Relevance Search is Next Generation Search

Introduction

In last release Microsoft introduced Next Generation Search (AKA Relevance Search) as a preview feature. In the latest release, i.e. November 2016 update, Microsoft has made this feature generally available. In this blog, I have covered this feature in detail.

Types of search

With relevance search, CRM now have 3 types of searches.

  1. Quick Find View – This search is restricted to particular entity. This is the basic search available since CRM 4.0
  2. Categorized search – This search was introduced in CRM 2015 release and is also known as global search. It searches the data across multiple entities using entity’s quick find view and shows the data categorized by particular entity.
  3. Relevance Search – With relevance search, you will get the results which are relevant to the search terms you are searching for. Relevance search will not be restricted to any entity type or any field, but it will search based on the relevance to your search terms.

What is Relevance Search?

  • Relevance search delivers fast and comprehensive search results based on the relevance across multiple entities.
  • It uses dedicated search service provided by Microsoft Azure which is external to Dynamics 365.
  • With relevance search, you can still use the other two types of searches.
  • Below are the benefits and enhancements provided by relevance search:
    • As external search service provided by Microsoft Azure is used to search the data, search performance is improved drastically and search is performed outside the D365 environment.
    • It uses semantic searches hence results may include inflectional words like win, won or winning, etc.
    • It returns results with matching the search term in any words and in any field in the entity.
    • The result is not categorized by the entity type, but it is returned and sorted based on the number of words matched and how the words are related to each other.
    • In the results, the words which are matched are highlighted.

Limitation

  • Relevance search is not available for D365 On Premise environments.
  • There is no limit on number of entities you can search, but there is a limit of 1000 index you can have per organization. While configuring entities for relevance search, you can see how many indexes are needed to enable the respective field and entity.
  • As relevance search is hosted on Microsoft Azure cloud platform, any changes to the data in D365 environment takes 15 minutes to get synced with Azure Search. Also, for average to large data it takes an hour or more to complete the full sync between D365 environment and Azure Search Service.

data-flow

  • Below fields are not supported for relevance search:

Find Fields

  1. Lookup
  2. Option set
  3. Rollup & Calculated fields
  4. Non-text

View Fields

  1. Lookup
  2. Option set
  3. Fields of related entities

Filter Fields

  1. Lookup
  2. Option set
  3. Fields of related entities

How Relevance Search makes a difference?

  • In quick find view search, i.e. entity level search, when user searches for a phrase, it tries to search all the sentence exactly in the fields of the entity and if all the words matches then only it returns the result.
  • In categorized search, i.e. global search, it also gives same result as quick find view, but it searches across the entities and maximum entities which can be searched is 10.
  • In relevance search, individual word from the search phrase is searched across the entities and in all the fields. The more relevant results are shown first. The relevancy is decided based on the number of words matched against the searched data and how close the words are to each other. For e.g., if the search term is found in an address and in an article, then result with address will be considered as more relevant data than the article.
  • There is no limit on how many number of entities you can search using relevance search.

How to configure relevance search?

Configuring relevance search is 3 steps process. First you need to enable the relevance search and then enable the entities which should be involved in the relevance search, and at last you need to configure the fields which should get included in relevance search.

Step 1: Enable relevance search

  1. You need to have system administrator/customizer role on D365 environment.
  2. Go to Settings > Administration > System Settings.administration_system-settings
  3. In General section, you will see an option as Enable Relevance Search. Select Yes and click OK on the popup shown.
    enable-searchpopup
  4. Once you enable the relevance search, it becomes your default searching option and can be changes later to categorized search from your personal settings options.
    personal-optionsdefault-search

Step 2: Enable entities for relevance search

  1. Go to Settings > Customizations > Customize the System.
  2. Under components, expand Entities, and click on Configure Relevance Search option.configure-relevance-search
  3. Select the entities and click okay. Publish all customizations.

Note: If you want to include any entity for relevance search, then below highlighted managed property of the entity should be set to True.

set-managed-properties

Step 3: Enable fields for relevance search

  1. Go to Settings à Customizations à Customize the System
  2. Go to Quick Find View of an entity for which you are configuring the relevance search.
  3. Select the find columns for the entity. By default, the fields you add in find columns of quick find view become part of relevance search. There is no limit on maximum number of fields allowed, but as explained earlier, there is limit on number of indexes involved in relevance search.

How to use relevance search?

  • Once you enable, relevance search becomes your default search. Just like global search, you can start searching your data and it will return data using relevance search.
    ordering-productshow-to-use
  • As you can see in above image, on left side you can filter the records and also select any particular record type to reduce your result set.
  • You can easily switch between relevance search and categorized search.swith-between-searches

Difference between Relevance Search & Categorized Search with few examples

  • Below screenshot shows my available data:

available-data

Case 1: I search for ‘need 110 item’ term.

  • Relevance search returns below datacase-1-1
  • Categorized search did not return anythingcase-1-2

Case 2: I search for term ‘ 10 items + need’

  • Relevance search returns below datacase-2-1
  • Categorized search did not return anything

case-2-2

References

Know more about relevance search

https://www.microsoft.com/en-us/dynamics/crm-customer-center/use-relevance-search-for-faster-comprehensive-search-results.aspx

https://technet.microsoft.com/library/5ec1a73d-06cb-4d40-adab-8bb293bcedbb.aspx

Learn how to search data with relevance search to get accurate data

https://docs.microsoft.com/en-us/rest/api/searchservice/Simple-query-syntax-in-Azure-Search

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s