chalk nlp

WordTagger

package nlpapp

import chalk.tools.sentdetect.SentenceDetectorME
import chalk.tools.sentdetect.SentenceModel
import java.io.FileInputStream
import chalk.tools.postag.POSModel
import chalk.tools.postag.POSTaggerME
import chalk.tools.tokenize.TokenizerModel
import chalk.tools.tokenize.TokenizerME

object WordTagger extends App {

  val tweetString = "CSSCorp is a global company. " +
    "The company's labs entity is providing services in Cloud and BigData Technologies."

  val fis = (x: String) => new FileInputStream(x)

  val model = (x: FileInputStream) => new SentenceModel(x)
  val detector = (x: SentenceModel) => new SentenceDetectorME(x)

  val tokenModel = (x: FileInputStream) => new TokenizerModel(x)
  val tokenDetector = (x: TokenizerModel) => new TokenizerME(x)

  val posmodel = (x: FileInputStream) => new POSModel(x)
  val posdetector = (x: POSModel) => new POSTaggerME(x)

  def splitSentence = {
    val senDet = detector(model(fis(this.getClass().getResource("/en-sent.bin").getPath())))
    def findSentences(sme: SentenceDetectorME) = (x: String) => sme.sentDetect(x)
    findSentences(senDet)
  }

  val sentences = splitSentence(tweetString)

  val postagr = posdetector(posmodel(fis(this.getClass().getResource("/en-pos-maxent.bin").getPath()))).tag(sentences(1))

  println(postagr)

  // output
  // The/DT company's/NNS labs/NNS entity/NN is/VBZ providing/VBG services/NNS in/IN Cloud/NNP and/CC BigData/NNP Technologies./NNP

}