Entity Extraction

Library to identify and label cyber-domain entities from unstructured text. This library makes use of [Stanford's CoreNLP] (http://nlp.stanford.edu/software/corenlp.shtml) and [Apache's OpenNLP] (https://opennlp.apache.org) libraries.

Entity Types

• Software
  • Vendor
  • Product
  • Version
• File
  • Name
• Function
  • Name
• Vulnerability
  • Name
  • Description
  • CVE
  • MS

Input

• Trained Apache OpenNLP perceptron model file in binary format that represents a cyber-domain entity model
• Default CoreNLP models for tokenizing, part-of-speech tagging, sentence splitting, and parse-tree building
• Text content of document to be annotated with cyber labels
• Predefined heuristics, including known-entity lists and regular expressions
• Mapping of tokens (i.e. words or punctuation) to a unique label, if found during training

Process

1. Use the CoreNLP library to tokenize, part-of-speech tag, and build the parse trees of the document's text.
2. Check the tokens (i.e. words and punctuation) against lists of known entities such as Google's Freebase data sets. If the token is found, label it appropriately.
3. Attempt to match a token, or set of tokens against regular expressions. If a match is found, then label the token, or set of tokens.
4. Check token against the token-to-unique-label map and label appropriately, if found.
5. If the token is still unlabeled, generate features/context for the token, and evaluate them against the maximum entropy model (MEM) to determine the label with the highest probability.

Features / Context Used

• Token (word or punctuation to be labeled)
• Prefix (first 6 characters of token)
• Suffix (last 6 characters of token)
• Part of speech tag
• Match current token against a set of regular expressions
• Match pervious token against a set of regular expressions

Output

An Annotation object that represents the document as a map, where annotator classnames are keys. The document map includes the following values:

• Text: original raw text
• Sentences: list of sentences
  • Sentence: map representing one sentence
    • Token: word within the sentence
    • POSTag: part-of-speech tag
    • CyberEntity: cyber domain label for the token
  • ParseTree: sentence structure as a tree

Usage

EntityLabeler labeler = new EntityLabeler();
Annotation doc = labeler.getAnnotatedDoc("My Doc", exampleText);

List<CoreMap> sentences = doc.get(SentencesAnnotation.class);
for ( CoreMap sentence : sentences) {
	for ( CoreLabel token : sentence.get(TokensAnnotation.class)) {
		System.out.println(token.get(TextAnnotation.class) + "\t" + token.get(PartOfSpeechAnnotation.class) + "\t" + token.get(CyberAnnotation.class));
	}
	
	System.out.println("Parse Tree:\n" + sentence.get(TreeAnnotation.class));			
}

See CoreNLP's [JavaDocs] (http://nlp.stanford.edu/nlp/javadoc/javanlp/), [Usage section] (http://nlp.stanford.edu/software/corenlp.shtml), or OpenNLP's [JavaDocs] (https://opennlp.apache.org/documentation/1.6.0/apidocs/opennlp-tools/index.html) for more information.

License

This software is freely distributable under the terms of the MIT License.

Copyright (c) UT-Battelle, LLC (the "Original Author")

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS, THE U.S. GOVERNMENT, OR UT-BATTELLE BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.