Neo4j Full Text Search Extra

Installation

• Copy the fts-extra-1.0.0.jar in the plugins directory of Neo4j.
• Restart the Neo4j database

Custom Analyzers

Standard No Stopwords

The default, standard analyzer but without stopwords. Tokenizes on non-letter and filters out English stop words and punctuation. Does no stemming, but takes care to keep likely product names, URLs and email addresses as single terms.

When applicable ?

The following text : To be or not to be will have no tokens indexed, because they're all stopwords and thus will never be returned by any search query.

Usage

Create a FTS index using the analyzer :

CALL db.index.fulltext.createNodeIndex('no-stopwords', ['Post'], ['text'], {analyzer:'standard-no-stopwords'})

Create some data :

CREATE (n:Post {text: 'To be or not to be'})

Query data :

CALL db.index.fulltext.queryNodes('no-stopwords', 'to be or not to be')

TODO

Move Czech Analyzer here