In writing these two wiki pages about Hierarchies, Faceted classification systems and Tagsonomies, I realized that the real quest is to understand How human beings turn data into knowledge. The search for the perfect organizational paradigm is a journey through the structures of the human brain and how people manage to see shapes, lines, patterns and coherence in the great soup of incoherence that is the world of information.

The end goal of all of this research, design, testing and rumination is not just a software system that is easy for people to get. (If that were the case, let's just give people 1s and 0s, it doesn't get much simpler or more generic than that.) The end goal is a data structure that sits firmly upon the deep-seeded, some might say, hard-wired, natural structures of the human information architecture. The stuff of linguistics and grammar.

A simpler way to put it is, when considering simplicity in design, we must always ask the question of: Simple for whom?

But, to prevent ourselves from going off the deep end of academic modeling and theory-making, we've also grounded our research in real people and the tangible artifacts of their attempts to organize and structure their information.

Glossary

Hierarchy An organizational system comprised of items and groups of items, bound together in containers where both items and containers are organized into a fixed structure of parent-child relationships.

Faceted classification system A classification system comprised of items and independent attributes and attribute values where items are grouped by attribute values and attribute values are typed by attributes and attributes are loosely bound to another only insofar as they intersect or overlap one another. Examples included: Chandler data model, Spreadsheets and tables and iTunes / iPod. The Browser in iTunes is an example of a Browser UI for Faceted systems.

Tagsonomy The anti-classificaton, anti-system comprised of items and independent labels or tags (attribute values in Chandler-speak), but no label types or tag types (attributes in Chandler-speak). Items are grouped into tags and tags are bound to one another only insofar as they intersect or overlap with one another. Examples include: del.icio.us, flickr, Gmail.

The papers in a nutshell:

Hierarchies are good at telling stories, precisely because they're so inflexible and immobile. Facets and Tags are great at being flexible, but horrible at telling stories precisely because they're so flexible and mobile.

How can we use the best parts of both systems in order to exploit their strengths and render their weaknesses irrelevant.

High-level outline: This is a tale of Order falling into Disorder and what Chandler is going to do about it.

• Why do we care about studying classification systems
• A brief overview of the 2 and 1/2 organizational systems we will look at: Hierarchies, Faceted systems, Tagsonomies
• Why do people organize? What are their organizational needs?

• How Hierarchies are good at meeting some organizational needs: Telling the Story of our Lives and a Guided Navigation experience
• How Hierarchies are horrible at meeting other organizational needs: Targeted search and retrieval, Favorites and Storing semantic data
• How trying to use Hierarchies as a one size fits all makes Hierarchies unusable for all organizational needs

• How Faceted systems are good at meeting some organizational needs: Targeted search and retrieval, Storing semantic data
• How Faceted systems are not as good at meeting other organizational needs: Telling the Story of our Lives and a Guided Navigation experience
• How Faceted systems are hard to use: A pain in the butt to add metadata and Hard for laypeople to design well

• How as a result, just like Hierarchies, Faceted systems ultimately disintegrate into Tagsonomies
• How Tagsonomies are overwhelming and fail to meet most organizational needs

Continue to 2 of 3: HierarchyPapers

---

Comments

See VirtualityPresentationImages for presentation slides