From Athenawiki

Jump to: navigation, search

Contents

A5. ORGANISE YOUR TERMS INTO A THESAURUS STRUCTURE

Error: must specify an image in the first line

Actions:

Structure your terms set by:

  • 1. Creating one micro-thesaurus (list of terms) for each domain or sub-domain you have defined at step A1: Define your domains.
  • 2. Gathering terms from these micro-thesauri out of the constraint of domain belongingness, but in a transversal or cross-domain approach (thematic approach)
  • 3. Making relations between the lists of terms according to the model of a network of micro-thesauri: "narrower", "broader", and "associated to"


Purpose:

The objective is to transform your simple lists of terms into a thesaurus, that is, a structured network of micro-thesauri. Indeed, among all existing types of terminology resources, we recommend the use of thesauri for museums in order to make their collections available to Europeana. First, this type of terminology is quite easy to SKOSify as the SKOS format is intended -- even mainly designed -- to handle thesauri, so that it can technically comply with the main requirement of Europeana ingestion process. Then, if we compare with the other terminology types, thesaurus features a good mix of richness and usability. Moreover, museums are generally already using this kind of terminology rather than ontology or classifications to describe their collections in a well-structured manner (47% of the 149 terminologies we have listed in our initial survey are thesaurus). Without forgetting that regular relations of equivalence and association are particularly relevant for multilinguality. So, even if Europeana is moving toward the ontologies, we keep recommending thesauri to the museums since it appears as a good "middle way" between controlled vocabulary (particularly appreciated in the museums) and ontologies (especially powerful for the retrieval).

A thesaurus can be defined as "a networked collection of controlled vocabulary terms". Thesauri allow the connection of terms using several types of relationships which can be hierarchical, associative, equivalence or definition. This means that a thesaurus uses associative relationships in addition to parent-child relationships. A parent-child relationship is expressed by a Broader Term (BT) /Narrower Term (NT) feature. Associative relationships in a thesaurus such as "Related Term" (RT) (e.g. term A is related to term B) are used to express relationships that are neither hierarchical nor equivalent. Equivalence is expressed by the USE (e.g. preferred term)/ Used For (UF) (e.g. non-preferred term). Additional information such as definition or remark can be included in a Scope Note (SN). The equivalence relationship is especially useful within multilingual thesauri.

Since the structure of a thesaurus is carried by links of hierarchy and of association between micro-thesauri, we recommend you to multiply the links in order to improve your terminology exploitation later. The more you have relations between your lists of terms, the more efficiently a search engine will return results to queries. In other words, we advice you to benefit from the thesaurus structure the most you can. If you respect both the hierarchical and the transversal approaches of the Thesaurus structure, later you will SKOSify your terminology quite easily. SKOS will allow you to manage an arborescence of lists of terms related to different concepts (ConceptScheme tree), and a non-hierarchical set of non-exclusive groups of terms coming from these lists (Collectionsfor thematic grouping).

You can consider that this step is over when there is no more term out of a list, and no more list of terms totally independent of the others.


For example

You have a set of terms describing objects of architecture. Among all the descriptors in your thesaurus you have "monument" and "habitation". For each of them you create one specific micro-thesaurus. In the micro-thesaurus about monuments you have several terms like "palace" which is a type of monument. And in the micro-thesaurus about habitations you have several terms like "apartment" which is a type of habitation. So terms like "monument", "habitation", "apartment" and "palace" are linked by hierarchical relations. But you can also consider that a transversal group of terms around the theme of "building" can be proposed. In this group you could have "palace" and "apartment" which do not belong to the same hierarchy of terms but can though be connected in a thematic group of terms.

 

h2>Methods and tools:</h2>

At the publication date of this deliverable, there is no specific tool guiding the logical design of your thesaurus. So if your vocabulary management tool does not provide this help, you will need to organise it with your daily tools (spreadsheets, ...).

You can have a look on the GEMET, General Multilingual Environmental Thesaurus, which propose you different ways to browse the thesaurus: thematic, alphabetic or hierarchical listings.


Navigation

We invite you to pursue the step by step process by going to the next step: A6: Find equivalent terms in other languages.

Error: must specify an image in the first line

The different tasks we detail are:

  • A1: Define your collection domain(s)
  • A2: Identify your users' expectations about your semantic descriptions
  • A3: Define your connection with the datamodel
  • A4: Choose the terms for the semantic description of your digital resources
  • A5: Organise your terms into a thesaurus structure
  • A6: Find equivalent terms in other languages
  • A7: Implement your thesaurus

You can also navigate through the recommendations by using the synoptic map below. This map will be available on each page of the recommendations process. In order to know the name of a step in particular, just rollover and stay a bit on the very box so that the name appears.

Error: must specify an image in the first line

This page was last modified on 3 June 2011, at 09:53.This page has been accessed 3,067 times.