de.psychometrica.lsa

Class VectorStore

java.lang.Object

de.psychometrica.lsa.VectorStore

public class VectorStore
extends java.lang.Object

This class serves for retrieving coordinates from the database file. The class is used as a singleton. Please use the getInstance() to use it's methods. Please set the path to the database file before doing any further actions.

Method Summary

Methods

Modifier and Type	Method and Description
void	disconnect() Disconnects the database.
boolean	exists(java.lang.String word) Checks, if the word is in the database
static VectorStore	getInstance() Initializes or retrieves the VectorStore
java.lang.String[]	getKnownWords(java.lang.String text) Return all known word in the text
java.lang.String[]	getMissingWords(java.lang.String text) Return all unknown word in the text
java.util.ArrayList<java.lang.String>	getMissingWordsList(java.lang.String text)
SemanticVector	getUnweightedSemanticVector(java.lang.String text) The pure SemanticVector for a the text is computed by adding the vectors of all single words.
SemanticVector	getWeightedSemanticVector(java.lang.String text) The SemanticVector for a the text is computed by adding the vectors of all single words.
SemanticVector	getWord(java.lang.String word) Retrieves the vector for a specific word
boolean	isConnected() Checks, if the connection to the database is still open
static void	setPath(java.lang.String directory) Please set up the path to the directory file before doing any computations.

Methods inherited from class java.lang.Object

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

Method Detail

getInstance

public static VectorStore getInstance()

Initializes or retrieves the VectorStore

Returns:

the VectorStore

setPath

public static void setPath(java.lang.String directory)

Please set up the path to the directory file before doing any computations.

Parameters:

directory - the directory, where all other database files reside

getWord

public SemanticVector getWord(java.lang.String word)

Retrieves the vector for a specific word

Parameters:

word - The word

Returns:

the semantic vector

disconnect

public void disconnect()

Disconnects the database.

isConnected

public boolean isConnected()

Checks, if the connection to the database is still open

Returns:

Returns true, if the connection is open, else return false.

getWeightedSemanticVector

public SemanticVector getWeightedSemanticVector(java.lang.String text)

The SemanticVector for a the text is computed by adding the vectors of all single words. Afterwards, the values on the different dimensions are weighted by the singular values. To our experience, this is a favorable approach when scoring essays.

Parameters:

text - The text, for which the vector should be computed

Returns:

the semantic vector

getUnweightedSemanticVector

public SemanticVector getUnweightedSemanticVector(java.lang.String text)

The pure SemanticVector for a the text is computed by adding the vectors of all single words. In contrast to getWeightedSemanticVector() the values on the different dimensions are not weighted by the singular values.

Parameters:

text - The text, for which the vector should be computed

Returns:

the semantic vector

getMissingWords

public java.lang.String[] getMissingWords(java.lang.String text)

Return all unknown word in the text

Parameters:

text - the text to check

Returns:

String-Array with all unknown words

getKnownWords

public java.lang.String[] getKnownWords(java.lang.String text)

Return all known word in the text

Parameters:

text - the text to check

Returns:

String-Array with all known words

exists

public boolean exists(java.lang.String word)

Checks, if the word is in the database

Parameters:

word - the word to check

Returns:

true id the word exists in the database, else return false

getMissingWordsList

public java.util.ArrayList<java.lang.String> getMissingWordsList(java.lang.String text)