RE: Re: Sentence classification with Lucene

Yes, something like lucene-classification [1].
But, there are multiple classifiers in this package.
Which one is better suited ? (Imagine I collect more samples per 
class... about... 30-40 samples per class)
Any good Java examples using these classifiers?


Another question:
in case I want my classification to work "semantically":

For example:
For the class "crypto" I can have these samples:
- crypto
- bitcoin
- stablecoin
- blockchain

and in case the input text contains "Eterium" - what happens in this 
case, will it match "crypto" ?

I mean, the models in lucene-classification: as far as I understand - 
they do not have knowledge about semantic similarity between words, right?




On 2025/02/19 13:45:52 Tommaso Teofili wrote:
 > Hi,
 >
 > if you have 30 classes with 10 samples per class, I'd say that's not an
 > optimal distribution.
 > Apart from that, you may use one of the text classifiers from
 > lucene-classification [1], is anything like this what you had in mind?
 > Alternatively you can also do things outside of Lucene and use Lucene 
only,
 > for example, to store vectors and find nearest neighbors.
 >
 > Regards,
 > Tommaso
 >
 > [1] :
 > 
https://lucene.apache.org/core/10_1_0/classification/org/apache/lucene/classification/package-summary.html
 >
 > On Mon, 17 Feb 2025 at 16:15, Dmitri Geller <[email protected]> wrote:
 >
 > > Hi all, I would like to classify a sentence into one or two categories.
 > > I see this classification roughly this way:
 > >
 > > ```
 > > unknown:
 > > example1
 > > example2
 > > ...
 > > exampleN
 > >
 > > class1:
 > > example1
 > > example2
 > > ...
 > > exampleN
 > >
 > > class2:
 > > example1
 > > example2
 > > ...
 > > exampleN
 > >
 > > ...
 > >
 > > classN:
 > > example1
 > > example2
 > > ...
 > > exampleN
 > >
 > > ...
 > > ```
 > >
 > > There are about 25-30 classes.
 > > About 10-30 examples per class.
 > > One sentence can get one or two classes assigned
 > >
 > > As far as I understand: this can be done with Lucene Core, should be
 > > quite a standard functionality.
 > > Can you point me to a Java example for this?
 > >
 > > Thanks in advance!
 > >
 > >
 > > ---------------------------------------------------------------------
 > > To unsubscribe, e-mail: [email protected]
 > > For additional commands, e-mail: [email protected]
 > >
 > >
 >