Search for phrases in Java

Question

I am trying to find collocations (PDF) in Java.

I know NLTK has a module collocations, but does not want to use Jython.

I looked at OpenNLP and GATE, but they didn't seem to have a collocation search.

Does anyone know what a free open source collocation search tool in Java is?

Answer 1

For me, the best ready-to-use xTract collating algorithm in DragonToolkit uses basic statistics collocation features such as lingpipe and more complex ones like POS marking

Answer 2

• http://www.cs.sunysb.edu/~algorith/implement/kolokacje/implement.shtml . ( java source code )

• LingPipe is a commercial library. http://alias-i.com/lingpipe/demos/tutorial/interestingPhrases/read-me.html

Answer 3

Through Ingersoll Grant Review of New Features in Mahout I discovered that Apache Mahout now has collocations . It is open source in Java, but it requires a Hadoop framework.