Q&A on “Qmax”, the Lexis® Securities Mosaic® text search syntax



What is Qmax?


Qmax is a proprietary search syntax that allows you to construct a text search using either Boolean or “natural language” syntax, whichever you prefer.  Boolean syntax offers the use of operators such as AND, OR, and NOT, as well as the ability to organize or “nest” phrases using parentheses. Natural language syntax means you just type words in and go (as you do on Google, for example).  Results are boosted by relevance ranking, unless you specify otherwise.



Where is Qmax?


Qmax debuted on our SEC Filings page in the Fall of 2012; and in the Spring of 2013 we deployed it to these search pages: Laws, Rules, & Agency Materials; SEC Comment Letters; Law Firm Memos; and Model Business Documents.

We will eventually deploy it to every search page on the Lexis® Securities Mosaic® website.



How is it special?


Qmax is special because it is a best-of-both-worlds approach to text searching.   Since both worlds – Boolean and natural language syntax – have distinct advantages, we wanted to incorporate both, thus offering the best search experience possible.  Essentially, Qmax works by recognizing whether or not a text search query is Boolean. If the user does not input any Boolean operators, the search will act like a Google search. 



This sounds complicated. Are there a lot of new operators to learn? Is Qmax  hard to use?


No. In fact, quite the opposite: Qmax pages are easier to use because knowledge of Boolean syntax becomes entirely optional.  You can still get good results just by entering terms and hitting the Search button.  And if you’re used to using Boolean logic on our search pages, you can keep doing what you’ve been doing.



What if I don’t want relevance boosting of my results? What if I’m old school and just want to live in the Boolean world?


You can easily exclude relevance boosting from your list of results after running your search. You can still live entirely in the Boolean world if you want to.



How will relevance boosting work? What constitutes a “stronger match”?


Relevance boosting is based primarily on the following criteria:


1)   Frequency or density of word/phrase matches.

The more times a word, phrase, or text string comes up in the document, the higher its relevance ranking.


2)   Phrase boosting.

 If you enter only a string of words (no Boolean operators), the search engine will boost results based on adjacency (words right next to each other). For example, suppose you type:

 merger agreement

Documents that contain the exact phrase “merger agreement” will tend to rank higher than others. Documents that contain the phrase “agreement and plan of merger” will also be returned, but will tend to rank lower.


3)   Document date.

 Relevance is also impacted moderately by the date of the document.   All else being equal, more recent documents will rank higher than older documents. (This is the current behavior on our Laws, Rules, & Agency Materials search page).


4)   Document title. 

On certain pages (such as Laws, Rules, & Agency Materials), text matches in the document title are boosted higher than text matches in the document.