The items are defined according to their relationship with the other items in the hierarchy. With data management, classification and taxonomy are both methods for organizing and categorizing large amounts of data in a form that humans are able to comprehend. They describe the items in a way that makes it easy for us to return to them later without having to analyze each piece of data every time we need to use it.
Taxonomies are more concerned with providing exhaustive lists while classification is not exhaustive. Taxonomies are based on providing a hierarchical relationship map between a multitude of items while classification usually only groups items according to one or two attributes.
The fundamental difference is that taxonomies describe relationships between items while classification simply groups items. How does this relate to data management or product information management? Products can belong to any number of Categories and Trees. The categorization can be based on any arbitrary consideration.
The power here is you can organize your products in a manner that is different and distinct from a "data-oriented" perspective. You build a category for all the "Holiday Stuff" to show on the website with items that you want to promote during the season. The items included with your Holiday Stuff category may not have any common attributes that distinguish them from all of the other products in the catalog, but their organization and presentation together may increase conversion rates, average order value, etc.
This creates an abstract relationship between these items, not something that is a natural characteristic that ties them together. Within Akeneo, a product can only be assigned to one family at a time. This is important in the context of data governance, because it provides constraints on the product attributes.
Specifically, the product should ONLY have the attributes defined in the assigned family. This control on what data is needed or allowed within a product is critical for ensuring high-quality product data. In addition, a taxonomy can provide a way to uniformly enrich products by expanding the attributes associated with products to include data.
This is a form of "inheritance" where attributes added to a family are automatically included to products already assigned to that family. This would show a more accurate hierarchical structure, though to normal people, we may not be able to see the relation between items. If you are a cell phone manufacturer, you might have a class of phone and specific phones within that phone line.
Hierarchical taxonomies have implicit relationships. In our phone example, the name iPhone is the parent, the broader term. The iPhone 3G is a child term of iPhone.
If you think your subject matter fits well into a taxonomy, use a hierarchical taxonomy. If, however, while you are creating this taxonomy, you find that you are forcing matters and that you are duplicating items everywhere, perhaps think about a faceted taxonomy.
To take an easy example, say you own a candle and home decor website where you want people to be able to find red candles and red throw pillows. Instead of making your hierarchical taxonomy deeper than it needs to be, turn this into a faceted taxonomy.
Colour becomes a facet and Product Type becomes a facet the products being Candles and Pillows. You can use a faceted taxonomy when multiple, similar values can be applied to dissimilar items. To quote Denton again,. Facets will handle three or more dimensions of classification. When, for the purposes of the classification, it is possible to organize the entities by three or more mutually exclusive and jointly exhaustive categories, then facets are probably the appropriate classification.
Facets can be used to organize the entire world of knowledge, or the clothes in your cupboard, or anything in between.
If we take a look at Zappos. For their shoes, they use categories like colour, size, width, height, and style to classify their shoes. No one aspect is more important than the other and there is no clear relationship between these aspects.
They do not require complete knowledge of the entities or their relationships; they are hospitable can accommodate new entities easily ; they are flexible; they are expressive; they can be ad hoc and free-form; and they allow many different perspectives on and approaches to the things classified. Admittedly, while working on faceted taxonomies, it is difficult to choose the right facets. You need to choose one that is broad enough to encompass the right things, but narrow enough so as not to be all encompassing.
How will I choose these things? How will the database know to combine all these values together?
0コメント