I have created a new azure cognitive search in azure which has index related to articles. I want to implement in way that if users search for any article first response it should have suggested content, like in Google if we search for any keywords first it wil show some response with ads tag the same way I want to implement in azure cognitive search, and need some idea to implement Thanks in advance.
I have tried azure cognitive search suggester but its expecting exact or partial match for example, search keyword is ’always feeling tired" it will retrieve related articles which has keyword tired in it.
but what i expect:
if user search for keyword "always feeling tired" 1st response should be
- Tiredness (one of article name this should be suggested first, note this article name doesn’t have keyword tired/feeling)
2
Answers
You can use Azure Cognitive Search scoring profiles to modify the ranking of search results in order to get the desired behavior. Scoring profiles allow you to assign weights to specific fields, boosting the relevance of documents that match your criteria.
Here’s a high-level outline of the steps you can follow:
Please refer to these documentations for more details.
Configure relevance scoring – Azure Cognitive Search | Microsoft Learn
Add scoring profiles – Azure Cognitive Search | Microsoft Learn
The core of your question is that you expect a response that does not contain all the words in the user’s input. Unfortunately, the suggested answer with scoring profiles won’t solve this. Scoring profiles can only boost content that is already part of the result set.
To get your Tiredness-article included in your result set for your sample query, you have to use the new beta feature for semantic search. Regular keyword search will never work for your use case.
https://learn.microsoft.com/en-us/azure/search/semantic-search-overview