# Document Similarity

**URL:** <https://discuss.elastic.co/t/document-similarity/11713>\
**Category:** Elasticsearch\
**Created:** [April 27, 2013, 6:42am UTC](https://discuss.elastic.co/t/document-similarity/11713 "2013-04-27T06:42:36Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Darran\_Kartaschew](https://avatars.discourse-cdn.com/v4/letter/d/5f9b8f/32.png) [@Darran\_Kartaschew](https://discuss.elastic.co/u/Darran_Kartaschew)\
**Post date:** [April 27, 2013, 6:42am UTC](https://discuss.elastic.co/t/document-similarity/11713/1 "2013-04-27T06:42:36Z")

</div>

Hi Guys,

Just like to find out what options there are to determine how similar two  
documents are when both documents are stored in ES?

While I can use the _more\_like\_this\_field_ query to determine a score based  
on a query, is there anyway that I can provide two document IDs and get the  
similarity score between them?

I have found this:  
[http://stackoverflow.com/questions/1844194/get-cosine-similarity-between-two-documents-in-luceneusing](http://stackoverflow.com/questions/1844194/get-cosine-similarity-between-two-documents-in-luceneusing) Lucene to compare two documents, so I can create my own query if  
required, (I'm maintaining my own ES fork for the time being), but was  
hoping there was something already in ES that could allow me to do this?

The context for wanting this, is that I need to build a similarity matrix  
of all documents stored in a single index in ES to determine how similar  
all documents are to each other. (This is to aid a machine learning  
activity). If I can have a query that does only two documents at a time,  
this is fine, but the creation of a similarity matrix directly in ES would  
be great! (Note: in my context, I only dealing with a small number of  
documents, less than 5000, but they are very large documents).

Kind regards

Darran

PS. My fork can be found here:

> **[dkartaschew / elasticsearch](https://bitbucket.org/dkartaschew/elasticsearch)**
>
> Modification of upstream elasticsearch for use with Bioinformatics

--  
You received this message because you are subscribed to the Google Groups "elasticsearch" group.  
To unsubscribe from this group and stop receiving emails from it, send an email to [elasticsearch+unsubscribe@googlegroups.com](mailto:elasticsearch+unsubscribe@googlegroups.com).  
For more options, visit [https://groups.google.com/groups/opt\_out](https://groups.google.com/groups/opt_out).

---

<div class="post-metadata">

**Author:** ![system](https://us1.discourse-cdn.com/elastic/original/3X/1/a/1ac57faf039f6b580b3f104ef42a2a89e41014de.png) [@system](https://discuss.elastic.co/u/system)\
**Post date:** [July 6, 2017, 2:39am UTC](https://discuss.elastic.co/t/document-similarity/11713/2 "2017-07-06T02:39:25Z")

</div>