Faceted Knowledge Base

Sub Banner

Overview
By convention, finding information is limited to searching or traversing a hierarchical structure of categories and sections. Both can be effective and have their place, but suffer from issues related to knowing the search terms to use and where articles could have a natural fit to various places within a single hierarchical organization.



However, there is a third way, in particular an alternative to the traditional single, and often multilevel hierarchical structure. Faceting is an approach that enables multi-classification of articles using a number of relevant dimensions within contextual scope of the knowledge base. It is also uses the principles of a taxonomy, a predefined classification as opposed to free form/keyword tagging.

By defining a number of orthogonal faceting dimensions, such as product type, article format, and tasks, the user can approach from any number of directions, apply these filters to narrow down to a candidate list of potential articles that could yield the answers sought.

End-user Experience
A Knowledge Base sub-landing page is provided as a focal starting point. This can include an introductory text and a range of useful additional methods to find information, including conventional keyword searching, lists of promoted articles, hot articles, and fresh content.

A left or right hand set of facet dimensions, typically three, with their first level decompositions can be used to start narrowing down suitable articles. Discreet number counts against each facet values gives an immediate sense of the breadth and width of the knowledge base. The user is able to select the most likely angle of attack, whether that is a product, format, or task type. Once a filter selection is made, a candidate list is returned and magically the faceted filters prune themselves to show only the applicable facet values and their counts which are common to the returned list. The remaining facet dimension can be applied to narrow the list further.

Show/hide snippets further enhances the experience and avoids loss of focus, as the shortened candidate list of articles can quickly be assessed without having to drill down to each article and risk losing the search criteria used to find it.

Keyword search still plays an important role. It can be used in conjunction with faceted filtering. Returned search results can be shown with the results accessed against the available common faceting terms. This provides a summary context for the results, rather than a just a count of hits and the option to use the facet filter to take over the task of refining the initial search results.

Implementation
Typically, but not exclusively as a hybrid approach can be taken, a single super forum/section is used to contain all articles. The article authors create content as normal, but are also given drop-down faceting dimension lists to multi-classify the content. Selections can be either mutually exclusive within a facet dimension, but can also be multi-select e.g. product version numbers.

Each time an article is created, updated or deleted, the CloudSET replacement indexing engine is updated and becomes effective within a minute or so of the update.

Set-up involves the design of a suitable faceting schema, a task that is supported by consultancy to help establish a suitable information architecture for your knowledge base. Changes to the faceting schema can be made, additions being the simplest, as well as minor terminology changes. Retiring terms can involve a little more work, as a reclassification of all articles that used the removed term is required.


  • Overview
  • End-user Experience
  • Implementation

Faceted Knowledge Base

Sub Banner

Faceted Knowledge Base

Overview
By convention, finding information is limited to searching or traversing a hierarchical structure of categories and sections. Both can be effective and have their place, but suffer from issues related to knowing the search terms to use and where articles could have a natural fit to various places within a single hierarchical organization.



However, there is a third way, in particular an alternative to the traditional single, and often multilevel hierarchical structure. Faceting is an approach that enables multi-classification of articles using a number of relevant dimensions within contextual scope of the knowledge base. It is also uses the principles of a taxonomy, a predefined classification as opposed to free form/keyword tagging.

By defining a number of orthogonal faceting dimensions, such as product type, article format, and tasks, the user can approach from any number of directions, apply these filters to narrow down to a candidate list of potential articles that could yield the answers sought.

End-user Experience
A Knowledge Base sub-landing page is provided as a focal starting point. This can include an introductory text and a range of useful additional methods to find information, including conventional keyword searching, lists of promoted articles, hot articles, and fresh content.

A left or right hand set of facet dimensions, typically three, with their first level decompositions can be used to start narrowing down suitable articles. Discreet number counts against each facet values gives an immediate sense of the breadth and width of the knowledge base. The user is able to select the most likely angle of attack, whether that is a product, format, or task type. Once a filter selection is made, a candidate list is returned and magically the faceted filters prune themselves to show only the applicable facet values and their counts which are common to the returned list. The remaining facet dimension can be applied to narrow the list further.

Show/hide snippets further enhances the experience and avoids loss of focus, as the shortened candidate list of articles can quickly be assessed without having to drill down to each article and risk losing the search criteria used to find it.

Keyword search still plays an important role. It can be used in conjunction with faceted filtering. Returned search results can be shown with the results accessed against the available common faceting terms. This provides a summary context for the results, rather than a just a count of hits and the option to use the facet filter to take over the task of refining the initial search results.

Implementation
Typically, but not exclusively as a hybrid approach can be taken, a single super forum/section is used to contain all articles. The article authors create content as normal, but are also given drop-down faceting dimension lists to multi-classify the content. Selections can be either mutually exclusive within a facet dimension, but can also be multi-select e.g. product version numbers.

Each time an article is created, updated or deleted, the CloudSET replacement indexing engine is updated and becomes effective within a minute or so of the update.

Set-up involves the design of a suitable faceting schema, a task that is supported by consultancy to help establish a suitable information architecture for your knowledge base. Changes to the faceting schema can be made, additions being the simplest, as well as minor terminology changes. Retiring terms can involve a little more work, as a reclassification of all articles that used the removed term is required.


