eu.slipo.athenarc.triplegeo.utils

Class Classification

java.lang.Object

eu.slipo.athenarc.triplegeo.utils.Classification

public class Classification
extends java.lang.Object

Parser for a (possibly multi-tier, i.e., hierarchical) classification scheme (category/subcategory/...). ASSUMPTION: Each category (at any level) must have a unique name, which is being used as its key in the derived dictionary.

Version:

2.0

Author:

Kostas Patroumpas

Field Summary

Fields

Modifier and Type	Field and Description
java.util.Map<java.lang.String,Category>	categories Dictionary for all categories and their parents; their name is used as the key in searching.
java.util.Map<java.lang.String,java.lang.String>	categoryUUIDs Auxiliary dictionary for searching the artificial UUID (value) of a given original category identifier (key).
private boolean	classifyByName
private Configuration	currentConfig
private char	indent
private static int	MAX_LEVELS
(package private) Assistant	myAssistant
(package private) Converter	myConverter
(package private) EmbeddedClassifier	myEmbeddedClassifier
private int	numTiers
private java.lang.String	outputFile
private java.lang.String	splitter

Constructor Summary

Constructors

Constructor and Description
Classification(Configuration config, java.lang.String classFile, java.lang.String outFile) Constructor of the classification hierarchy representation to be used in transformation
Classification(java.lang.String classFile, boolean classifyFlag) Constructor of the classification hierarchy representation used for testing its validity

Method Summary

Modifier and Type	Method and Description
int	countCategories() Returns the number of categories in the dictionary representation of the classification scheme
int	countTiers() Returns the depth, i.e., max number of levels in any path of the given multi-tier classification hierarchy
private void	executeParser4Graph() Parses each item in the classification hierarchy and streamlines the resulting triples according to the given YML mapping.
private void	executeParser4RML() Parses each item in the classification hierarchy and streamlines the resulting triples according to the given RML mapping.
private void	executeParser4Stream() Parses each item in the classification hierarchy and streamlines the resulting triples according to the given YML mapping.
java.lang.String	findUUID(java.lang.String categoryId) For a given category identifier in the classification scheme, find its respective UUID
java.lang.String	getEmbeddedCategory(java.lang.String categoryName) For a given category name, identifies its respective category in the embedded (default) classification scheme
java.lang.String	getUUID(java.lang.String categoryName) For a given category name, identifies its respective UUID in the classification scheme
void	parseCSVFile(java.lang.String classificationFile) Parses input CSV file containing the classification hierarchy
void	parseYMLFile(java.lang.String classificationFile) Parses input YML file containing the classification hierarchy
private void	printDescendants(java.lang.String parent_id, int level) Given a parent identifier, recursively find all its descendants and print them in a YML-like fashion
void	printHierarchyYML() Prints the entire classification scheme representation to the standard output in a YML-like fashion
Category	searchById(java.lang.String categoryId) For a given category identifier, finds the respective entry in the classification scheme
Category	searchByName(java.lang.String categoryName) For a given category name, finds the respective entry in the classification scheme

Methods inherited from class java.lang.Object

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

Field Detail

myConverter

Converter myConverter

myAssistant

Assistant myAssistant

myEmbeddedClassifier

EmbeddedClassifier myEmbeddedClassifier

outputFile

private java.lang.String outputFile

currentConfig

private Configuration currentConfig

classifyByName

private boolean classifyByName

numTiers

private int numTiers

MAX_LEVELS

private static final int MAX_LEVELS

See Also:

Constant Field Values

splitter

private java.lang.String splitter

indent

private char indent

categories

public java.util.Map<java.lang.String,Category> categories

Dictionary for all categories and their parents; their name is used as the key in searching.

categoryUUIDs

public java.util.Map<java.lang.String,java.lang.String> categoryUUIDs

Auxiliary dictionary for searching the artificial UUID (value) of a given original category identifier (key).

Constructor Detail

Classification

public Classification(java.lang.String classFile,
                      boolean classifyFlag)

Constructor of the classification hierarchy representation used for testing its validity

Parameters:

classFile - Input file (CSV or YML) containing the user-specified classification scheme.

classifyFlag - Boolean value: True, if the actual name of the category is used in the hierarchy; False, if identifiers are used instead.

