package edu.stanford.nlp.coref.data; import java.util.List; import java.util.Map; import edu.stanford.nlp.coref.docreader.CoNLLDocumentReader.CoNLLDocument; import edu.stanford.nlp.pipeline.Annotation; /** * Input document read from input source (CoNLL, ACE, MUC, or raw text) * Stores Annotation, gold info (optional) and additional document information (optional) * * @author heeyoung * */ public class InputDoc { public Annotation annotation; /** * Additional document information possibly useful for coref. * (e.g., is this dialog? the source of article, etc) * We can use this as features for coref system. * This is optional. */ public Map<String, String> docInfo; /** * Gold mentions with coreference information for evaluation. * This is optional. */ public List<List<Mention>> goldMentions; /** optional for CoNLL document */ public CoNLLDocument conllDoc; public InputDoc(Annotation anno) { this(anno, null, null, null); } public InputDoc(Annotation anno, Map<String, String> docInfo) { this(anno, docInfo, null, null); } public InputDoc(Annotation anno, Map<String, String> docInfo, List<List<Mention>> goldMentions) { this(anno, docInfo, goldMentions, null); } public InputDoc(Annotation anno, Map<String, String> docInfo, List<List<Mention>> goldMentions, CoNLLDocument conllDoc) { this.annotation = anno; this.docInfo = docInfo; this.goldMentions = goldMentions; this.conllDoc = conllDoc; } }