The data set stored in ES is in the form of Q & A.
The Question data is indexed to query later with the important keywords and its synonyms.
When the data becomes more and similar words are repeated in the multiple questions, ES is giving same _score value to the multiple questions, resulting wrong search result. (expectation is with a different query needed to be popped as highest _score value).
Q1. What is MMM?
Q2. What is MMM size?
Q3. What is the compatibility for MMM?
Is there anything needs to be tuned in ES parameters or any other methods to give a proper _score for each questions, since all are having different meaning and the result expectation is w.r.t that question?