aida-disambiguation

import mpi.aida.config.settings.PreparationSettings;
import mpi.aida.preparator.Preparator;
import mpi.aida.data.PreparedInput;
import mpi.aida.config.settings.DisambiguationSettings;

import java.io.FileInputStream;
import java.io.FileOutputStream;
import java.io.OutputStream;
import java.io.PrintStream;
import java.util.HashSet;
import java.util.Map;
import java.util.Set;

import org.apache.commons.io.IOUtils;

import mpi.aida.Disambiguator;
import mpi.aida.access.DataAccess;
import mpi.aida.data.DisambiguationResults;
import mpi.aida.data.Entities;
import mpi.aida.data.EntityMetaData;
import mpi.aida.data.KBIdentifiedEntity;
import mpi.aida.util.htmloutput.ResultMention;
import mpi.aida.data.ResultEntity;
import mpi.aida.config.settings.preparation.StanfordHybridPreparationSettings;
import mpi.aida.config.settings.disambiguation.CocktailPartyDisambiguationSettings;
import mpi.aida.config.settings.disambiguation.FastCocktailPartyDisambiguationSettings;
import mpi.aida.config.settings.disambiguation.FastLocalKeyphraseBasedDisambiguationSettings;
import mpi.aida.config.settings.disambiguation.ImportanceOnlyDisambiguationSettings;
import mpi.aida.config.settings.disambiguation.LocalKeyphraseBasedDisambiguationSettings;
import mpi.aida.config.settings.disambiguation.PriorOnlyDisambiguationSettings;


public class test_2 {

	public static void main(String[] args) throws Exception  {
		// Define the input.
		//FileInputStream inputStream = new FileInputStream("D:/files_final_inputs/files/training-monolingual.tokenized.shuffled/news.txt");
		//FileInputStream inputStream = new FileInputStream(args[0]);
		//final OutputStream os = new FileOutputStream(args[1]);

		FileInputStream inputStream = new FileInputStream("D:/try2.txt");
		final OutputStream os = new FileOutputStream("D:/tryOutFast.txt");
		final PrintStream printStream = new PrintStream(os);


		try {
		    String everything = IOUtils.toString(inputStream);
		    PreparationSettings prepSettings = new StanfordHybridPreparationSettings();
			Preparator p = new Preparator();
			PreparedInput input = p.prepare(everything, prepSettings);

			DisambiguationSettings disSettings = new CocktailPartyDisambiguationSettings();
			//disSettings.setMaxEntityRank(0.15);
			//disSettings.getSimilaritySettings().setMaxEntityKeyphraseCount(100);
			// Below is new.
			//disSettings.getCoherenceSimilaritySetting().setMaxEntityKeyphraseCount(100);

			//DisambiguationSettings disSettings = new ImportanceOnlyDisambiguationSettings();
			Disambiguator d = new Disambiguator(input, disSettings);
			DisambiguationResults results = d.disambiguate();


			// Print the disambiguation results.
			Set<KBIdentifiedEntity> entities = new HashSet<KBIdentifiedEntity>();
			for (mpi.aida.data.ResultMention rm : results.getResultMentions()) {
			  entities.add(results.getBestEntity(rm).getKbEntity());
			}

			Map<KBIdentifiedEntity, EntityMetaData> entitiesMetaData =
			  DataAccess.getEntitiesMetaData(entities);

			for (mpi.aida.data.ResultMention rm : results.getResultMentions()) {
			  KBIdentifiedEntity entity = results.getBestEntity(rm).getKbEntity();
			  EntityMetaData entityMetaData = entitiesMetaData.get(entity);

			  if (Entities.isOokbEntity(entity)) {
			    //System.out.println("\t" + rm + "\t NO MATCHING ENTITY");
			  } else {
				  int l=rm.getCharacterLength()+rm.getCharacterOffset();
				  printStream.print( rm.getCharacterOffset() + " " + l + " " + rm.getMention() + ":" + entity.getIdentifier() + "\r\n");
			  }
			}
		} finally {
		    inputStream.close();
		    printStream.close();
		    os.close();
		}

	}

}