Princeton Service Portal Search


Service Portal Improvements

In December 2020, a series of search and filter improvements were made to the Princeton Service Portal.  These include:

  1. Enhanced keyword search algorithm allowing for more precise search results. 
  2. Search results are now prioritized across articles, forms, and services and will present the most relevant results across all three types.  
  3. Forms appear more prominently in search results.
  4. Search suggestions appear based on what you have typed and what others are looking for.
  5. Filters have moved to the left and results are now presented in blocks.
  6. New filters for the last updated date and campus unit.  
  7. Helpful preview information provided such as:  hyperlinked title, share button, KB #.  
  8. Newly published articles will include a “New” badge for 30 days.

Feedback about Search?

Please use the blue portal Feedback button to report any findings or issues. (You must be logged in. Each survey will create an incident for the ServiceNow team.)

Overview

The Princeton Service Portal uses several features which affect how a reader finds articles when they search. 

  1. Article Can/ Cannot Read.
  2. Reader Authentication (Log In).
  3. Search Engine (Search Criteria & Relevancy).

Can/ Cannot Read

Can/ Cannot Read settings determine the visibility of an article to a particular audience when a reader performs a search.  Content authors can specify the can/ cannot read audience for each article.  To learn how to change the can/ cannot read setting read article KB0012910 Article Publication Checklist.

Authentication (Login)

Search Engine

When a reader inputs search criteria, the Princeton Service Portal performs a search across Services, Forms (Catalog Items/ Requests) and Knowledge Articles to return the content which most closely matches the search criteria. Only published articles will return in a search.

A search algorithm determines which articles, forms and services are most relevant based on the criteria entered. The search engine algorithm evaluates the search term based on assigned weighting. The three most heavily weighted fields are Article Title, Abstract, Metadata.  The words in the article body are also searched, but this is the least weighted criteria.  

The algorithm is also set to ignore certain high-use words to alleviate unwanted results. 

How Readers Typically Search

Based on our analytics, readers typically search for and navigate to portal content by using direct article links (links sent to them via email, or available from other websites), Princeton University Search, or Google Search.  Direct searches in the Princeton Service Portal are also performed by readers more familiar with using the portal.