Manual para compilar y correr el sistema de resumen automático de Sandeep Sripada (Stanford)

 $ git clone https://github.com/sandeepsripada/Summarization

 $ cd Summarization
 $ javac -classpath src -d bin src/com/constants/*.java src/com/decoding/stackdecoder/*.java src/com/interfaces/*.java src/com/main/*.java src/com/model/DataSet.java src/com/processor/SentenceProcessor.java src/com/score/importance/*.java src/com/score/similarity/*.java src/com/util/*.java

 $ java -Xmx100m -classpath bin com.main.Start data/

La clase SentenceProcessor llama a LSP para tokenizar… actualmente el llamado no ocurre, pero si se da el caso habrá que buscar LSP e instalarlo (o sustituirlo por el tokenizer de NLTK)

4
 $ find . -name "*.java" -exec grep -H LSP {} \;
 ./src/com/constants/Constants.java:	public static final boolean runLSP = false;
 ./src/com/processor/SentenceProcessor.java:	private static File getLSPPath(String filename)
 ./src/com/processor/SentenceProcessor.java:			String command = "perl external/runLSP.pl " + filename + " " + tempDir;
 ./src/com/processor/SentenceProcessor.java:		if(Constants.runLSP)
 ./src/com/processor/SentenceProcessor.java:			path = getLSPPath(filename);
 ./src/com/processor/SentenceProcessor.java:		if(Constants.runLSP)
 ./src/com/processor/SentenceProcessor.java:			path = getLSPPath(filename);