org.terrier.applications
Class TrecTerrier

java.lang.Object
  org.terrier.applications.TrecTerrier

public class TrecTerrier
extends java.lang.Object

The text-based application that handles querying with Terrier, for TREC-like test collections. TrecTerrier, indexing TREC collections with Terrier.
usage: java TrecTerrier [flags in any order]

-h --help print this message
-V --version print version information
-i --index index a collection
-r --retrieve retrieve from an indexed collection
-e --evaluate evaluates the results in the directory
var/results with the specified qrels file
in the file etc/trec.qrels

If invoked with '-i', then both the direct and
inverted files are build, unless it is specified which
of the structures to build.
-d --direct creates the direct file
-v --inverted creates the inverted file, from an already existing direct

If invoked with '-r', there are the following options.
-c value parameter value for term frequency normalisation.
If it is not specified, then the default value for each
weighting model is used, eg PL2 => c=1, BM25 b=> 0.75
-q --queryexpand applies query expansion

If invoked with '-e', there is the following option.
-p --perquery reports the average precision for each query separately.
filename.res restrict evaluation to filename.res only.

If invoked with one of the following options, then the contents of the
corresponding data structure are shown in the standard output.
--printdocid prints the contents of the document index
--printlexicon prints the contents of the lexicon
--printinverted prints the contents of the inverted file
--printdirect prints the contents of the direct file
--printstats prints statistics about the indexed collection


Author:

Vassilis Plachouras

Field Summary

protected static int	ARGUMENTS_OK
protected double	c The value of the term frequency normalisation parameter.
protected boolean	direct Specifies whether to build the direct file only.
protected static int	ERROR_CONFLICTING_ARGUMENTS
protected static int	ERROR_DIRECT_FILE_EXISTS
protected static int	ERROR_DIRECT_FILE_NOT_EXISTS
protected static int	ERROR_DIRECT_NOT_INDEXING
protected static int	ERROR_EXPAND_NOT_RETRIEVE
protected static int	ERROR_GIVEN_C_NOT_RETRIEVING
protected static int	ERROR_HADOOP_NOT_RETRIEVAL
protected static int	ERROR_HADOOP_ONLY_INDEX
protected static int	ERROR_INVERTED_NOT_INDEXING
protected static int	ERROR_LANGUAGEMODEL_NOT_RETRIEVE
protected static int	ERROR_NO_ARGUMENTS
protected static int	ERROR_NO_C_VALUE
protected static int	ERROR_PRINT_DIRECT_FILE_NOT_EXISTS
protected static int	ERROR_PRINT_DOCINDEX_FILE_NOT_EXISTS
protected static int	ERROR_PRINT_INVERTED_FILE_NOT_EXISTS
protected static int	ERROR_PRINT_LEXICON_FILE_NOT_EXISTS
protected static int	ERROR_PRINT_STATS_FILE_NOT_EXISTS
protected static int	ERROR_UNKNOWN_OPTION
protected boolean	evaluation Specifies whether to perform trec_eval like evaluation.
protected boolean	evaluation_per_query Specifies whether to perform trec_eval like evaluation, reporting only average precision for each query.
protected java.lang.String	evaluation_type Specifies if the evaluation is done for adhoc or named-page finding retrieval task.
protected java.lang.String	evaluationFilename The file to evaluation, if any
protected boolean	hadoop use Hadoop indexing
protected boolean	indexing Specifies whether to index a collection
protected boolean	inverted Specifies whether to build the inverted file from an already created direct file.
protected boolean	isParameterValueSpecified Indicates whether there is a specified value for the term frequency normalisation parameter.
protected boolean	printdirect Specifies whether to print the direct file
protected boolean	printdocid Specifies whether to print the document index
protected boolean	printHelp Specifies whether a help message is printed
protected boolean	printinverted Specifies whether to print the inverted file
protected boolean	printlexicon Specifies whether to print the lexicon
protected boolean	printmeta whether to print the meta index
protected boolean	printstats Specifies whether to print the statistics file
protected boolean	printVersion Specified whether a version message is printed
protected boolean	queryexpand Specifies whether to apply query expansion
protected boolean	retrieving Specifies whether to retrieve from an indexed collection
protected boolean	singlePass Specifies whether to build the inverted file from scrach, sigle pass method
protected java.lang.String	unknownOption The unkown option

Constructor Summary

TrecTerrier()

Method Summary

void	applyOptions(int status) Apply the option resulted from processing the command line arguments
static void	main(java.lang.String[] args) The main method that starts the application
protected int	processOptions(java.lang.String[] args) Processes the command line arguments and sets the corresponding properties accordingly.
void	run() Calls the required classes from Terrier.
protected void	usage() Prints a help message that explains the possible options.
protected void	version() Prints the version information about Terrier

