==Major Terms, Minor Terms, and Relevance Scores==LexisNexis SmartIndexing assigns a relevance score to each index term that is assigned to a document. The score is based on a caluclation of how important the concept behind the term is to the document. If a term is given a relevance score of 85% or higher, it is considered a Major Term. If it has a lower score, it is a Minor Term.
==The Index Terms List==When you open a full-text document in LexisNexis Academic, you will usually find list of index terms at the bottom. If the terms are accompanied by parenthetical number and percent sign, you know that these are index terms assigned by LexisNexis SmartIndexing and the number shown is their relevance score. If there is no parenthetical number, the terms were supplied by the orignial publisher.
[[Image:Relevance_score1.jpg | center | 100% | frame | Example of Relevance Scores]]
==The Index Terms Table==
In many news and business-related documents you will also see a table of index terms at the bottom of the document. There are checkboxes next to the terms and buttons that will use the terms you check off to narrow your current search or build a new search (a technique that is outside the topic of this article). By default, only the Major Terms are shown. You will also see a pair of links below this table that will change the display to show the Minor Terms and the revlevance scores.
[[Image:Show_scores.jpg | center| 100% | frame | Example of Show/Hide Scores]]
==Searching on Relevance Scores==But what if you want to find documents with a 70% or higher score? Because the relevance scores appear in the "Terms" segment of the full-text document, you can use LexisNexis Terms and Connectors to search for a string that can only occur when the relevance score is 70% or better. When building this kind of query, it is best to work from an example of a document that has the index term(s) you want to search on. The example below shows a query for documents that contain the word "infrastructure" and also have the index term "Iraq" posted to them with a 70% or better relevance score. Because we know how the index term and its score will be displayed in the list at the bottom of the document, we can build a query that matches. The two tricks involved here are the use of the "Terms" segment, which searches only against index terms, and the "*" wildcard character, which we use to replace the characters 0-9 in the search string. We are also using the "OR" operator to cover three ranges of numbers at once.