocabulary Control
| Indexing type | Description |
|---|---|
| Assigned – Pre-coordinate | Index terms are combined by the indexer at the time of indexing into a fixed heading string (e.g., chain indexing) |
| Post-coordinate | Terms are assigned individually and combined by the searcher at the time of search (as in most database keyword systems) |
| Derived – Title-based | Index terms are extracted directly from the document's title (e.g., KWIC, KWOC) |
Vocabulary control (via thesauri and authority files) ensures consistency by mapping synonyms and related terms to a single preferred term, improving retrieval precision and recall.
Abstracting: Types and Guidelines
| Type | Description |
|---|---|
| Indicative | Describes what the document covers, without giving results/conclusions |
| Informative | Summarises the actual findings, methods, and conclusions |
| Critical/Evaluative | Adds the abstractor's assessment of the document's value or quality |
Good abstracting guidelines call for conciseness, accuracy, objectivity, and inclusion of key terms that aid retrieval.
Information Retrieval System: Features, Components, Models, Evaluation
An IR system's core components are the document collection, an indexing subsystem, a query subsystem, and a matching/ranking mechanism. Common retrieval models include the Boolean model (exact match using AND/OR/NOT), the Vector Space Model (documents and queries as weighted term vectors, ranked by similarity), and the Probabilistic model (ranks by estimated probability of relevance).
Precision = relevant retrieved ÷ total retrieved. Recall = relevant retrieved ÷ total relevant in the collection. The two typically trade off against each other.