/// <summary>
/// Summarizes the specified input using the specified <paramref name="sentenceDetector"/> and <paramref name="tokenizer"/>.
/// </summary>
/// <param name="input">The input string to be summarized.</param>
/// <param name="sentenceDetector">The sentence detector.</param>
/// <param name="tokenizer">The tokenizer.</param>
/// <returns>The summarized string.</returns>
/// <exception cref="System.ArgumentNullException">
/// <paramref name="sentenceDetector"/>
/// or
/// <paramref name="tokenizer"/>
/// </exception>
public string Summarize(string input, ISentenceDetector sentenceDetector, ITokenizer tokenizer)
{
if (string.IsNullOrEmpty(input))
{
return(string.Empty);
}
if (sentenceDetector == null)
{
throw new ArgumentNullException(nameof(sentenceDetector));
}
if (tokenizer == null)
{
throw new ArgumentNullException(nameof(tokenizer));
}
var doc = new Document("x-unspecified", input);
var anl = new AggregateAnalyzer {
new SentenceDetectorAnalyzer(sentenceDetector),
new TokenizerAnalyzer(tokenizer)
};
anl.Analyze(doc);
return(ProcessSummarization(doc));
}
public static NaturalLanguageData AnalyzeMessage(Chat chat)
{
var document = new Document("en", chat.message);
var sentences = new List <Sentence>();
try
{
analyzer.Analyze(document);
foreach (var s in document.Sentences)
{
var sentence = new Sentence();
sentence.tokens = new List <Token>();
foreach (var t in s.Tokens)
{
var token = new Token();
token.POSTag = t.POSTag;
token.Lexeme = t.Lexeme;
token.Stem = wordStemmer.GetSteamWord(t.Lexeme);
sentence.tokens.Add(token);
}
sentence.chunks = new List <Chunk>();
foreach (var c in s.Chunks)
{
var chunk = new Chunk();
chunk.tag = c.Tag;
chunk.tokens = new List <Token>();
foreach (var t in c.Tokens)
{
var token = new Token();
token.POSTag = t.POSTag;
token.Lexeme = t.Lexeme;
token.Stem = wordStemmer.GetSteamWord(t.Lexeme);
chunk.tokens.Add(token);
}
sentence.chunks.Add(chunk);
}
sentence.interrogative = isInterrogative(s);
sentence.triplets = tripletService.GetSentenceTriplets(sentence);
sentences.Add(sentence);
}
}
catch (AnalyzerException)
{
}
var naturalLanguageData = new NaturalLanguageData();
naturalLanguageData.sentences = sentences;
return(naturalLanguageData);
}
public void TestEverything()
{
var doc = new Document("en",
"Bart, with $10,000, we'd be millionaires! We could buy all kinds of useful things like... love!");
analyzer.Analyze(doc);
Assert.NotNull(doc);
Assert.AreEqual(2, doc.Sentences.Count);
Assert.AreEqual(true, doc.IsTokenized);
Assert.AreEqual(true, doc.IsTagged);
Assert.AreEqual(true, doc.IsChunked);
Assert.AreEqual(true, doc.IsParsed);
Assert.AreEqual(1, doc.Sentences[0].Entities.Count);
}
/// <summary>
/// Summarizes the specified input using the specified <paramref name="sentenceDetector"/> and <paramref name="tokenizer"/>.
/// </summary>
/// <param name="input">The input string to be summarized.</param>
/// <param name="sentenceDetector">The sentence detector.</param>
/// <param name="tokenizer">The tokenizer.</param>
/// <returns>The summarized string.</returns>
/// <exception cref="System.ArgumentNullException">
/// <paramref name="sentenceDetector"/>
/// or
/// <paramref name="tokenizer"/>
/// </exception>
public string Summarize(string input, ISentenceDetector sentenceDetector, ITokenizer tokenizer) {
if (string.IsNullOrEmpty(input))
return string.Empty;
if (sentenceDetector == null)
throw new ArgumentNullException("sentenceDetector");
if (tokenizer == null)
throw new ArgumentNullException("tokenizer");
var doc = new Document("x-unspecified", input);
var anl = new AggregateAnalyzer {
new SentenceDetectorAnalyzer(sentenceDetector),
new TokenizerAnalyzer(tokenizer)
};
anl.Analyze(doc);
return ProcessSummarization(doc);
}
public IList <string> ScrapeMovieTitles(string actorWikipediaUrl, string[] ignoreWords)
{
var result = new List <string>();
var web = new HtmlWeb();
var page = web.Load(actorWikipediaUrl);
//read the content of "Career" section of the article
var careerNode = page.DocumentNode.SelectSingleNode("//*/h2[starts-with(.,'Career')]");
if (careerNode != null)
{
string text = "";
var node = careerNode.NextSibling;
while (node != null && node.Name != "h2")
{
text += node.InnerText;
node = node.NextSibling;
}
var doc = new SharpNL.Document("en", text);
_analyzer.Analyze(doc);
foreach (var sentence in doc.Sentences)
{
Token prevToken = null;
string title = "";
bool withinTitle = false;
foreach (var token in sentence.Tokens)
{
if (!withinTitle && IsTitleStart(token, prevToken))
{
withinTitle = true;
title = "";
}
else if (withinTitle)
{
if (IsTitleEnd(token, prevToken))
{
if (prevToken != null && prevToken.POSTag.IsNamePosTag())
{
title = title.AddWord(prevToken.Lexeme);
}
if (IsMovieTitle(title))
{
result.Add(title);
}
title = "";
withinTitle = false;
}
else
{
title = title.AddWord(prevToken.Lexeme);
}
}
prevToken = token;
}
}
}
return(result);
}
