What Is MongoDB Search？ A Beginner's Guide

Ever felt the frustration of a search that just doesn't get what you're looking for? MongoDB Atlas Search is a lexical search service that allows you to create sophisticated search experiences without managing separate infrastructure. But to fully leverage the capabilities of Atlas Search, we first need to understand how it works. In this video, we'll explore what Atlas Search is, provide an overview of how it works under the hood, and examine some key use cases. Let's get started. Atlas Search is most useful for scenarios where users need to find relevant information quickly through keyword search. Some common use cases include e-commerce product discovery, content management systems, knowledge bases, and any application where traditional exact match database queries fall short of delivering intuitive search experiences. To understand why it's useful in these scenarios, let's look at an example. Imagine we're working on a recipe app and we want to allow users to search for different foods and recipes. We could use find with regex to query the database for matching documents. For example, we can look up pasta using this query and we'll find documents that contain the word pasta, like this one, prosciutto and parmigiana pasta. However, with this regex query, we have to scan all characters in the name field, which can be very inefficient with large datasets. Additionally, these types of queries don't support the advanced search features that users expect from modern applications, like autocomplete or fuzzy search. But with MongoDB Atlas Search, we can create a search index, which will provide fast, relevant results, and build search experiences with all of the features that users expect. Let's take a closer look at how this works. To use Atlas Search, we need to create a search index on a MongoDB collection and write search queries using the search or search meta aggregation stages to retrieve and display relevant results. In this example query, we'll be using search. So now, let's use this Atlas Search index and query to find matches for the word pasta, just like we did earlier with that regex query, except that Atlas Search is going to make this search much more efficient and easier to manage. By now, you probably know how indexes and queries work in MongoDB, but in Atlas Search, they work a little differently. Atlas Search has its own indexing technology based on Apache Lucene. The Atlas Search process, MongoT, acts as a bridge between MongoDB's query engine and Lucene's search engine, handling both text indexing and querying. When creating a search index, the data from the collection is transformed into a structured format using a process called tokenization, where text fields are broken into smaller components, like words or phrases. We can customize how data is tokenized with different analyzers. These tokens are then stored in the index along with relevant metadata for quick retrieval during search queries. Let's take a closer look at our example to learn more about how this works. To implement Atlas Search, the first thing we need to do is create a search index. For our recipe app, let's create a search index for the food collection with the createSearchIndex method in the MongoDB shell. First, we come up with an index name so that we can use it in our search queries. Let's use foodNameIndex for the index name. Next, we can decide to index specific fields or all fields in the document. We'll talk more about this later when we discuss static and dynamic mappings. Here, we're only indexing the dish field so that users can use keywords to find dishes. We'll also need to provide the data type of the dish field, which, in this case, is a string. Once we create this index, Atlas uses the Lucene engine to process and organize data in the food collection. With the search index set, the MongoT process can query documents we want to search. So now, how do we execute a search query? We need to use another powerful tool known as aggregation. If you haven't already or simply need a refresher, check out our aggregation skill. To use aggregation with search, we use the search stage or search meta stage. Search is used to return documents that match the search criteria, while search meta is used to retrieve metadata about the search results. It's important to note that the search stages must be first in the aggregation pipeline for a query to execute correctly. For example, if we were to place a match stage before a search stage, we would get an error. This is because the match stage creates a new data set that no longer has access to the search index. The search stage performs a full-text search on a specified field or fields that are defined within the search index. Let's look at the search query from our example earlier. Here, we created an aggregation pipeline that uses the search stage with the food name index index created earlier. Next, we specify our query criteria. Here, we're using the text operator to perform a text search. We'll explore more available operators in future lessons. Finally, we define the query and the field path we want to search on. We're searching for mentions of pasta within the dish field for now. When we run this search query, every document that's returned has pasta mentioned in the dish field. Atlas Search assigns a relevance-based score to each document in the result set. It then returns these documents ordered from highest to lowest score, making sure we see the most relevant results first. So, in our example, we know that these two documents at the top of the return set have higher scores and are most closely associated with the search term pasta. This relevance scoring uses the BM25 ranking algorithm, which is designed to analyze the relationship between query terms and index data in your MongoDB collection. MongoDB Atlas Search also has additional controls for customizing scoring behavior, allowing you to tailor relevance ranking for specific fields or attributes. Check out our docs to learn more about how to adjust scoring and ranking. Once we've created our search index and built our search query, there are a few methods that can help you manage Atlas Search. To view all of your search indexes, you can use the getSearchIndexes method. To update an existing search index, use the updateSearchIndex method, along with the new index definition. And to delete a search index, we can use the dropSearchIndex method. But make sure to evaluate the impact on your application first. You can also easily manage your search indexes and test out your search queries with MongoDB Atlas and Compass. Ready to master the fundamentals of Atlas Search? Click the link in the description to claim your skill badge and boost your expertise. In this skill, we walked you through how to build search indexes and queries with these considerations in mind, so you can take full advantage of Atlas Search. First, let's recap what we learned in this lesson. MongoDB Atlas Search is an integrated full-text search service within Atlas, enabling sophisticated search experiences without separate infrastructure management. The MongoT process facilitates the ability to create search indexes and search queries. To use Atlas Search with your Atlas cluster, you must first create a search index on a collection. You can then perform search queries using the search and search meta aggregation pipeline stages. Finally, we showed you shell methods that you can use to manage Atlas Search indexes, including getSearchIndexes, updateSearchIndex, and dropSearchIndex.