Methods inherited from class java.lang.Object

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

Field Detail

unknownOption

protected java.lang.String unknownOption

The unkown option

evaluationFilename

protected java.lang.String evaluationFilename

The file to evaluation, if any

queryexpand

protected boolean queryexpand

Specifies whether to apply query expansion

printHelp

protected boolean printHelp

Specifies whether a help message is printed

printVersion

protected boolean printVersion

Specified whether a version message is printed

indexing

protected boolean indexing

Specifies whether to index a collection

singlePass

protected boolean singlePass

Specifies whether to build the inverted file from scrach, sigle pass method

hadoop

protected boolean hadoop

use Hadoop indexing

retrieving

protected boolean retrieving

Specifies whether to retrieve from an indexed collection

printdocid

protected boolean printdocid

Specifies whether to print the document index

printlexicon

protected boolean printlexicon

Specifies whether to print the lexicon

printinverted

protected boolean printinverted

Specifies whether to print the inverted file

printdirect

protected boolean printdirect

Specifies whether to print the direct file

printstats

protected boolean printstats

Specifies whether to print the statistics file

printmeta

protected boolean printmeta

whether to print the meta index

evaluation_per_query

protected boolean evaluation_per_query

Specifies whether to perform trec_eval like evaluation, reporting only average precision for each query.

evaluation_type

protected java.lang.String evaluation_type

Specifies if the evaluation is done for adhoc or named-page finding retrieval task. adhoc by default.

inverted

protected boolean inverted

Specifies whether to build the inverted file from an already created direct file.

direct

protected boolean direct

Specifies whether to build the direct file only.

c

protected double c

The value of the term frequency normalisation parameter.

evaluation

protected boolean evaluation

Specifies whether to perform trec_eval like evaluation.

isParameterValueSpecified

protected boolean isParameterValueSpecified

Indicates whether there is a specified value for the term frequency normalisation parameter.

ARGUMENTS_OK

protected static final int ARGUMENTS_OK

See Also:

Constant Field Values

ERROR_NO_ARGUMENTS

protected static final int ERROR_NO_ARGUMENTS

See Also:

Constant Field Values

ERROR_NO_C_VALUE

protected static final int ERROR_NO_C_VALUE

See Also:

Constant Field Values

ERROR_CONFLICTING_ARGUMENTS

protected static final int ERROR_CONFLICTING_ARGUMENTS

See Also:

Constant Field Values

ERROR_DIRECT_FILE_EXISTS

protected static final int ERROR_DIRECT_FILE_EXISTS

See Also:

Constant Field Values

ERROR_DIRECT_FILE_NOT_EXISTS

protected static final int ERROR_DIRECT_FILE_NOT_EXISTS

See Also:

Constant Field Values

ERROR_PRINT_DOCINDEX_FILE_NOT_EXISTS

protected static final int ERROR_PRINT_DOCINDEX_FILE_NOT_EXISTS

See Also:

Constant Field Values

ERROR_PRINT_LEXICON_FILE_NOT_EXISTS

protected static final int ERROR_PRINT_LEXICON_FILE_NOT_EXISTS

See Also:

Constant Field Values

ERROR_PRINT_INVERTED_FILE_NOT_EXISTS

protected static final int ERROR_PRINT_INVERTED_FILE_NOT_EXISTS

See Also:

Constant Field Values

ERROR_PRINT_STATS_FILE_NOT_EXISTS

protected static final int ERROR_PRINT_STATS_FILE_NOT_EXISTS

See Also:

Constant Field Values

ERROR_PRINT_DIRECT_FILE_NOT_EXISTS

protected static final int ERROR_PRINT_DIRECT_FILE_NOT_EXISTS

See Also:

Constant Field Values

ERROR_UNKNOWN_OPTION

protected static final int ERROR_UNKNOWN_OPTION

See Also:

Constant Field Values

ERROR_DIRECT_NOT_INDEXING

protected static final int ERROR_DIRECT_NOT_INDEXING

See Also:

Constant Field Values

ERROR_INVERTED_NOT_INDEXING

protected static final int ERROR_INVERTED_NOT_INDEXING

See Also:

Constant Field Values

ERROR_EXPAND_NOT_RETRIEVE

protected static final int ERROR_EXPAND_NOT_RETRIEVE

See Also:

Constant Field Values

ERROR_GIVEN_C_NOT_RETRIEVING

protected static final int ERROR_GIVEN_C_NOT_RETRIEVING

See Also:

Constant Field Values

ERROR_LANGUAGEMODEL_NOT_RETRIEVE

protected static final int ERROR_LANGUAGEMODEL_NOT_RETRIEVE

See Also:

Constant Field Values

ERROR_HADOOP_NOT_RETRIEVAL

protected static final int ERROR_HADOOP_NOT_RETRIEVAL

See Also:

Constant Field Values

ERROR_HADOOP_ONLY_INDEX

protected static final int ERROR_HADOOP_ONLY_INDEX

See Also:

Constant Field Values

Constructor Detail

TrecTerrier

public TrecTerrier()

Method Detail

version

protected void version()

Prints the version information about Terrier

usage

protected void usage()

Prints a help message that explains the possible options.

main

public static void main(java.lang.String[] args)

The main method that starts the application

Parameters:

args - the command line arguments

processOptions

protected int processOptions(java.lang.String[] args)

Processes the command line arguments and sets the corresponding properties accordingly.

Parameters:

args - the command line arguments.

Returns:

int zero if the command line arguments are processed successfully, otherwise it returns an error code.

run

public void run()
         throws java.lang.Exception

Calls the required classes from Terrier.

Throws:

java.lang.Exception

applyOptions

public void applyOptions(int status)
                  throws java.lang.Exception

Apply the option resulted from processing the command line arguments

Parameters:

status - the status after process the command line arguments.

Throws:

java.lang.Exception