Package net.thexcoders.aladin_bot_backend.nlp_models.categorizer

Class CategorizerImpl

java.lang.Object

net.thexcoders.aladin_bot_backend.nlp_models.NLPModel

net.thexcoders.aladin_bot_backend.nlp_models.categorizer.CategorizerImpl

All Implemented Interfaces:

Categorizer

public class CategorizerImpl
extends NLPModel
implements Categorizer

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static class	CategorizerImpl.CategoryResult	this inner class is used to store the result of the getCategory(String[], Language, String) method it stores the values to be stored or logged.

Nested classes/interfaces inherited from interface net.thexcoders.aladin_bot_backend.nlp_models.categorizer.Categorizer

Categorizer.Language

Field Summary

Fields inherited from class net.thexcoders.aladin_bot_backend.nlp_models.NLPModel

fileName, PATH

Constructor Summary

Constructors

Modifier	Constructor	Description
protected	CategorizerImpl()	Categorizer constructor that prepares the files needed to train the Model in the languages added to the Language Enumeration Class.

Method Summary

Modifier and Type	Method	Description
CategorizerImpl.CategoryResult	getCategory(String[] tokens, Categorizer.Language lang, String sentence)	this method is used to get the category of the tokenized sentence using the Tokenizer class.
static CategorizerImpl	getInstance()	gets the instance of the class, if the instance is null it calls the constructor to create a new instance
opennlp.tools.doccat.DoccatModel	train(String lang)	This method takes a lot of time to execute since it prepares the training Model for the categorizer based on the training Data provided in the CategorizerImpl() constructor} For optimization purposes, we Have used a singleton

Methods inherited from class net.thexcoders.aladin_bot_backend.nlp_models.NLPModel

fileExists, generateFile

Methods inherited from class java.lang.Object

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

Constructor Detail

CategorizerImpl

protected CategorizerImpl()

Categorizer constructor that prepares the files needed to train the Model in the languages added to the Language Enumeration Class.

This class is private due to the method train(String)

Method Detail

getInstance

public static CategorizerImpl getInstance()

Description copied from interface: Categorizer

gets the instance of the class, if the instance is null it calls the constructor to create a new instance

Returns:

Categorizer the instance of the class.

train

public opennlp.tools.doccat.DoccatModel train(String lang)

Description copied from interface: Categorizer

This method takes a lot of time to execute since it prepares the training Model for the categorizer based on the training Data provided in the CategorizerImpl() constructor} For optimization purposes, we Have used a singleton

Specified by:

train in interface Categorizer

Parameters:

lang - a string value representing the language in which the model will be trained. It also references the file to be used to train the Model

Returns:

instance of DoccatModel which is the trained model.

See Also:

ObjectStream, DoccatModel, TrainingParameters, DoccatFactory, DocumentSampleStream, DocumentSample, DocumentCategorizerME, MarkableFileInputStreamFactory, PlainTextByLineStream

getCategory

public CategorizerImpl.CategoryResult getCategory(String[] tokens,
                                                  Categorizer.Language lang,
                                                  String sentence)

Description copied from interface: Categorizer

this method is used to get the category of the tokenized sentence using the Tokenizer class.

Specified by:

getCategory in interface Categorizer

Parameters:

tokens - an array of tokens retrieved from the tokenizer

lang - the Categorizer.Language in which the tokens are in

sentence - the sentenced (a single sentence from the user's input) from which the tokens are retrieved.

Returns:

an instance of CategorizerImpl.CategoryResult representing the result to be displayed to the user.