Taxonomy Trifecta - Your guide dog to the best software available

HOME

. Evaluation criteria

. Evaluations

. . Navigation

. . Editing

. . Terms

. . Relationships

. . Import/export

. . Collaboration

. . User friendliness

. Summary

. References

INF440: Introduction to Information Architecture

Assignment 3

Timothy Godfrey

Student no 11348115

This website was produced with Macromedia Dreamweaver v8

Evaluation criteria

What makes a good thesaurus builder?

Aside from general considerations concerning software such as operating system, security and price, Ganzmann (1990) outlined three basic criteria for the evaluation of thesaurus building software: term related attributes and the degree to which relations can be managed in the software, features concerning the capture of data either through manual editing or the import of a set of thesaurus terms, and the output of terms either through a screen display or through reports.

Ganzmann’s criteria were expanded into an accompanying checklist of 11 criteria, each broken up into further subdivisions. Although many of the criteria are no longer relevant as a result of changes in technology since 1990, the import and export capacity of the software and its ability to manage terms and their relations remains a useful model for evaluation.

Morante and Walker (2003) mentioned six factors to consider in evaluating thesaurus and taxonomy building software: whether it is a “standalone” package or part of a larger content management system, whether it supports multiple users or is intended for single users, has a browser based or Windows based interface, import/export capability and formats, whether the software can track changes to a thesaurus and the cost of the software.

The US National Information Standards Association (NISA) includes in its summary of standard requirements for construction of thesauri and controlled vocabularies a list of criteria for selection of systems to manage thesauri (NISA, 2005 p 131-135). Recommended features in controlled vocabulary management systems include the capacity to sort and display terms in their alphabetical or heirarchical context and to display all the relationships (NT, BT etc) specified in the standard. The system should be able to distinguish between candidate and approved terms and to check for inconsistencies between terms, such as conflicting and duplicate references and inheritance by terms when related terms are modified or deleted. The system should also allow the cusomisation of term records to include, for example, definitions and codes assigned in other systems. Further, the system should allow truncated searching and viewing a term in the context of its relationship with other terms. It should allow reports which list terms, display term relationships and provide statistical information about the number and types of terms, and the number of modifications since a given date.

Perhaps the most comprehensive and up to date set of criteria is provided by the UK Office of the e-Government Unit (2006). Their main criteria include criteria for terms and relationships, including the number of characters in terms, terms, heirarchical levels and term relationships that can be supported by a thesaurus management system. Additional criteria include the capacity of the system to manage term notes, codes and notation and map relationships, and editing and navigation features. In addition to the criteria outlined in the guideline, an article by Darin Stewart (2007) stressed the value of providing thesauruses and controlled vocabularies in xml format given their increasing use in search engines and content management systems. For this reason the xml format has been included in import/export criteria.

The Taxonomy Dog has come up with the following features for evaluation thesaurus management software based largely on the Office of e-Government criteria. Criteria covering the multilingual capacity and cost of software were not considered.

Navigation and reporting features

Does the software:

display or produce reports in a sequence of user defined identifiers or notations
display parts of a vocabulary under a facet or node assigned by the user
produce reports and displays including terms, scope notes and relationships in alphabetical and hierarchical formats
search for and display all instances of a given term including the option for truncated searching
browse from a term to any of its connected terms

Editing features

Does the software:

support standard word processing features for fields in term records
display a term's complete hierarchical context while editing a term
move a term from one position in the hierarchy at the same time updating the modified relationships that result
navigate to and select a related term from the edit screen without having to key in the name of the term
seek confirmation before deleting a term

Term management features

Does the software:

allow an unlimited number of thesauruses, hierarchical levels and terms
allow an unlimited number and range of characters in term names and scope notes
automatically assign a unique, preferably user defined notation to each new term
allow the user to define custom numeric, date or alphabetical fields
allow the inclusion of markers to related terms mentioned in scope notes

Term relationship management features

Does the software:

support an unlimited number of standard thesaurus relationships for one term: BT/NT, RT and UF
automatically reject duplicate or inadmissable term relationships
automatically enter reciprocal relationships when a new relationship is assigned to a term
set up a user defined reciprocal relationship
map terms to terms borrowed or adapted from another thesaurus and edit mapped relationships in bulk

Import/export features

Does the software:

import datasets from exisiting vocabularies, including terms, scope notes and relationships
export data to xml format
export terms, scope notes and relationships, including sub sets of a vocabulary such as non preferred terms or terms deleted after a certain date

Collaboration features

Does the software: