How to create effective taxonomy

Knowing that you need information taxonomy is one thing. Knowing how to build it is another. These tips will help you build a profitable Web site.

Knowing that you need information taxonomy is one thing. Knowing how to build it is another. These tips will help you build a profitable Web site.

Taxonomy is the technical term for the guiding principles behind the organisation of information—a key concern for Web developers. Every Web developer should know how to harness the basic principles of taxonomy to design a logical, organised, efficient Web infrastructure. By understanding information taxonomy and using it to optimise Web sites, Web developers can maximise the value and capability of their work.
The two key aspects of information taxonomy are taxonomy structure and taxonomy view. Taxonomy structure provides a classification schema for categorising content within the content management process. Taxonomy view is a conceptual model illustrating the types of information, ideas, and requirements to be presented on the Web. It represents the logical grouping of content visible to a site visitor and serves as input for Web site design and search engineering. Together, these concepts can guide your Web development efforts to maximise return on investment. Build it right, and they will come.
Trinity compass for the road of taxonomy development
The goal of taxonomy is to implement structure in an "unstructured" world of information. The road of taxonomy development has many twists and turns. To be successful and keep on the right path, you must arm yourself with a compass.
Many articles, books, and presentations provide details on taxonomy methodologies and techniques, but the central, underlying theme always revolves around the three key factors of information architecture: business context, users, and content (Figure A). These factors reflect the fundamental business requirements for most taxonomy projects. Strategically, they provide a "trinity compass" for the road of taxonomy development. Here's a description of each factor:

Business context is the business environment for the taxonomy efforts in terms of business objectives, Web applications where taxonomy will be used, corporate culture, past or current taxonomy initiatives, and artifacts within the organisation and across the industry.

Users refers to the target audience for the taxonomy, user profiles, and user characteristics in terms of information usage patterns.

Content is the type of information that will be covered by the taxonomy or that the taxonomy will be built upon.

Figure A

Trinity compass for the road of taxonomy development
Taxonomy development process
Guided by the trinity compass, we can define and follow a taxonomy development process that addresses business context, content, and users. Typically, a taxonomy development process involves the steps illustrated in Figure B. Here's a look at each one.
Figure B

Taxonomy development process
Assemble a team
Assembling a team is a critical step and deserves much attention. Successful taxonomy development requires both taxonomy expertise and in-depth knowledge of the corporate culture and content. However, finding a group of people who posses both the taxonomy expertise and organisational knowledge is often an unrealistic goal. Further, having taxonomy and subject matter experts on the same team causes inefficiencies because they are each focusing on different aspects of the project, leading to conflict and confusion as to the prioritisation of objectives. Although taxonomy experts are interested in the actual construction of taxonomy, subject matter experts are concerned with content coverage, terminology, and labeling.
In addition, taxonomy interrelates with several aspects of Web development, including Web site design, content management, and Web search engineering. Therefore, a taxonomy team must collaborate with Web development groups responsible for these areas.
The ideal way to address all of these skills requirements is to build a taxonomy community by involving the different skills profiles through a dedicated taxonomy development team, a taxonomy interest group, and collaboration with other Web development groups. Specific criteria for member selection will vary from project to project. Common considerations are overall project scope, target audience, existing organisational taxonomy initiatives, and corporate culture.
The taxonomy development team should consist of members from both the technical and business communities, including, for example, information professionals, librarians, Web architects, and information architects. This team is responsible for the actual development of both the taxonomy structure and taxonomy view. Ideally, an information architect who is fluent in both information management and Web development will lead the project. The designated information architect coordinates with the taxonomy interest group and other Web groups.
The taxonomy interest group should be made up of subject matter experts or content experts from the business community who have in-depth knowledge of corporate culture and content. For small projects, the group may simply be part of a user focus group that is concentrating on the taxonomy track.
The actual structure of the taxonomy community will need to be adjusted according to the size of the organisation. Figure C shows a sample taxonomy community, and Table A describes the recommended skill profiles needed to develop a taxonomy structure and a taxonomy view.
Figure C

Sample taxonomy community

Define scope
Once the taxonomy team is established, the first agenda item should be defining the scope of the taxonomy project. The taxonomy development team and the taxonomy interest group should jointly create this definition. Sample scope questions are listed below, grouped by the three guiding elements of the trinity compass.
Business context

What is the purpose of the taxonomy?

How is the taxonomy going to be used?

What are the input sources for taxonomy design? (Possibilities include the Library of Congress classification system, existing listings of terms, a glossary, etc.)

Content

What is the content scope? (Possibilities include companywide, within an organisational unit, etc.)

What content sources will the taxonomy be built upon? (Specifically, the locations of the content to be covered in the taxonomy.)

User

Who will be using the taxonomy? (Possibilities include employees, customers, partners, etc.)

What are the user profiles?

This step should also define metrics for measuring the taxonomy values. For enhanced Web sites, baselines should be established for later comparison with the new site. An example would be the number of clicks it takes a site visitor to locate certain information.
Create taxonomy
Taxonomy creation can either be manual, automated, or a combination of both. It involves analysing context, content, and users within the defined scope. The analysis results serve as input for the taxonomy design, including both taxonomy structure and taxonomy view. The taxonomy development team is responsible for the actual mechanics of taxonomy design, whereas the taxonomy interest group is responsible for providing consultation on content inclusion, nomenclature, and labeling.
The design of the taxonomy structure and taxonomy view may run in tandem, depending on the resources available and project timeframe. All concepts presented through the taxonomy view need to be categorised properly according to the taxonomy structure. This will ensure that every content item is organised centrally through the same classification schema.
Along with taxonomy structure and taxonomy view, standards and guidelines must be defined. There should be a categorising rule for each category in taxonomy view and taxonomy structure. In short, you must define what type of content should go under any given category. Content managers can then refer to these rules when categorising content. If an automated tool is used for content tagging, these rules can be fed to the tagging application. Standards and guidelines help ensure classification consistency, an important attribute of a quality content management system and search engineering process.
Implement the taxonomy
The next step includes setting up the taxonomy and tagging content against it. This is often referred to as "populating" the taxonomy. Similar to taxonomy creation, implementation can be manual, automated, or a combination of both. The goal here is to implement the taxonomy into the Web site design, search engineering, and content management.
For Web site design, taxonomy view provides the initial design for the site structure and interface. The focus is on the concepts and groupings, not so much on nomenclature, labeling, or graphics. There may be a need to go through multiple iterations, moving from general to specific in defining levels of detail for the content. Types of taxonomy view include site diagrams, navigation maps, content schemes, and wire frames. The final site layout is built by applying graphical treatment to the last iteration of taxonomy view.
For search engineering, implementation can be accomplished in various ways. Taxonomy structure as a classification schema can be fed into a search engine for training purposes or integrated with the search engine for a combination of category browsing and searching. In the latter case, the exposed taxonomy structure is essentially a type of taxonomy view. One of the most challenging aspects of taxonomy implementation is the synchronisation between the search engine and the taxonomy, especially for search engines that do not take taxonomic content tagging in the indexing process. In such cases, a site visitor may receive different results from searching and browsing the same category, which could prove confusing.
Taxonomy structure needs to be integrated within the content management process. Content categorisation should be one of the steps within the content management workflow, just like review and approval. If a content management tool is available, the taxonomy structure is loaded into the tool, either through a manual setup process, or imported from a taxonomy created externally. Through the content management process, content is tagged manually or automatically against the taxonomy. In other words, the taxonomy is populated with content.
Test, test, test
The goal of testing is to identify errors and discrepancies. The test results are then used to refine the taxonomy design. The testing should be incorporated into the usability testing process for the entire Web application, including backend content management testing and front-end site visitor testing. Here is a sample checklist of testing topics:

Given specific information topics, can the site visitors find what they need easily, in terms of coverage and relevancy?

Given specific information topics, how many clicks does it take before a site visitor arrives at the desired information?

Given specific tasks, can the site visitors accomplish them within a reasonable timeframe?

Do the labels convey the concepts clearly or is there ambiguity?

Are the content priorities in sync with the site visitors' needs?

Does the structure allow content managers to categorise content easily?

Testing results are recorded and can later be compared with the baseline statistics to derive the measurements of improvements.
Maintain
Taxonomy design and fine-tuning is an ongoing process similar to content management. As an organisation grows or evolves, its business context, content, and users change. New concepts, nomenclature, and information need to be incorporated into the taxonomy. A change management process is critical to ensure consistency and currency.
Better structure equals better access
Taxonomy serves as a framework for organising the ever-growing and changing information within a company. The many dimensions of taxonomy can greatly facilitate Web site design, content management, and search engineering. If well done, taxonomy will allow for structured Web content, leading to improved information access.