public void StanfordCoreNlpDemoThatChangeCurrentDirectory()
{
const string Text = "Kosgi Santosh sent an email to Stanford University. He didn't get a reply.";
// Annotation pipeline configuration
var props = new Properties();
props.setProperty("annotators", "tokenize, ssplit, pos, lemma, ner, parse, dcoref");
props.setProperty("sutime.binders", "0");
// we should change current directory so StanfordCoreNLP could find all the model files
var curDir = Environment.CurrentDirectory;
Directory.SetCurrentDirectory(Config.JarRoot);
var pipeline = new edu.stanford.nlp.pipeline.StanfordCoreNLP(props);
Directory.SetCurrentDirectory(curDir);
// Annotation
var annotation = new Annotation(Text);
pipeline.annotate(annotation);
// Result - Pretty Print
using (var stream = new ByteArrayOutputStream())
{
pipeline.prettyPrint(annotation, new PrintWriter(stream));
Console.WriteLine(stream.toString());
}
this.CustomAnnotationPrint(annotation);
}
public void StanfordCoreNlpDemoThatChangeCurrentDirectory()
{
const string Text = "Kosgi Santosh sent an email to Stanford University. He didn't get a reply.";
// Annotation pipeline configuration
var props = new Properties();
props.setProperty("annotators", "tokenize, ssplit, pos, lemma, ner, parse, dcoref");
props.setProperty("sutime.binders", "0");
// we should change current directory so StanfordCoreNLP could find all the model files
var curDir = Environment.CurrentDirectory;
Directory.SetCurrentDirectory(Config.JarRoot);
var pipeline = new edu.stanford.nlp.pipeline.StanfordCoreNLP(props);
Directory.SetCurrentDirectory(curDir);
// Annotation
var annotation = new Annotation(Text);
pipeline.annotate(annotation);
// Result - Pretty Print
using (var stream = new ByteArrayOutputStream())
{
pipeline.prettyPrint(annotation, new PrintWriter(stream));
Console.WriteLine(stream.toString());
}
this.CustomAnnotationPrint(annotation);
}
public static void Execute(string option, string text, bool disableLogging = true)
{
if (disableLogging)
{
RedwoodConfiguration.current().clear().apply();
}
var jarRoot = @"../../../data/paket-files/stanford-corenlp-3.9.1-models/";
var props = new Properties();
props.setProperty("annotators", option);
props.setProperty("ner.useSUTime", "0");
// We should change current directory, so StanfordCoreNLP could find all the model files automatically
var curDir = Environment.CurrentDirectory;
Directory.SetCurrentDirectory(jarRoot);
var pipeline = new StanfordNLP.StanfordCoreNLP(props);
Directory.SetCurrentDirectory(curDir);
// Annotation
var annotation = new StanfordNLP.Annotation(text);
pipeline.annotate(annotation);
//get sentencesAnnotation to get sentences
var sentencesAnnotation = new CoreAnnotations.SentencesAnnotation().getClass();
//get tokensAnnotaion to get tokens in each sentence
var tokensAnnotaion = new CoreAnnotations.TokensAnnotation().getClass();
//get posAnnotation to get POS result of each token
var posAnnotation = new CoreAnnotations.PartOfSpeechAnnotation().getClass();
//get nerAnnotation to get NER result of each token
var nerAnnotaion = new CoreAnnotations.NamedEntityTagAnnotation().getClass();
var deparseAnnotation = new TreeCoreAnnotations.TreeAnnotation().getClass();
//deparseAnnotation = new TypedDependency().getClass();
var sentences = annotation.get(sentencesAnnotation) as ArrayList;
foreach (CoreMap sentence in sentences.toArray())
{
var tokens = (ArrayList)sentence.get(tokensAnnotaion);
Console.WriteLine("Token-POS-NER: ");
foreach (CoreLabel token in tokens)
{
Console.Write($"{token.value()}-{token.get(posAnnotation)}-{token.get(nerAnnotaion)} ");
}
Console.WriteLine("\n\n\n");
var parsedText = (Tree)sentence.get(deparseAnnotation);
if (parsedText != null)
{
