Cataloging and Classification/Metadata

Information resources have properties that are useful for cataloguing which are referred to as "elements" or "attributes".

Typical attributes about an information resource are:

Author (who created it).
Title (what is its title).
Subject (what is it about).

All such attributes are collectively referred to as metadata.

A very large number of elements or attributes of books could be collected as metadata. The usual metadata attributes for a book would be title, author, publisher, publication date, ISBN, index words, subject, and any other relevant information that readers may need to locate the book. In addition, there would be metadata collected which would meet the administrative needs of the library, for example, the cost of the book, whom it was bought from, and where the funds came from. If the publication details are important, then edition, publishing house, whether it was part of a series, date of original publication, original publisher, so on and so forth.

Maintaining the metadata of the collection of information resources held in a library or any other institution is a basic task of cataloguing. It is vital that the metadata must be complete, accurate and informative.

Choosing Metadata

It must be understood that metadata are used not just in the bibliography of books and information resources, but in every information database, depending upon what data objects are being saved in the information database, and what aspects of their objects are important.

For example, a database of famous paintings would choose to have fields not only such such as creator, date/year of creation, title, material used (paper, canvas, cloth), drawing or coloring material used (charcoal sketch, gouache, oil etc.), and monetary value, but also school of painting (cubism, impressionism, etc.), provenance, location, owner, copyright holder and so on. So many of these would not be applicable to books or to other information resources stored in a library.

Similarly, a collection of fossils would have location of excavation, strata in which it was found, age of strata, estimated age of fossil, method of estimating that age, condition, taxonomic identification and person doing the identification, present location (this Museum or elsewhere), storage location (cupboard, shelf and box information), whether it is a type specimen, whether described in a research paper, and if so, which one, and so on and so forth.

As is obvious, metadata is of various kinds and the exact level of detail would depend upon the purpose of the institution. For example, while all institutions would wish to know the author, title and subject of a work, all institutions would not be interested in every possible element of a book such as location of the setting of the book, name of illustrators (if any), time period, and so on.

So they choose the kinds of Metadata which they want and organise them into the structure which is called as a record.

Organising metadata into records

In the 21st century, metadata typically refers to digital forms, but traditional card catalogs contain metadata, with cards holding information about books in a library (author, title, subject, etc.).

For small collections, one can choose an excel sheet or a small flat-file database and create a structure for the record. In a spreadsheet such as Excel or Calc, one can do so by selecting columns, the providing titles in the topmost row, and deciding the format that the data in that column would be stored under, and by assigning that format to all the cells in the column, less the column title. Likewise, in a database, one would select a number of fields, and decide the format that the data would be stored in that field for all the records. This approach however is simplistic, small-scale and unable to scale up, with very limited features and not at all suitable for modern libraries

However, most libraries all over the world normally use computerised online catalogs/library management software, either custom-made, proprietary or open source. These would have a wide variety of features, such as ability to work on networks or the web, use them for information interchange and also to interact with each other. To do this, computerized library catalogs and library management software need to use industry-wide standards; at present four standards are closely involved—MARC, AACR2/RDA and ISDB.

The structure of the bibliographic catalog records are almost universally as per the industry-wide standard called MARC (machine-readable cataloging), so that bibliographic information can be shared freely between computers. MARC standards are a set of digital formats for the description of information resources catalogued by libraries, such as books.

Other standards work in conjunction with MARC, for example, Anglo-American Cataloguing Rules (AACR)/Resource Description and Access (RDA) provide guidelines on formulating bibliographic data into the MARC record structure, while the International Standard Bibliographic Description (ISBD) provides guidelines for displaying MARC records in a standard, human-readable form.

Filling in the values of Metadata

Once an institution has decided the structure of a record (about an information resource) and which metadata elements it wants to save, this is usually implemented in some variety of software catalog, such as Koha.