  • Overview
  • End-user Experience
  • Implementation

Faceted Knowledge Base

Sub Banner

Overview
By convention, finding information is limited to searching or traversing a hierarchical structure of categories and sections. Both can be effective and have their place, but suffer from issues related to knowing the search terms to use and where articles could have a natural fit to various places within a single hierarchical organization.



However, there is a third way, in particular an alternative to the traditional single, and often multilevel hierarchical structure. Faceting is an approach that enables multi-classification of articles using a number of relevant dimensions within contextual scope of the knowledge base. It is also uses the principles of a taxonomy, a predefined classification as opposed to free form/keyword tagging.

By defining a number of orthogonal faceting dimensions, such as product type, article format, and tasks, the user can approach from any number of directions, apply these filters to narrow down to a candidate list of potential articles that could yield the answers sought.

End-user Experience
A Knowledge Base sub-landing page is provided as a focal starting point. This can include an introductory text and a range of useful additional methods to find information, including conventional keyword searching, lists of promoted articles, hot articles, and fresh content.

A left or right hand set of facet dimensions, typically three, with their first level decompositions can be used to start narrowing down suitable articles. Discreet number counts against each facet values gives an immediate sense of the breadth and width of the knowledge base. The user is able to select the most likely angle of attack, whether that is a product, format, or task type. Once a filter selection is made, a candidate list is returned and magically the faceted filters prune themselves to show only the applicable facet values and their counts which are common to the returned list. The remaining facet dimension can be applied to narrow the list further.

Show/hide snippets further enhances the experience and avoids loss of focus, as the shortened candidate list of articles can quickly be assessed without having to drill down to each article and risk losing the search criteria used to find it.

Keyword search still plays an important role. It can be used in conjunction with faceted filtering. Returned search results can be shown with the results accessed against the available common faceting terms. This provides a summary context for the results, rather than a just a count of hits and the option to use the facet filter to take over the task of refining the initial search results.

Implementation
Typically, but not exclusively as a hybrid approach can be taken, a single super forum/section is used to contain all articles. The article authors create content as normal, but are also given drop-down faceting dimension lists to multi-classify the content. Selections can be either mutually exclusive within a facet dimension, but can also be multi-select e.g. product version numbers.

Each time an article is created, updated or deleted, the CloudSET replacement indexing engine is updated and becomes effective within a minute or so of the update.

Set-up involves the design of a suitable faceting schema, a task that is supported by consultancy to help establish a suitable information architecture for your knowledge base. Changes to the faceting schema can be made, additions being the simplest, as well as minor terminology changes. Retiring terms can involve a little more work, as a reclassification of all articles that used the removed term is required.


  • Overview
  • End-user Experience
  • Implementation

Faceted Knowledge Base

Sub Banner

Faceted Knowledge Base

Overview
By convention, finding information is limited to searching or traversing a hierarchical structure of categories and sections. Both can be effective and have their place, but suffer from issues related to knowing the search terms to use and where articles could have a natural fit to various places within a single hierarchical organization.



However, there is a third way, in particular an alternative to the traditional single, and often multilevel hierarchical structure. Faceting is an approach that enables multi-classification of articles using a number of relevant dimensions within contextual scope of the knowledge base. It is also uses the principles of a taxonomy, a predefined classification as opposed to free form/keyword tagging.

By defining a number of orthogonal faceting dimensions, such as product type, article format, and tasks, the user can approach from any number of directions, apply these filters to narrow down to a candidate list of potential articles that could yield the answers sought.

End-user Experience
A Knowledge Base sub-landing page is provided as a focal starting point. This can include an introductory text and a range of useful additional methods to find information, including conventional keyword searching, lists of promoted articles, hot articles, and fresh content.

A left or right hand set of facet dimensions, typically three, with their first level decompositions can be used to start narrowing down suitable articles. Discreet number counts against each facet values gives an immediate sense of the breadth and width of the knowledge base. The user is able to select the most likely angle of attack, whether that is a product, format, or task type. Once a filter selection is made, a candidate list is returned and magically the faceted filters prune themselves to show only the applicable facet values and their counts which are common to the returned list. The remaining facet dimension can be applied to narrow the list further.

Show/hide snippets further enhances the experience and avoids loss of focus, as the shortened candidate list of articles can quickly be assessed without having to drill down to each article and risk losing the search criteria used to find it.

Keyword search still plays an important role. It can be used in conjunction with faceted filtering. Returned search results can be shown with the results accessed against the available common faceting terms. This provides a summary context for the results, rather than a just a count of hits and the option to use the facet filter to take over the task of refining the initial search results.

Implementation
Typically, but not exclusively as a hybrid approach can be taken, a single super forum/section is used to contain all articles. The article authors create content as normal, but are also given drop-down faceting dimension lists to multi-classify the content. Selections can be either mutually exclusive within a facet dimension, but can also be multi-select e.g. product version numbers.

Each time an article is created, updated or deleted, the CloudSET replacement indexing engine is updated and becomes effective within a minute or so of the update.

Set-up involves the design of a suitable faceting schema, a task that is supported by consultancy to help establish a suitable information architecture for your knowledge base. Changes to the faceting schema can be made, additions being the simplest, as well as minor terminology changes. Retiring terms can involve a little more work, as a reclassification of all articles that used the removed term is required.


  • Overview
  • End-user Experience
  • Implementation