The SearchStax Site Search solution honors the search industry’s use of stopwords to help focus the results of a website search. Stopwords are common words that should be ignored when evaluating a query because they dilute the relevance of the results.
Stopwords Example
If we search the SearchStax technical document collection for the keywords “sort rank” we get 28 hits.
That’s 28 documents that contain either sort OR rank (or both).
If we search for sort and rank we get about ten times as many hits:
The pages containing sort or rank are buried among all the pages that contain and. A naive user might have some trouble finding the useful pages in the results list.
To alleviate this situation, we go to the Site Search Basic Relevance > Stopwords Tab and add “and” to the list of stopwords.
After publishing the stopwords list, and refreshing the Search UI Kit > Search UI App, the same search remains focused on sort OR rank content. We’re back to 28 hits again.
This is the benefit of stopwords in keyword search.
Stopword Lists
By default, Site Search does not have any stopwords configured out-of-the-box. Why is that?
Site Search supports multi-language experiences, meaning that you can index documents or webpages in over forty languages. Stopwords are language-specific, so SearchStax leaves that step to you.
It is not difficult to find stopword lists on the Internet, but you can also find them in searchstax-client-master.zip, a downloadable file we provide as a source for the Zookeeper Command Line Interface (zkcli.exe).
Download this file (about 280 MB) and navigate to one of the >searchstax-client-master\solr-n\configsets\_default\conf\lang subdirectories. Look for stopword files named using ISO 639 language codes.
You might need to edit the file to remove the #comments at the beginning. Then return to the Basic Relevance > Stopwords Tab to load the file into Site Search.
You must remember to click the Publish button at the bottom of the screen to load the list into Solr.
Questions?
Do not hesitate to contact the SearchStax Support Desk.