pitt.search.semanticvectors
Class BuildPositionalIndex
java.lang.Object
pitt.search.semanticvectors.BuildPositionalIndex
public class BuildPositionalIndex
- extends java.lang.Object
Command line utility for creating semantic vector indexes using the
sliding context window approach (see work on HAL, and by Shutze).
Method Summary |
static void |
main(java.lang.String[] args)
Builds term vector stores from a Lucene index - this index must
contain TermPositionVectors. |
static void |
usage()
Prints the following usage message:
BuildPositionalIndex class in package pitt.search.semanticvectors
Usage: java pitt.search.semanticvectors.BuildPositionalIndex PATH_TO_LUCENE_INDEX
BuildPositionalIndex creates file termtermvectors.bin in local directory. |
Methods inherited from class java.lang.Object |
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait |
BuildPositionalIndex
public BuildPositionalIndex()
usage
public static void usage()
- Prints the following usage message:
BuildPositionalIndex class in package pitt.search.semanticvectors
Usage: java pitt.search.semanticvectors.BuildPositionalIndex PATH_TO_LUCENE_INDEX
BuildPositionalIndex creates file termtermvectors.bin in local directory.
Other parameters that can be changed include windowlength (size of context window),
vector length (number of dimensions), seed length (number of non-zero
entries in basic vectors), and minimum term frequency.
To change these use the following command line arguments:
-dimensions [number of dimensions]
-seedlength [seed length]
-mintermfreq [minimum term frequency]
-windowradius [window half size]
-positionalmethod [type of index: basic (default), directional (HAL), permutation (Sahlgren 2008)
main
public static void main(java.lang.String[] args)
throws java.lang.IllegalArgumentException
- Builds term vector stores from a Lucene index - this index must
contain TermPositionVectors.
- Parameters:
args
-
- Throws:
java.lang.IllegalArgumentException