Return only high quality results in ElasticSearch query

[Shreya_Jain]

I'm sending search queries to my ES index and get multiple results back. A lot of times the results with lower scores are irrelevant and I want to remove these results and return only high-quality results (which mostly have a higher score).

My index contains 1000 documents of type text of 100-500 words. For example - 'AVENGERS: ENDGAME is set after Thanos' catastrophic use of the Infinity Stones randomly wiped out half of Earth's population in Avengers: Infinity War. Those left behind are desperate to do something -- anything -- to bring back their lost loved ones. But after an initial attempt -- with extra help from Captain Marvel -- creates more problems than solutions, the grieving, purposeless Avengers think all hope is lost.'

If the user searches for 'Captain Marvel aka Brie Larson kills Thanos in the movie', the above document should be returned as a result since it contains similar terms.

Currently, I am using min_score to set the threshold, but I know it's not best practice and the scores vary depending on the number of documents in the index (which will keep growing). So this approach doesn't seem scalable.

I also tried multiple ways of tuning the query to get high-quality results back, such as More Like This functionality -

"must": 
[{"more_like_this" : {
"fields" : field_list,
"like" : query_data,
"min_term_freq" : 1,
"max_query_terms" : 50,
"min_doc_freq" : 1,
"minimum_should_match" : '50%'}}]}}

But I'm still getting results with low scores like 1.5, whereas a good quality result usually has a score of 20. Is there a good way to tune the query further or adjust the min_score to be dynamic to only return highly relevant documents? Any help would be appreciated!

[polyfractal (Zachary Tong)]

It's true, min_score is usually not a good method for ensuring good results. The scores are all relative to the documents that match a particular query, so you can easily get two different queries that return two different ranges of scores.

Generally, better scoring is accomplished by boosting more important criteria. Some examples:

• Exact phrase matches (match_phrase query) are better than exact term matches (match), which are better than typo-tolerant matches (fuzzy query, fuzziness)
• Dynamically ignoring "stop words" can help reduce bad matches (cutoff_frequency)
• Matches in a short title field are usually better than a long paragraph field
• min_should_match parameter (e.g. on boolean should clauses) tend to improve results, since more exclusive queries == less off-target matches
• Similarly, terms_set is often better over terms since you can specify a min number that should match

Bad results are usually due to off-target or poorly matching query components. Things like fuzzy queries can generate a lot of poor matches, or a very large query which matches many random terms.

The Explain API can help show you how different terms generate scores in the different parts of your query.