Faceted search
8/23/2012A user's first search query is often of a general nature, and there is a need for help drilling down in the results. Faceted search is a well established design pattern for guiding the user to the results she expects. When used for e-commerce and product catalogues this pattern will give the best conditions for quick and exact search results.
When using faceted search, the user is presented with a number of categories, called facet values, divided into one or more groups, called facets, after having performed the first search query. If the website contains a lot of information, it is possible to categorise in several steps, and in that way help the user to define the search query even further.
- Use patterns
- Implementation Details
- Integration Module v3
- .Net Integration
User Needs
As a user of the search function, your knowledge about the exact contents of a website is often limited. The objective of the first search is often to gain a general sense of what kind of information you can find on the website and how the search function works. One of the challenges for search page designers is to catch these users and guide them to the relevant information.
The goal should be to start a dialogue with the user, where the user enters a search query which leads to a number of search results. In addition, the user should also be able to pinpoint the search query by choosing a number of categories shown on the search page as facets.
Facets can be placed as tabs on the search page, like in this example. The number displayed in each tab indicates the number of results in that specific tab.
The image above shows an example of categorising in multiple levels with the help of category groups. The idea is to give the user the option to limit the number of results based on several criteria, in this case page type and subject. The user that is looking for a publication about radon gas from 2009, does not have to browse through 227 pages to find the relevant information. If you choose to limit the search to Publications and the Publishing year 2009, only seven pages remains for the search query radon.
