University at Buffalo (UB)
Transforming UB’s digital communications.
Unexpected items in my site's search results
Unexpected items appear in your site's Search Results.
Revised June 3, 2022
The Issue
Normally sites limit their Search Results to pages that occur within their own site.
Pages can be included in these results for routine reasons, such as being included in a module or list. But rarely the results may include a page that is not on your site, and whose inclusion has no obvious cause.
Workaround
We have discovered one way that this can occur.
List Host Page - e.g. your Search Results page includes a page with this format (note HOST and DETAIL in the path):
http://www.buffalo.edu/your-site/some-page.host.html/content/shared/another-site/another-page.detail.html
The presence of HOST and DETAIL in the URL path indicate a List is 'hosting' a page, so when someone clicks on the link, the page will be displayed as if it were originally part of the 'host' site, on that 'host' page.
When the search engine updates its master index, pages are grouped based on their path, and a 'hosted' page is grouped with the host's index, even though it may not actually be part of that site.
The following example was in Management's site search when searching for +v:host.html +v:university.
One can exclude very specific or broad sets of host page results by modifying the Site filter field in the Search component's dialog box.
Specifically, enter in another comma separated value starting with -host.html/content and construct the remainder of the URL to match the beginning of the set of hosted pages that shall be excluded.
- For example, to exclude all hosted /content/shared/university pages, add -host.html/content/shared/university to the Site filter list.
- For example, to exclude only the undesirable result from above, add -host.html/content/shared/university/page-content/the-buffalo-niagara-region/things-to-do-in-buffalo/food_and_dining.detail.html .
Additionally, individuals may contact the DCT Help Team about restricting the search to a specific search collection if the above option is not sufficient.