search - Document Similarity in ElasticSearch -
I want to calculate the similarity between the two documents indexed in elasticsearch I know that this word is used by vectors in Lucene What can be done is the direct way to do this?
I found that there is one similarity module that does this:
How do I integrate it into my system? I'm using the Pilichtishchurch to call elasticsearch commands, but if I want to open the REST API for equality to use.
I think elastic search documents can be easily misplaced.
There is no comparison of "equality" documents or fields, but there is a mechanism for running documents that match the terms that match the query. / P>
Documentation states:
Symmetry algorithm that supports the alißsecretor, potential model based on word distribution in cache (index) Are there.
Regarding the word vector, it can also be misinterpreted.
See statistics for the conditions of a document called "word vectors" which can be easily inquired. It seems that any similarity measurement in the term vectors must be posted in your post-query. Word Vector State Documents:
If you want a demonstrator (fast) similarity metric on a very large corpus, then you consider the low-rank Embeds your documents stored in an index for the closest neighbor searches. After your KNN lookup, which makes the candidate set very low, you can calculate more expensive metrics for ranking.
This is an excellent resource for evaluating estimated KNN solutions:
Comments
Post a Comment