© 2007 by ContextualAnalysis, LLC Why Are Taxonomies Necessary? By Fred Leise ContextualAnalysis, LLC
May 12, 2015
© 2007 by ContextualAnalysis, LLC
Why Are Taxonomies Necessary?
By Fred Leise
ContextualAnalysis, LLC
© 2007 by ContextualAnalysis, LLC
Taxonomies are sets of terms (controlled vocabularies or CVs) used to tag documents or other content objects.
Taxonomies may also be used as browsing hierarchies or for search enhancement.
What Are Taxonomies?
© 2007 by ContextualAnalysis, LLC
Taxonomy terms are collected into groups called attributes. Each attribute (or facet) describes one property of your content.
What Are Taxonomies?
© 2007 by ContextualAnalysis, LLC
Example:
Attribute: Office Location
Terms: London
New York City (NYC, Big Apple)Washington, DC
What Are Taxonomies?
Alternate Terms
© 2007 by ContextualAnalysis, LLC
In this example, “NYC” and Big Apple” are given as variants for “New York.”
Variant terms are used to expand search queries. If a user enters “New York” the search system expands to search “New York or NYC or Big Apple.
What Are Taxonomies?
© 2007 by ContextualAnalysis, LLC
Search query expansion ensures that more relevant information is found, even though it might use terms the searcher hasn’t thought of.
What Are Taxonomies?
© 2007 by ContextualAnalysis, LLC
Other typical attributes include:
Author
Creation Date
Audience
Version Number
Subject
What Are Taxonomies?
© 2007 by ContextualAnalysis, LLC
There is an international standard for metadata, the Dublin Core Metadata Element Set, consisting of 15 attributes.
What Are Taxonomies?
© 2007 by ContextualAnalysis, LLC
Good metadata schemas (collections of attributes) will adhere as closely as possible to the Dublin Core standard.
More information is available at: www.dublincore.org
What Are Taxonomies?
© 2007 by ContextualAnalysis, LLC
Well designed taxonomies:
1. Enable users to find relevant information quickly and efficiently (improved retrieval)
2. Lead users to additional relevant information, providing upselling and cross-selling opportunities
What Are Taxonomies?
© 2007 by ContextualAnalysis, LLC
Well designed taxonomies:
3. Assists authors in consistently tagging content
What Are Taxonomies?
© 2007 by ContextualAnalysis, LLC
Proper use of taxonomies results in:
Less time wasted searching for information
Fewer failed searches
Fewer abandoned interactions
Increased income
Reduced customer assistance costs
What Are Taxonomies?
© 2007 by ContextualAnalysis, LLC
English is rich in words that mean the same or nearly the same thing
feline/cat
car/automobile
travel/journey/excursion/trip
jeans/denims/Levi's/501s
Why Are Taxonomies Important?
© 2007 by ContextualAnalysis, LLC
Result: scattering of information.
No matter what term you use in a free-text search, you get only part of the relevant information.
The rest is not retrieved because it uses different terms to describe the same concept.
Why Are Taxonomies Important?
© 2007 by ContextualAnalysis, LLC
Consider the example of mobile devices.
There are many ways that users can refer to them:
Personal digital assistants
Handheld computers
Blackberries
PDAs
Why Are Taxonomies Important?
© 2007 by ContextualAnalysis, LLC
If users don’t know the term you use to label the information they are looking for, they waste time browsing or give up their search completely.
They are victims of a communication chasm.
Why Are Taxonomies Important?
© 2007 by ContextualAnalysis, LLC
You use the term “cat.” I use “feline.” If we each search a recipe database that uses both terms with equal frequency, we will get back only half the appropriate recipes, a recall ratio of 50%
Why Are Taxonomies Important?
© 2007 by ContextualAnalysis, LLC
Solution: Add a controlled vocabulary to the search system that gives “feline” and “cat” as equivalent terms.
Search queries will be expanded appropriately.
Why Are Taxonomies Important?
© 2007 by ContextualAnalysis, LLC
English is rich in words that have more than one disparate meaning
Pitch
To throw a baseball
A tar-like substance
A salesman’s monologue
Why Are Taxonomies Important?
© 2007 by ContextualAnalysis, LLC
Bank
Where you store money
The side of a river
To carom a cue ball off a pool table rail
To prepare a fire for the night
To maneuver a plane for a turn
Why Are Taxonomies Important?
