Is there a plan for elasticsearch to support LTR (learning to rank)

Elastic Stack Elasticsearch
1

Hi,
Solr 6.4 enabled support for LTR machine learning model with the goal of improved search relevancy . I was wondering if elasticsearch or prelert will support this in future?

" Machine Learning:

Configurable Learning-To-Rank (LTR) support: upload feature definitions, extract feature values, upload your own machine learnt models and use them to rerank search results. " - from http://lucene.apache.org/solr/news.html
-Sujith Joseph

3

Hey,

I only took a very quick glance at the above link, and I do not think you will find this exact feature in Elasticsearch, but you should take a look at others which may offer ways to implement similar functionality.

You might want to take a look at the rescore query as well as in this work in progress branch (which is not yet released) called rank-eval

--Alex

4

Thanks Alex. LTR provides a personalized relevancy experience per user. rank-eval seems like an alternative for NDCG based rank scorer implementation.

5

Hey @Sujith_Joseph in fact we just released a Learning to Rank plugin, you can read more in this blog post. Here's the github repo.

6

Thanks for sharing, Doug. Great Work! Read about this is reddit today. You might also be interested in Vowpal Wabbit (https://github.com/JohnLangford/vowpal_wabbit/wiki/Learning-to-Search-Sub-System), if you are looking for features to add to your ranklib based implementation.
-Sujith