Feature Overview

Similar Documents feature represents a mechanism of returning additional query results that are:

• Of different types
• From different backends
• Similar to the initial results set for the user’s input query

Based on the configuration settings, the metadata of the results returned by the user’s query is computed and transformed into a “similar documents” request with the scope of querying the backend The search engine your SmartHub instance uses to perform queries. SmartHub can be configured to use more than one search engine. for results that are similar with the set returned, but do not necessarily match the user’s query.

Similar Documents results are displayed in Content-by-Search components on the SmartHub Results page. 

How Similar Documents Work

The “similar documents” returned from a query do not necessarily match the terms from the query.

• Similar documents work on category matching, not keyword matching.

• Terms in the same category as the terms queried are therefore returned as "similar documents."

Note: There is no hit highlighting performed in the UI because keyword matching is not used by this feature.

Example

Queries for the terms “Hiroshima bomb” may return related results for “Nagasaki” without the Nagasaki documents containing the term "Hiroshima."

• This happens because similar matches work on category matching rather than keyword matching.

• In this example both “Hiroshima bomb” and “Nagasaki” are contained in the category “atomic bombs” or “WW2.”

• Terms in the same category as the terms queried ("Hiroshima bomb") are therefore returned as "similar documents."