My goal is to ensure that only those document whose value exactly matches
(including word position) some part/substring of the keywords.
To illustrate, I have a index/type that contains the following documents:
----- ----- ---------------------
concept entity value
----- ----- ---------------------
part part Mirror Motor
part brand Honda Collection
part brand Honda Apparel
vehicle make Honda
part brand Honda by Moose
part part Mirror
part part Telescoping Mirror
part part Mirror Back
part part Mirror Bra
part part Mirror Hardware
part part Mirror Mount
part part Spot Mirror
I want to ensure that when the keywords is "Honda Civic Mirror" only the
following documents are returned (or at least gets the highest _score
value)::
vehicle make Honda
part part Mirror
If the keywords is "Mirror Mount for Honda", only the following documents
are to be returned (or at least gets the highest _score value):
vehicle make Honda
part part Mirror Mount
Is there any index and search analyzer that will allow me to achieve this
goal?
My goal is to ensure that only those document whose value exactly matches
(including word position) some part/substring of the keywords.
To illustrate, I have a index/type that contains the following documents:
----- ----- ---------------------
concept entity value
----- ----- ---------------------
part part Mirror Motor
part brand Honda Collection
part brand Honda Apparel
vehicle make Honda
part brand Honda by Moose
part part Mirror
part part Telescoping Mirror
part part Mirror Back
part part Mirror Bra
part part Mirror Hardware
part part Mirror Mount
part part Spot Mirror
I want to ensure that when the keywords is "Honda Civic Mirror" only the
following documents are returned (or at least gets the highest _score
value)::
vehicle make Honda
part part Mirror
If the keywords is "Mirror Mount for Honda", only the following documents
are to be returned (or at least gets the highest _score value):
vehicle make Honda
part part Mirror Mount
Is there any index and search analyzer that will allow me to achieve this
goal?
Normal text query works. I actually use query_string. However when
executing a query based on supplied keyword, the engine will actually try
to whole or parts of the keywords against the searchable documents field.
Right now, with the current ES index that I have, searching for "Honda
Civic Mirror" produces the following results:
concept entity value _score
vehicle model Civic 0.638050
vehicle make Honda 0.579641
part brand Honda Collection 0.526310
part brand Honda Apparel 0.524273
part brand Honda by Moose 0.507186
part part Mirror 0.476971
vehicle model Civic del Sol 0.449377
part part Mirror Arm 0.417349
part part Mirror Switch 0.417349
I was hoping that to find a solution on excluding "part/brand = Honda
Collection" to be not part of the list because in my keyword there is no
word "Collection".
I am looking looking for the reverse of search ( but not using
percolation). where document fields are checked to make sure that their
value is a substring (or exact match) of the keywords. This should ensure
that "Honda Collection" or "Mirror Arm" will not appear because "Arm" or
"Collection" is not part of my example keywords.
concept entity value
vehicle model Civic
vehicle make Honda
part part Mirror
Wwhat I am looking for here is hard to explain, I apologize. I'm interested
if someone out there has done something similar to this.
My goal is to ensure that only those document whose value exactly matches
(including word position) some part/substring of the keywords.
To illustrate, I have a index/type that contains the following documents:
----- ----- ---------------------
concept entity value
----- ----- ---------------------
part part Mirror Motor
part brand Honda Collection
part brand Honda Apparel
vehicle make Honda
part brand Honda by Moose
part part Mirror
part part Telescoping Mirror
part part Mirror Back
part part Mirror Bra
part part Mirror Hardware
part part Mirror Mount
part part Spot Mirror
I want to ensure that when the keywords is "Honda Civic Mirror" only the
following documents are returned (or at least gets the highest _score
value)::
vehicle make Honda
part part Mirror
If the keywords is "Mirror Mount for Honda", only the following documents
are to be returned (or at least gets the highest _score value):
vehicle make Honda
part part Mirror Mount
Is there any index and search analyzer that will allow me to achieve this
goal?
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.