In January, I wrote a primer about clustering. I explained that:
… clustering is the automatic organization of search results into sets of results that have something in common. Some search engines and some federated search engines provide clustering features.
I also introduced faceted search, also known as faceted navigation:
This technology guides a user to relevant content by organizing search results in a hierarchical structure and providing labeled choices of paths in the hierarchy to follow. A faceted search system might have a series of pulldown menus that guide a user from the broad category of “Iraq” to “Iraq -> Geography”, to “Iraq -> Geography -> Maps” to “Iraq -> Geography -> Maps -> Baghdad.” Endeca is one vendor that provides faceted searching.
Faceted search is a powerful extension to clustering, and is of particular value in federated search, especially in the enterprise. Wikipedia provides an overview about faceted classification, which is required to build faceted navigation systems. Content management platform Drupal.org has a page describing their faceted search capability. Drupal-specific details aside, I found the page to provide an excellent description and summary of the motivation for providing a faceted search capability.
Faceted searching is of particular value in federated search for several reasons:
- Many users of federated search are researchers; research-oriented searches are not typically one-shot searches; they lend themselves to interactivity, which faceted searching provides.
- Faceted navigation can guide federated search users in a discovery process, starting with a broad subject and leading users to more specialized subtopics.
- When article meta data is available, as it is at times with federated search sources, faceted navigation, like clustering, can organize results by author, publication date, journal, or other meta data attribute.
Search company Endeca has an outstanding 7 minute video that illustrates the value of faceted navigation and reinforces the points above. Click on the “Watch Now” link in the center of their home page; you’ll need to register with Endeca to view the demo. Note: I have not used any Endeca products so I am not endorsing their technology.
The Endeca video makes some important points:
- Faceted navigation allows users to conduct effective searches even when they don’t know exactly what they’re looking for. In the video, an engineer is looking for a replacement part. He doesn’t know what the replacement part number might be or even if such a part exists. He doesn’t even know how to describe the part. He only knows some attributes of the replacement would be.
- Faceted navigation is critical in the enterprise search environment. I wrote recently about federated search in the enterprise and made the point which the Endeca video reinforces:
the popular search engines perform full text searches of unstructured text but enterprise content is much more structured than content in the Internet at large, it often contains fielded data in databases, and it is often hierarchically organized. Federated search vendors that want to sell into the enterprise need to consider this important difference.
- Guided navigation, a close cousin to faceted navigation, is particularly useful for solving problems where one can ask questions with a number of discrete possible answers and apply constraints to the possible solutions. In this type of situation, an iterative approach can lead to a useful solution. This only works to the extent that knowledge in an organization can be tagged with attributes.
- Creating an effective search solution may need to comprise a number of elements, such as faceted navigation, clustering, a search box, and range filters for attributes.
Faceted search can add tremendous value to an enterprise’s data but the design and development work to create powerful faceted search systems that are intuitive to use is very challenging. Ontologies and taxonomies (topics for future articles) are among the complexities that must be addressed in building powerful faceted search systems. While I am not surprised that few sophisticated faceted search solutions exist, especially in the federated search industry, I predict that it will only be a matter of time before this feature becomes as commonplace as clustering is becoming.