How do scoring results generate points in Azure Search?

Question

I want to add a rating profile to my index on Azure Search. More specifically, each document in my index has a weighttype field Edm.Double, and I want to increase them according to this value. I do not want to simply sort them in relation to weight, because the relevance of the search term is also important.

So, just to test this, I created an evaluation profile with an amplitude function with a boost value of 1000 (just to find out how I did it), linear interpolation, initial value 0 and final value 1. What am I expecting that the boost value will be added to overall search rating. Thus, a document with a weight of 0.5 will receive an increase of 500, while a document with a weight of 0.125 will receive an increase of 125. However, the resulting ratings were not nearly as intuitive.

In this case, I have a few questions:

1) How is the function score created in this case? I have documents with weights close to each other (say, 0.5465 and 0.5419), but the differences between their final grades are about 100-150, while I expect it to be around 4-5.

2) How are indicators and weights aggregated into the final result for each search result?

Answer 1

Thanks for the information provided. What were the basic meanings of these two documents?

, , , . , , , @search.score , 0,5 0,2, 0,5465 0,5419 . 0, 1, 1000. , , :

1: base search_score (0.5) * boost_factor (1000) * ( (0,5465) - min (0))/max - min (1) = final_search_score (273.25)

2: base_search_score (0.2) * boost_factor (1000) * ( (0,5419) - min (0))/max - min (1) = final_search_score (108.38)

, , , , . !