Plural vs Singular names

[martinschaef]

Going through a bunch of results, I think we have to distinguish plural from singular names for clustering of classes and fields.
Is there a good way to determine if a word (or class name) is singular or plural that works well with the stemming that we do (apart from checking if it ends on 's' before stemming)?

[hsanchez]

"Is there a good way to determine if a word (or class name) is singular or plural that works well with the stemming that we do (apart from checking if it ends on 's' before stemming)?"

You can play with the class Noun (which is part of the library that I am using to extract frequent and typical words). Here is the source code (latest version): https://github.com/vesperin/introspector/blob/cleanup2/src/main/java/com/vesperin/text/nouns/Noun.java

Example

System.out.println(Noun.get().isSingular("Trial") => true
System.out.println(Noun.get().isPlural("Trial") => false
// or

if(Noun.get().isSingular("Trial")){
  System.out.println(Noun.get().pluralOf("Trial")); => Trials
  System.out.println(Noun.get().singularOf("Trial")); => Trial
}

Side comment: Maybe the name of this class should "Term". instead of Noun..

[hsanchez]

You can use it within the clusterer project.