使用 nlp 提取年龄相关信息

Question

我是 NLP 新手，我一直在尝试从原始文本中提取与年龄相关的信息。我用谷歌搜索并没有得到任何语言的可靠库来满足这个要求。如果我能在这方面得到任何帮助，那就太好了。我对任何语言都持开放态度，这不是一种约束。它也可以是 Java、Python 或任何其他语言。任何帮助将非常感激。提前致谢。干杯！

更新：

我尝试将斯坦福帮助中提到的注释器添加到我的 java 解析器中，但我遇到了以下异常：

    ERROR: cannot create CorefAnnotator!
    java.lang.RuntimeException: Error creating coreference system
    at 

 edu.stanford.nlp.scoref.StatisticalCorefSystem.fromProps(StatisticalCorefSystem.java:58)
    at edu.stanford.nlp.pipeline.CorefAnnotator.<init>(CorefAnnotator.java:66)
    at edu.stanford.nlp.pipeline.AnnotatorImplementations.coref(AnnotatorImplementations.java:220)
    at edu.stanford.nlp.pipeline.AnnotatorFactories.create(AnnotatorFactories.java:515)
    at edu.stanford.nlp.pipeline.AnnotatorPool.get(AnnotatorPool.java:85)
    at edu.stanford.nlp.pipeline.StanfordCoreNLP.construct(StanfordCoreNLP.java:375)
    at edu.stanford.nlp.pipeline.StanfordCoreNLP.<init>(StanfordCoreNLP.java:139)
    at edu.stanford.nlp.pipeline.StanfordCoreNLP.<init>(StanfordCoreNLP.java:135)
    at com.dateparser.SUtime.SUAgeParser.makeNumericPipeline(SUAgeParser.java:85)
    at com.dateparser.SUtime.SUAgeParser.<clinit>(SUAgeParser.java:60)
Caused by: java.lang.RuntimeException: Error initializing coref system
    at edu.stanford.nlp.scoref.StatisticalCorefSystem.<init>(StatisticalCorefSystem.java:36)
    at edu.stanford.nlp.scoref.ClusteringCorefSystem.<init>(ClusteringCorefSystem.java:24)
    at edu.stanford.nlp.scoref.StatisticalCorefSystem.fromProps(StatisticalCorefSystem.java:48)
    ... 9 more
Caused by: java.io.IOException: Unable to open "edu/stanford/nlp/models/hcoref/md-model.ser" as class path, filename or URL
    at edu.stanford.nlp.io.IOUtils.getInputStreamFromURLOrClasspathOrFileSystem(IOUtils.java:485)
    at edu.stanford.nlp.io.IOUtils.readObjectFromURLOrClasspathOrFileSystem(IOUtils.java:323)
    at edu.stanford.nlp.hcoref.md.DependencyCorefMentionFinder.<init>(DependencyCorefMentionFinder.java:38)
    at edu.stanford.nlp.hcoref.CorefDocMaker.getMentionFinder(CorefDocMaker.java:149)
    at edu.stanford.nlp.hcoref.CorefDocMaker.<init>(CorefDocMaker.java:61)
    at edu.stanford.nlp.scoref.StatisticalCorefSystem.<init>(StatisticalCorefSystem.java:34)
    ... 11 more

我升级到 1.6.0 版，并将 stanford-corenlp-models-current.jar 添加到类路径中。如果我遗漏了什么，请告诉我

更新 1：

升级到 3.9.1 后，异常已修复。但是我得到的输出是 per:duration 关系而不是 per:age

private static AnnotationPipeline makePipeline() {

    Properties props = new Properties();
    props.setProperty("annotators",
             "tokenize,ssplit,pos,lemma,ner,depparse,coref,kbp");
    StanfordCoreNLP pipeline = new StanfordCoreNLP(props);
    return pipeline;


}

public static void parse(String str) {
    try {
        Annotation doc = new Annotation(str);
        pipeline.annotate(doc);
        ArrayList<CoreMap> resultRelations = new ArrayList<CoreMap>();
        List<CoreMap> mentionsAnnotations = doc.get(MentionsAnnotation.class);
        for (CoreMap currentCoreMap : mentionsAnnotations) {
            System.out.println(currentCoreMap.get(TextAnnotation.class));
            System.out.println(currentCoreMap.get(CharacterOffsetBeginAnnotation.class));
            System.out.println(currentCoreMap.get(CharacterOffsetEndAnnotation.class));
            System.out.println(currentCoreMap.get(NamedEntityTagAnnotation.class));
        }
    } catch (Exception e) {

    }
}

这是正常行为还是我做错了什么？

Answer 1

您可能会发现 KBP 关系提取器很有用。

示例文本：

Joe Smith is 58 years old.

命令：

java -Xmx12g edu.stanford.nlp.pipeline.StanfordCoreNLP -annotators tokenize,ssplit,pos,lemma,ner,depparse,coref,kbp -file example.txt -outputFormat text

这应该通过 per:age 关系将 Joe Smith 附加到 58 years old。

使用 nlp 提取年龄相关信息

Extracting age related information from using nlp

nlp

nltk

stanford-nlp

opennlp