As promised, here is another installment covering the exciting improvements made to RSA Archer search in the 6.6 release. This article will specifically cover enhancements made to global search.
Background
One of the major frustrations is that it takes too much time to find the “right” record in RSA Archer, even when using the global search tool in the main Archer toolbar. Often the record that would seem to be the most relevant to your search term appears halfway down on page 7 of the global search results instead at the top of page 1. Global search uses a keyword index to find matches across the entire RSA Archer instance. Prior to 6.6, an Archer record was indexed as one document. There was no way to weight anything within the document. This affected the relevancy of matches, and therefore the order of the global search results. So when searched for Australian Physical because you wanted to read the core policy section on Australian Government Physical Security, the record where Australian Physical policy in the Section name was not necessarily considered the most relevant. Because this section is related to many other records in Archer it could end up on the 7th page of the result since the cross-reference field carries the same weight for keyword matches.
The following two features were added to improve the experience when using global search
- Weight relevancy for global search indexing
- Type-Ahead in global search tool
Let’s take a deeper dive into each.
Weight Relevancy For Global Search Indexing
With the introduction of Weight relevancy for keyword indexing, Global Search can now prioritize content that matches the most relevant data within the content. Relevancy rank is defined by the following table:
This means records where the search term occurs in the key field are gathered at the top of the result set. Records where the search term occurs in a tracking id field are given second priority. This is helpful for the many cases where you know the specific record id you need to find. Matches against all other fields are prioritized third. This means matches where the relevant record is referenced will now follow the relevant record in the results.
Highlights
- Weight Relevancy is enabled by default, and requires no user configuration.
- If Elasticsearch is not selected as the Search Index in the RSA Archer Control Panel, you must rebuild your search indices to utilize this feature.
- If the RSA Archer instance is using an Elasticsearch server, you do not need to rebuild your search indices.
- The same set of results return, but are prioritized with the relevant results at the top.
To ease the burden of rebuilding the search indexes, a new option was added to the Archer Control Panel to rebuild the Lucene keyword indexes in place. The process indexes all the content in the same manner, it just does not flush the current index first. The index is just updated with new weighting data. Because the indexes are not flushed at the beginning, global and keyword search are essentially uninterrupted while the weighting data is added.
Type-Ahead for Global Search
Making this new relevancy feature even more impactful is the new type-ahead capability added to global search.
Highlights
- As user types in global search a drop-down is presented to show the 6 most relevant matches.
- Type ahead adds an implied * at end of term
- Drop-down shows application, key field, and tracking id of matching record.
- Results update as user types more characters.
- User can select item from drop-down to immediately view the record.
- Selecting View All Matches executes global search and displays the global search results
These two features make global search a much more powerful tool. The number of clicks and time it takes to find the “right” content is drastically reduced. Check back tomorrow for one more article covering more search enhancements. Note that all of the search enhancements will be covered in this week’s Free Friday Tech Huddle so register for that event to see these search features in action. WEBEX Registration
