使用 LingPipe 在 Gate 中进行命名实体识别

Named Entity recognition in Gate using LingPipe

我正在使用 GATE NLP 处理我的文档,我想使用实体名称作为候选标签 在 Gate 中有 OpenNLPLingPipe 当我阅读答案表时 here @Shashikant Kore 回答他说

if you have the sentence "My friend Joe Smith went to the Walmart store", OpenNLP identifies two named entities - "Joe Smith" and "Walmart". I couldn't get it tag "Joe Smith" as Person and "Walmart" as Organization.

并建议使用 LingPipe 所以我使用了 Gate NLP 中提供的 LingPipe 喜欢这里

SerialAnalyserController pipeline = (SerialAnalyserController) Factory.createResource("gate.creole.SerialAnalyserController");
        pipeline.add((ProcessingResource) Factory.createResource("gate.lingpipe.TokenizerPR"));
        pipeline.add((ProcessingResource) Factory.createResource("gate.lingpipe.NamedEntityRecognizerPR"));
        pipeline.add((ProcessingResource) Factory.createResource("gate.lingpipe.POSTaggerPR"));
        pipeline.add((ProcessingResource) Factory.createResource("gate.lingpipe.SentenceSplitterPR"));
        Corpus corpus = Factory.newCorpus("SegmenterCorpus");
        Document document = Factory.newDocument(handler.toString());
        corpus.add(document); 
        pipeline.setCorpus(corpus); 
        pipeline.execute();

但是,当我 运行 我的程序出现此异常时

Exception in thread "main" gate.creole.ResourceInstantiationException: No model file provided! at gate.lingpipe.NamedEntityRecognizerPR.init(NamedEntityRecognizerPR.java:55) at gate.lingpipe.NamedEntityRecognizerPR.init(NamedEntityRecognizerPR.java:55)

没有提供模型文件是什么意思?? 抱歉,因为我在问这个问题,但我对这个领域完全陌生 我刚刚了解了 ANNIE,当我用它来提取 POS 标签时,它不需要任何文件 有帮助吗??

终于对我有用了。

我在 GATE GUI 中尝试了 File -> Ready Made Applications -> LingPipe -> LingPipe IE System。并且已经加载成功。

工作模型是: $gatehome$plugins/LingPipe/resources/models/ne-en-news-muc6.AbstractCharLmRescoringChunker

gate.lingpipe.NamedEntityRecognizerPR 的完整相关配置是:

...    
<string>modelFileUrl</string>
<gate.util.persistence.PersistenceManager-URLHolder>
  <urlString>$gatehome$plugins/LingPipe/resources/models/ne-en-news-muc6.AbstractCharLmRescoringChunker</urlString>
</gate.util.persistence.PersistenceManager-URLHolder>
...