org.terrier.applications

Class TRECIndexing

java.lang.Object

org.terrier.applications.TRECIndexing

public class TRECIndexing
extends Object

This class creates the indices for a test collection.

Properties:

• trec.indexer.class - name of the class to use as the indexer. This only applies to the Index method.
• trec.collection.class - name of the class to use as the Collection.

Author:

Gianni Amati, Vassilis Plachouras, Ben He, Craig Macdonald

Constructor Summary

Constructors

Constructor and Description
TRECIndexing() A default constructor that initialised the data structures to use for indexing.
TRECIndexing(String _path, String _prefix) A constructor that initialised the data structures to use for indexing.
TRECIndexing(String _path, String _prefix, String collectionSpec) A constructor that initialised the data structures to use for indexing.

Method Summary

Methods

Modifier and Type	Method and Description
void	createDirectFile() Builds the direct file and lexicon.
void	createInvertedFile() Building the inverted file.
void	createSinglePass() Builds the inverted file from scratch, single pass method
void	index() Calls the method index(Collection[]) of the class Indexer in order to build the data structures for a set of collections.
static void	main(String[] args) Used for testing purposes.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

TRECIndexing

public TRECIndexing(String _path,
            String _prefix)

A constructor that initialised the data structures to use for indexing.

Parameters:

_path - Absolute path to where the index should be created

_prefix - Prefix of the index files, usually "data"

TRECIndexing

public TRECIndexing(String _path,
            String _prefix,
            String collectionSpec)

A constructor that initialised the data structures to use for indexing.

Parameters:

_path - Absolute path to where the index should be created

_prefix - Prefix of the index files, usually "data"

TRECIndexing

public TRECIndexing()

A default constructor that initialised the data structures to use for indexing.

Method Detail

index

public void index()

Calls the method index(Collection[]) of the class Indexer in order to build the data structures for a set of collections. This particular method of the Indexer uses a set of builders for a subset of the collection and builds separate data structures, which are later merged.

createInvertedFile

public void createInvertedFile()

Building the inverted file.

createDirectFile

public void createDirectFile()

Builds the direct file and lexicon. This method goes through the input files specified in the collections.spec file and processes them in groups of n documents, where n is specified by the property bundle.size. Then, it merges the temporary lexicon files. If it necessary, it calls for the optimisation of the identifiers assigned to terms.

createSinglePass

public void createSinglePass()

Builds the inverted file from scratch, single pass method

main

public static void main(String[] args)

Used for testing purposes.

Parameters:

args - the command line arguments.