Classification

public Classification(Configuration config,
                      java.lang.String classFile,
                      java.lang.String outFile)

Constructor of the classification hierarchy representation to be used in transformation

Parameters:

config - User-specified configuration for the transformation process.

classFile - Input file (CSV or YML) containing the user-specified classification scheme.

outFile - Output file containing the RDF triples resulted from transformation of the classification scheme.

Method Detail

parseYMLFile

public void parseYMLFile(java.lang.String classificationFile)

Parses input YML file containing the classification hierarchy

Parameters:

classificationFile - Path to YML file specifying the classification scheme. ASSUMPTION: Each line in the YML file corresponds to a category; levels are marked with a number of indentation characters at the beginning of each line; no indentation signifies a top-tier category.

parseCSVFile

public void parseCSVFile(java.lang.String classificationFile)

Parses input CSV file containing the classification hierarchy

Parameters:

classificationFile - Path to CSV file specifying the classification scheme. ASSUMPTION: Each line (record) corresponds to a full path from the top-most to the bottom-most category; at each level, two attributes are given: first a (usually numeric) identifier, then the name of the category. ASSUMPTION: This CSV file must use ',' as delimiter character between attributes values enclosed in double quotes ("..."). EXAMPLE ROW for a 3-tier classification: 1,"Food",103,"Restaurant",103005,"Chinese Restaurant".

searchByName

public Category searchByName(java.lang.String categoryName)

For a given category name, finds the respective entry in the classification scheme

Parameters:

categoryName - The category name to search in the classification hierarchy. CUATION! Category names must be unique amongst all levels and are used as keys.

Returns:

The entry in the classification scheme representation corresponding to the given category

searchById

public Category searchById(java.lang.String categoryId)

For a given category identifier, finds the respective entry in the classification scheme

Parameters:

categoryId - The category identifier to search in the classification hierarchy. Category identifiers must be unique amongst all levels.

Returns:

The entry in the classification scheme representation corresponding to the given category

getUUID

public java.lang.String getUUID(java.lang.String categoryName)

For a given category name, identifies its respective UUID in the classification scheme

Parameters:

categoryName - The category name to search in the classification hierarchy. Category names must be unique amongst all levels.

Returns:

The UUID corresponding to this category in the classification scheme

findUUID

public java.lang.String findUUID(java.lang.String categoryId)

For a given category identifier in the classification scheme, find its respective UUID

Parameters:

categoryId - The original category identifier to search in the classification hierarchy. Category identifiers must be unique amongst all levels.

Returns:

The UUID corresponding to this category identifier in the classification scheme

getEmbeddedCategory

public java.lang.String getEmbeddedCategory(java.lang.String categoryName)

For a given category name, identifies its respective category in the embedded (default) classification scheme

Parameters:

categoryName - The category name to search in the user-defined classification hierarchy. Category names must be unique amongst all levels.

Returns:

The embedded category corresponding to that category in the user-defined classification scheme

countCategories

public int countCategories()

Returns the number of categories in the dictionary representation of the classification scheme

Returns:

Total number of categories at all levels

countTiers

public int countTiers()

Returns the depth, i.e., max number of levels in any path of the given multi-tier classification hierarchy

Returns:

Number of levels for a multi-tier classification hierarchy.

printDescendants

private void printDescendants(java.lang.String parent_id,
                              int level)

Given a parent identifier, recursively find all its descendants and print them in a YML-like fashion

Parameters:

parent_id - Original identifier for the parent of a category

level - The level of the category in the classification scheme (0: top-tier)

printHierarchyYML

public void printHierarchyYML()

Prints the entire classification scheme representation to the standard output in a YML-like fashion

executeParser4RML

private void executeParser4RML()

Parses each item in the classification hierarchy and streamlines the resulting triples according to the given RML mapping. Applicable in RML transformation mode.

executeParser4Graph

private void executeParser4Graph()

Parses each item in the classification hierarchy and streamlines the resulting triples according to the given YML mapping. Applicable in GRAPH transformation mode.

executeParser4Stream

private void executeParser4Stream()

Parses each item in the classification hierarchy and streamlines the resulting triples according to the given YML mapping. Applicable in STREAM transformation mode.