Console.WriteLine("Parsed Text: ");
new TreePrint("penn,typedDependenciesCollapsed").printTree(parsedText);
}
}
}
public void StanfordCoreNlpDemoManualConfiguration()
{
Console.WriteLine(Environment.CurrentDirectory);
const string Text = "Kosgi Santosh sent an email to Stanford University. He didn't get a reply.";
// Annotation pipeline configuration
var props = new Properties();
props.setProperty("annotators", "tokenize, ssplit, pos, lemma, ner, parse, dcoref");
props.setProperty("pos.model", Config.GetModel(@"pos-tagger\english-bidirectional\english-bidirectional-distsim.tagger"));
props.setProperty("ner.model", Config.GetModel(@"ner\english.all.3class.distsim.crf.ser.gz"));
props.setProperty("parse.model", Config.GetModel(@"lexparser\englishPCFG.ser.gz"));
props.setProperty("dcoref.demonym", Config.GetModel(@"dcoref\demonyms.txt"));
props.setProperty("dcoref.states", Config.GetModel(@"dcoref\state-abbreviations.txt"));
props.setProperty("dcoref.animate", Config.GetModel(@"dcoref\animate.unigrams.txt"));
props.setProperty("dcoref.inanimate", Config.GetModel(@"dcoref\inanimate.unigrams.txt"));
props.setProperty("dcoref.male", Config.GetModel(@"dcoref\male.unigrams.txt"));
props.setProperty("dcoref.neutral", Config.GetModel(@"dcoref\neutral.unigrams.txt"));
props.setProperty("dcoref.female", Config.GetModel(@"dcoref\female.unigrams.txt"));
props.setProperty("dcoref.plural", Config.GetModel(@"dcoref\plural.unigrams.txt"));
props.setProperty("dcoref.singular", Config.GetModel(@"dcoref\singular.unigrams.txt"));
props.setProperty("dcoref.countries", Config.GetModel(@"dcoref\countries"));
props.setProperty("dcoref.extra.gender", Config.GetModel(@"dcoref\namegender.combine.txt"));
props.setProperty("dcoref.states.provinces", Config.GetModel(@"dcoref\statesandprovinces"));
props.setProperty("dcoref.singleton.predictor", Config.GetModel(@"dcoref\singleton.predictor.ser"));
props.setProperty("dcoref.big.gender.number", Config.GetModel(@"dcoref\gender.data.gz"));
var sutimeRules = new[] {
Config.GetModel(@"sutime\defs.sutime.txt"),
Config.GetModel(@"sutime\english.holidays.sutime.txt"),
Config.GetModel(@"sutime\english.sutime.txt")
};
props.setProperty("sutime.rules", String.Join(",", sutimeRules));
props.setProperty("sutime.binders", "0");
var pipeline = new edu.stanford.nlp.pipeline.StanfordCoreNLP(props);
// Annotation
var annotation = new Annotation(Text);
pipeline.annotate(annotation);
// Result - Pretty Print
using (var stream = new ByteArrayOutputStream())
{
pipeline.prettyPrint(annotation, new PrintWriter(stream));
Console.WriteLine(stream.toString());
}
this.CustomAnnotationPrint(annotation);
}
private void button1_Click(object sender, EventArgs e)
{
text = InputText.Text.ToString();
option = OutputType.Text.ToString();
if (text != "")
{
PWF = new PleaseWaitForm();
PWF.Show();
this.Visible = false;
var annotation = new Annotation(text);
pipeline.annotate(annotation);
using (var stream = new ByteArrayOutputStream())
{
if (option == "Pretty")
{
pipeline.prettyPrint(annotation, new PrintWriter(stream));
}
else if (option == "Json")
{
pipeline.jsonPrint(annotation, new PrintWriter(stream));
}
else if (option == "XML")
{
pipeline.xmlPrint(annotation, new PrintWriter(stream));
}
else if (option == "Conll")
{
pipeline.conllPrint(annotation, new PrintWriter(stream));
}
else
{
pipeline.conllPrint(annotation, new PrintWriter(stream));
}
OutputText.Text = stream.ToString();
stream.close();
}
PWF.Close();
this.Visible = true;
}
}
public static void FindCoreferenceResolution(string text)
{
//var text = "Barack Obama nominated Hillary Rodham Clinton as his secretary of state on Monday. He chose her because she had foreign affairs experience as a former First Lady.";
