TF-IDF or BM25 works by extracting terms from queried text then Scoring the matching documents based on the presence of those terms and takes into account the length of the field.
I want queried text's term frequencies to influence the document scoring. For example:
query text: 'Define water and salts of water'
Here, water has tf 2.
I want the documents having tf of water closer to the tf of water in queried text to be scored higher. Similarly, for other terms present in the query.
By default elasticsearch considers the length of the field into scoring but not the queried text's term frequencies.
