It is probably only known to insiders in the advertising industry: the IAB taxonomy. The taxonomy is the result of many years of standardization work by the Internal Advertising Bureau. This is an international trade association of the online advertising industry based in New York, which is committed, among other things, to unification and standardization in this industry. A German sub-organization is the “Online Vermarkter Kreis OVK”. The aim of the taxonomy is to assign WWW content uniformly to different classes (segments).
What Is a Taxonomy Anyway?
This is a classification scheme with which real objects or theoretical concepts can be categorized into specific classes (taxa) and subclasses. Most taxonomies have a hierarchical structure and are represented as a tree structure with individual branches. The deeper you go down the tree branches, the more specific the information becomes and the better the object is described.
When creating a schema, the challenge is to recognize the necessary main classes and subclasses and thus to capture the total quantity of objects or concepts and their interdependencies in the most tailored way possible. Another task is to find descriptive identifiers for all class levels. This makes it easier to navigate, find and store objects.
Taxonomy example: Excerpt for the “Business and finance” domain
What Is the Great Benefit of a Taxonomy?
A disorganized set of diverse objects (e.g. the World Wide Web with its pages, documents, images, etc.) is transformed into an orderly, manageable and comprehensible system that depicts relationships and dependencies and takes special cases into account.
Several different taxonomies are often conceivable for an unordered set of objects. For example, Lego bricks can be sorted by:
-
- Color of the Lego brick:
White, Green – (lime green, medium green, light green, grass green), Transparent – (blue, yellow, red …), Fluorescent …
- Color of the Lego brick:
-
- Belongs to a Lego building set:
Basic, Classic – (fire department, castle, train …), Creator – (vehicles – (articulated lorry, car ferry …), Buildings – (lighthouse, vacation home …)), . ..
- Belongs to a Lego building set:
The classification of objects in a taxonomy is often anything but trivial, even with simple taxonomies. Particularly in the case of documents or websites, it is often not possible to clearly assign them. A document can be assigned to different segments. Taxonomies that contain several levels mean even more effort for the classification process. However, they provide much more enriched information for the application.
What Is the Difference Between Taxonomy and Ontology?
A taxonomy is classically arranged hierarchically and describes the relationship “is subset of” or “is superset of”.
An ontology is regarded as a network. It can map multi-layered and interwoven data models and display any number of cross-connections. The cross-connections can also be stored with integrity rules and relationship types.
What Form Does the IAB Taxonomy Take?
IAB has developed a comprehensive taxonomy – i.e. a ready-made classification scheme – for World Wide Web content and makes it freely available on the web in Excel format. Although it recognizes the focal points of the advertising industry and the views of the American way of life, it is ideal for classifying content for a wide range of use cases.
The current taxonomy, which is also used by MORESOPHY, is IAB version 2.0. Customer requirements change over the years, setting different weightings or giving rise to completely new trends and products. This should of course be reflected in the taxonomy.
Version 2.0 combines 29 classifications, which correspond to so-called market segments . The depth of the associated taxonomy tree extends down to four levels, and the number of leaves on the tree, i.e. the last and most specialized elements, is just under 700. The most detailed market segments are Business & Finance, Sports and Style & Fashion.
What Is the Value of the IAB Taxonomy?
In conjunction with intelligent classification algorithms, the application of the IAB taxonomy is not limited to the field of advertising. Its universal applicability has proven to be a major advantage in practice. In the area of Human Resources, for example, the knowledge learned from the Careers class can be used. The Personal Finance segment is used for solutions in the banking and insurance sectors. If, for example, the “Sports – Football” class is combined with the “Style & Fashion – Men’s Fashion – Men’s Clothing – Sportswear” class, the topic of “Men playing soccer” is covered.
Behind MORESOPHY’s CONTEXTSUITE are complex artificial intelligence models that can be used to automatically assign any content – from the web such as documents or emails in the company – to the segments of the IAB taxonomy. In doing so, the software learns better and better what exactly distinguishes the content in the various segments from one another and what is really relevant in them. The IAB taxonomy provides a stable basis for using the knowledge acquired across many applications and use cases with a high degree of reusability and low costs.
In line with the topic: An article on the importance of customer-specific data for machine learning.
Computer scientist and computer linguist
Ulrike is a computer scientist and computer linguist. The thematic focus of her blog posts is on the comprehensibility and usability of the human-machine interface.
More articles from Content in Context
