在 stanford-nlp 中训练 NER 模型

Question

我一直在尝试使用 stanford Core NLP。我希望训练我自己的 NER 模型。从 SO 的论坛和官方网站上描述了使用属性文件来做到这一点。我将如何通过 API 做到这一点？。

Properties props = new Properties();
props.setProperty("annotators", "tokenize, ssplit, pos, lemma, ner, parse, sentiment, regexner");
props.setProperty("regexner.mapping", "resources/customRegexNER.txt");

StanfordCoreNLP pipeline = new StanfordCoreNLP(props);      

String processedQuestion = "Who is the prime minister of Australia?"

//Annotation annotation = pipeline.process(processedQuestion);
Annotation document = new Annotation(processedQuestion);
pipeline.annotate(document);
List<CoreMap> sentences = document.get(SentencesAnnotation.class);
for (CoreMap sentence : sentences) {

    // To get the tokens for the parsed sentence
    for (CoreMap tokens : sentence.get(TokensAnnotation.class)) {           
        String token = tokens.get(TextAnnotation.class);
        String POS = tokens.get(PartOfSpeechAnnotation.class);      
        String NER = tokens.get(NamedEntityTagAnnotation.class);            
        String Sentiment = tokens.get(SentimentClass.class);            
        String lemma = tokens.get(LemmaAnnotation.class);

1. 如何以及在哪里添加道具文件？
2. N-gram 标记化（例如，将总理视为单个标记，稍后将此标记传递给 POS，NER 而不是传递两个标记（总理和部长））？

Answer 1

我认为它可以与该代码一起使用：

val props = new Properties()
  props.put("annotators", "tokenize, ssplit, pos, lemma, ner, regexner")
  props.put("ner.model", "/your/path/ner-model.ser.gz");
  val pipeline = new StanfordCoreNLP(props)