2

We're encountering a challenge with the Sitecore search function. Specifically, we're having difficulties when trying to locate web pages using certain keywords. Our investigation reveals that the root cause of this problem seems to be the absence of space between words at the end of each component.

  • For illustration, let's consider a situation where one component ends with the word "worldwide." and the subsequent component begins with "Facts and Figures". In this case, the Sitecore search fails to find the page using these two words, as it interprets the content as "worldwide.Facts".

We attempted to resolve this by adding a non-breaking space element at the end of each component from the frontend, but this failed to deliver the desired outcome.

In our attempts to address this issue within the Document Extractor of Sitecore search, we have experimented with various JavaScript methods to retrieve content. At present, we are using 'body': $('#content').text(). We have also tested innerText(), textContent(), and prop("innerText"), but none of these methods have resolved the problem. We would appreciate any suggestions on alternative methods or approaches that could help resolve this issue. Thank you in advance for your assistance.

4
  • @Yaswanth can you share your analyzer, JS extraction script and a URL you are using to validate? Or the equivalent HTML code that can be used for validating. I'd like to test if it's an issue with the extraction code, or the HTML being returned form the you are accessing but I can't really troubleshoot with the provided infromation
    – searchGeek
    Commented Jun 11 at 16:28
  • @MarcelGruber Thank you for your willingness to assist with the issue. However, due to the confidential nature of the content and scripts involved, I'm unable to share them publicly on this platform. I attempted to reach out to you through your social media to discuss this privately, but it seems that direct messaging isn't available. Could you please provide an alternative way to contact you directly.
    – Yaswanth
    Commented Jun 12 at 6:35
  • @searchGeek and Yaswanth, it would be helpful for both of you to update your SSE profiles with additional contact information. Commented Jun 12 at 15:05
  • @MarcelGruber I've added email address in my about section or please join me in Slack community
    – Yaswanth
    Commented Jun 17 at 6:32

2 Answers 2

2

Please check Analyzers for attribute inside Administration -> Domain Setting -> Feature Configuration -> Textual Relevance in CEC portal.

Trying changing/adding attribute analyzer to Standard, Multi locale standard or with more advanced Ngram based matching.

Below is the screen of different types of analyzer in Textual Relevance. enter image description here

1
  • 1
    we are already using Multi Locale Standard Analyzer for Body Attribute.
    – Yaswanth
    Commented Jun 12 at 6:25
1

This is way the analyzers work (https://unicode.org/reports/tr29/#Word_Boundary_Rules - WB6-WB7c). You should add a space in "worldwide.Facts" to be "worldwide. Facts".

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.