The Solr search provider in Sitecore 7 allows users to rank the search results using different types of boosting methods. This article shows how to use the Query-Time Boosting to sort the search results and highlights the common mistakes that developers should avoid.

Assume that you want to run a search query over a set of Sitecore items, each of which has a title and description fields, and you would like the score of the results that contain the search keyword in the title to be higher than the ones contain the search keyword in the description.

The search query that uses Query-Time Boosting should look like:

var query = "guitar";
var dataQuery = context.GetQueryable<SearchResultItem>()
.Where(i => i.Title == query.Boost(3.0f) || i.Description == query.Boost(1.0f))

Read More Sitecore 7 & Solr – Sorting Search Results by Score

If you are about to implement an auto-compelte text box for your search page, then you are in the right place. This article will show you how to display search suggestions using Sitecore 7 and Solr Search Provider. Solr is a very powerful search platform which provides many indexing and query time analyzers. The NGram analyzer is used here to implement search auto-complete feature. Before I go through the details, please don’t do the following:

Don’t use String.Contains()

var context = index.CreateSearchContext();
var results = context.GetQueryable().Where(i=> i.Title.Contains(searchText));
...

Please don’t ever do this, your client will come back complaining that, when their customers search for “pen” they get irrelevant suggestions such as “Dispenser”, “Open” and “Sharpener”.

Don’t use String.StartsWith()

var context = index.CreateSearchContext();
var results = context.GetQueryable().Where(i=> i.Title.StartsWith(searchText));
...

This won’t work as well because it will fail to return results such as “blue pen” where the title doesn’t start with the word “pen”.

The proper way (Using NGram)

These are the steps to properly implement this feature:
Read More Sitecore 7 & Solr search auto-complete using NGram