var jarRoot = @"../../../data/paket-files/stanford-corenlp-3.9.1-models/";
var propsFile = Path.Combine(jarRoot, "StanfordCoreNLP.properties");
var props = new Properties();
//props.setProperty("annotators", "coref");
props.load(new FileReader(propsFile));
props.setProperty("ner.useSUTime", "0");
// We should change current directory, so StanfordCoreNLP could find all the model files automatically
var curDir = Environment.CurrentDirectory;
Directory.SetCurrentDirectory(jarRoot);
var pipeline = new StanfordNLP.StanfordCoreNLP(props);
Directory.SetCurrentDirectory(curDir);
// Annotation
var annotation = new Annotation(text);
pipeline.annotate(annotation);
// Result (detail)
//using (var stream = new ByteArrayOutputStream())
//{
// pipeline.prettyPrint(annotation, new PrintWriter(stream));
// Console.WriteLine(stream.toString());
// stream.close();
//}
// Result
var corefChainAnnotation = new CorefCoreAnnotations.CorefChainAnnotation().getClass();
Console.WriteLine("---");
Console.WriteLine("coref chains");
var corefChain = annotation.get(corefChainAnnotation) as Map;
foreach (CorefChain cc in corefChain.values().toArray())
{
Console.WriteLine($"\t{cc}");
}
}
public void StanfordCoreNlpDemoManualConfiguration()
{
Console.WriteLine(Environment.CurrentDirectory);
const string Text = "Kosgi Santosh sent an email to Stanford University. He didn't get a reply.";
// Annotation pipeline configuration
var props = new Properties();
props.setProperty("annotators", "tokenize, ssplit, pos, lemma, ner, parse, dcoref");
props.setProperty("pos.model", Config.GetModel(@"pos-tagger\english-bidirectional\english-bidirectional-distsim.tagger"));
props.setProperty("ner.model", Config.GetModel(@"ner\english.all.3class.distsim.crf.ser.gz"));
props.setProperty("parse.model", Config.GetModel(@"lexparser\englishPCFG.ser.gz"));
props.setProperty("dcoref.demonym", Config.GetModel(@"dcoref\demonyms.txt"));
props.setProperty("dcoref.states", Config.GetModel(@"dcoref\state-abbreviations.txt"));
props.setProperty("dcoref.animate", Config.GetModel(@"dcoref\animate.unigrams.txt"));
props.setProperty("dcoref.inanimate", Config.GetModel(@"dcoref\inanimate.unigrams.txt"));
props.setProperty("dcoref.male", Config.GetModel(@"dcoref\male.unigrams.txt"));
props.setProperty("dcoref.neutral", Config.GetModel(@"dcoref\neutral.unigrams.txt"));
props.setProperty("dcoref.female", Config.GetModel(@"dcoref\female.unigrams.txt"));
props.setProperty("dcoref.plural", Config.GetModel(@"dcoref\plural.unigrams.txt"));
props.setProperty("dcoref.singular", Config.GetModel(@"dcoref\singular.unigrams.txt"));
props.setProperty("dcoref.countries", Config.GetModel(@"dcoref\countries"));
props.setProperty("dcoref.extra.gender", Config.GetModel(@"dcoref\namegender.combine.txt"));
props.setProperty("dcoref.states.provinces", Config.GetModel(@"dcoref\statesandprovinces"));
props.setProperty("dcoref.singleton.predictor", Config.GetModel(@"dcoref\singleton.predictor.ser"));
props.setProperty("dcoref.big.gender.number", Config.GetModel(@"dcoref\gender.data.gz"));
var sutimeRules = new[] {
Config.GetModel(@"sutime\defs.sutime.txt"),
Config.GetModel(@"sutime\english.holidays.sutime.txt"),
Config.GetModel(@"sutime\english.sutime.txt")
};
props.setProperty("sutime.rules", String.Join(",", sutimeRules));
props.setProperty("sutime.binders", "0");
var pipeline = new edu.stanford.nlp.pipeline.StanfordCoreNLP(props);
// Annotation
var annotation = new Annotation(Text);
pipeline.annotate(annotation);
// Result - Pretty Print
using (var stream = new ByteArrayOutputStream())
{
pipeline.prettyPrint(annotation, new PrintWriter(stream));
Console.WriteLine(stream.toString());
}
this.CustomAnnotationPrint(annotation);
}