© 2007 by ContextualAnalysis, LLC
Result: Lots of false drops (irrelevant information), resulting in poor precision.
Why Are Taxonomies Important?
© 2007 by ContextualAnalysis, LLC
Solution: use a CV that includes scope notes (definitions) or that uses facets.
Example: Think about searching for the term “Rembrandt.” You might get the following results.
Why Are Taxonomies Important?
© 2007 by ContextualAnalysis, LLC
Why Are Taxonomies Important?
Rembrandt GoSearch
The painter Rembrandtwas one of the greatestof all the Dutch realists….
If you want to whitenand brighten yourteeth, there is no betterbrand than Rembrandt.
© 2007 by ContextualAnalysis, LLC
Why Are Taxonomies Important?
You probably are interested in only one of these “Rembrandts.” So half of your search results are irrelevant.
Now consider what happens if you were able to specify the type of object you are looking for, either an artist or a toothpaste brand.
© 2007 by ContextualAnalysis, LLC
Why Are Taxonomies Important?
The painter Rembrandtwas one of the greatestof all the Dutchrealists….
If you want to whitenand brighten yourteeth, there is no betterbrand than Rembrandt.
Artist
Brand Name
Rembrandt
Rembrandt
© 2007 by ContextualAnalysis, LLC
Why Are Taxonomies Important?
You get only results relevant to what you are interested in.
Here, having search boxes identified by attribute (faceted searching) lets you hone in quickly on the particular information you want.
© 2007 by ContextualAnalysis, LLC
Why Are Taxonomies Important?
You could also use one search and let users filter or narrow results after their search.
© 2007 by ContextualAnalysis, LLC
© 2007 by ContextualAnalysis, LLC
Roles for Taxonomies
Tagging documents for a content management system
Provides administrative metadata to control authoring and publishing processes
How are Taxonomies Used?
© 2007 by ContextualAnalysis, LLC
Roles for Taxonomies
Administrative metadata: example
Document # AuthorDepartment Creation datePublication date Expiration date
How are Taxonomies Used?
© 2007 by ContextualAnalysis, LLC
Roles for Taxonomies
Tagging document contents for a content management system
Provides metadata to support search
Ensures inter-indexer consistency
How are Taxonomies Used?
© 2007 by ContextualAnalysis, LLC
Roles for Taxonomies
Tagging document contents for a content management system
Controls subject scattering
Increases search results relevance: tags “aboutness” not just mentions of a word
How are Taxonomies Used?
© 2007 by ContextualAnalysis, LLC
Roles for Taxonomies
Search engine component
Translates user’s terms into those used to tag items (increases precision and recall)
Offers options for expanding or reducing scope of search using broader or narrower terms
How are Taxonomies Used?
© 2007 by ContextualAnalysis, LLC
Roles for Taxonomies
Search engine component
Differentiates between multiple meanings of terms
How are Taxonomies Used?
© 2007 by ContextualAnalysis, LLC
Taxonomy Use: Search Results
rei.com
© 2007 by ContextualAnalysis, LLC
Roles for Taxonomies
Operating as a browsing hierarchy
Organizes content using taxonomy terms as category labels
Represents taxonomy hierarchy by browsing levels
How are Taxonomies Used?
© 2007 by ContextualAnalysis, LLC
rei.com
Level 1
Level 4Level 3
Level 2
© 2007 by ContextualAnalysis, LLC
Synonym Ring
Identifies words with equivalent meanings (in a given context)
rock = stone
CD-ROM = CD = disk
money = dough = bucks = greenbacks = legal tender
Types of Taxonomies
© 2007 by ContextualAnalysis, LLC
Synonym Ring
When one of the words in a synonym ring is searched for, the search engine expands the search and returns items containing any of the words in the ring.
Types of Taxonomies
© 2007 by ContextualAnalysis, LLC
Authority File
Has all the features of a synonym ring, plus the identification of preferred terms (approved terms/descriptors/keywords) for tagging content.
Types of Taxonomies
© 2007 by ContextualAnalysis, LLC
Taxonomy
Also called hierarchy or classification.
All features of authority files, plus the broader term (BT) and narrower term (NT) relationships.
Types of Taxonomies
© 2007 by ContextualAnalysis, LLC
Taxonomy
All terms must be part of a hierarchical relationship (no orphan terms).
Taxonomies may be presented in hierarchical or alphabetical format.
Types of Taxonomies
© 2007 by ContextualAnalysis, LLC
total compensation . compensation . . base salary (salary) . . deferred payments (deferred compensation) . . variable pay . benefits . . 401(k) plan . . health benefits . . . dental plan . . . disability insurance
Types of Taxonomies: Taxonomy Example
© 2007 by ContextualAnalysis, LLC
Thesaurus
Plural form: thesauri
All the features of taxonomies, plus the associative relationship of related terms (RT)
Types of Taxonomies
© 2007 by ContextualAnalysis, LLC
Types of Taxonomies:Thesaurus Example, Alphabetical
Building Permits BT Permits
Business Licenses BT Licenses
Business Taxes BT Taxes
Fees RT Taxes
Licenses NT Business Licenses RT Permits
Operating Permits BT Permits
Permits NT Building Permits; Operating Permits RT Licenses
Taxes NT Business Taxes RT Fees
© 2007 by ContextualAnalysis, LLC
Types of Taxonomies:Thesaurus Example, Hierarchical
Vocabulary Terms Related Terms
Licenses, Permits & Taxes
. Fees Taxes
. Licenses Permits
. . Business Licenses
. Permits Licenses
. . Building Permits
. . Operating Permits
. Taxes Fees
. . Business Taxes
© 2007 by ContextualAnalysis, LLC
Synonym Ring
+ preferred terms
= Authority File
+ broader/narrower terms
= Taxonomy
+ related terms
= Thesaurus
Types of Taxonomies—Summary
© 2007 by ContextualAnalysis, LLC
Facets are fundamental categories by which an object or concept may be described
Example: some facets describing a toy ball:
size, weight, shape, color, texture, material
Taxonomies and Facets
© 2007 by ContextualAnalysis, LLC
Uses of Facets: Browsing Hierarchies
Facets allow users to follow the path best matching the way they think (their mental model).
Taxonomies and Facets
© 2007 by ContextualAnalysis, LLC
Uses of Facets: Browsing Hierarchies
Example: epicurious.com > recipes > browse
Main ingredient Cuisine Preparation method Season/occasion Course/dish
Taxonomies and Facets
© 2007 by ContextualAnalysis, LLC
Taxonomies and Facets
epicurious.com
© 2007 by ContextualAnalysis, LLC
Uses of Facets: Fielded Search
Allows for greater specificity, thus increasing search precision.
But this is usually more complicated for users than simple searching, so it is often introduced as option on results page.
Taxonomies and Facets
© 2007 by ContextualAnalysis, LLC
alibris.com Advanced Search
© 2007 by ContextualAnalysis, LLC
epicurious.com Advanced Search
© 2007 by ContextualAnalysis, LLC
Requirements for Browsing/Search Facets
Development of metadata schema
Development of appropriate controlled vocabularies
Proper content tagging
Taxonomies and Facets
© 2007 by ContextualAnalysis, LLC
Aitchison, Jean. Thesaurus Construction and Use: A Practical Manual. 4th ed. Chicago: Fitzroy Dearborn Publishers
Resources
© 2007 by ContextualAnalysis, LLC
Resources
International standard for metadata: Dublin Core Metadata Element Set (ISO Standard 15836-2003)
http://www.niso.org/international/SC4/n515.pdf
© 2007 by ContextualAnalysis, LLC
National Information Standards Organization. ANSI/NISO Z39.19:1993. Guidelines for the Construction, Format and Management of Monolingual Thesauri. Bethesda, MD: NISO Press, 1994
Rosenfeld, Lou, and Peter Morville. Information Architecture for the World Wide Web: Designing Large-Scale Websites. 3d ed. O’Reilly Publishers, 2006.
Resources
© 2007 by ContextualAnalysis, LLC
Sinha, Rashmi. Beyond Cardsorting: Free-listing Methods to Explore User Categorizations
Available at: http://www. boxesandarrows.com/archives/ beyond_cardsorting_freelisting_ methods_to_explore_user_categorizations.php
Steckel, Mike, Karl Fast and Fred Leise. Creating a Controlled Vocabulary. 2002
Available at: http://www.boxesandarrows.com/archives/ creating_a_controlled_vocabulary.php
Resources
© 2007 by ContextualAnalysis, LLC
Contact Information
Fred Leise
www.contextualanalysis.com
@ChicagoIndexer