org.activemath.content
Class NLG

java.lang.Object
  org.activemath.content.NLG

public class NLG
extends Object

The class contains methods used for Natural Language Generation, e.g., title generation, bridge texts, etc.

Field Summary

protected static Logger	log
static String	regex_TITLE We define here the regular expression for a title, the case is assumed to be lower case.

Constructor Summary

NLG()

Method Summary

static String	generateBridgeText(Item before, Item behind) This method generate the bridge text between two Content Items of a Book.
static Map	generateContentMap(String phrase) Returns text.NLGGenerator.
static Map	generateMapForPhrase(String phrase) Returns phrase in all available languages.
static String	generateText(String lang, String phrase) Returns the text of the phrase in the given language.
static Map<String,String>	generateTextMap(String userId, String queryname, List<String> contentIds) generate text for given user and queryname and contentIds.
static Map	generateTitleMap(String phrase) Returns text.NLGGenerator.
static Map	generateTitleMapForContentId(String contentId) Currently we check the type of the contentId.

Methods inherited from class java.lang.Object

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

Field Detail

log

protected static Logger log

regex_TITLE

public static final String regex_TITLE

We define here the regular expression for a title, the case is assumed to be lower case. This String can be always expanded to more cases for more languages.

See Also:

Constant Field Values

Constructor Detail

NLG

public NLG()

Method Detail

generateTitleMapForContentId

public static Map generateTitleMapForContentId(String contentId)

Currently we check the type of the contentId. If it starts with Mbase signature, then we call the mbase to get the title Map. Otherwise we use Mediator to get the title by calling the propriate repository. More generally and logically, we should just call the mediator to find the propriate repository to get the title. But there's some factors we considered:

• Calling Mbase directly might be faster as we call mediator and let him analyse and then do the same thing.
• Calling Mediator self can be remote, and this takes also time.

Parameters:

contentId -

Returns:

A Map of Titles

generateBridgeText

public static String generateBridgeText(Item before,
                                        Item behind)

This method generate the bridge text between two Content Items of a Book. It is not yet implemented

Parameters:

before -

bebind -

Returns:

BridgeText

generateText

public static String generateText(String lang,
                                  String phrase)

Returns the text of the phrase in the given language.

generateTextMap

public static Map<String,String> generateTextMap(String userId,
                                                 String queryname,
                                                 List<String> contentIds)

generate text for given user and queryname and contentIds.

Parameters:

userId -

queryname -

contentIds -

Returns:

generateTitleMap

public static Map generateTitleMap(String phrase)

Returns text.NLGGenerator.phrase.Title in all available languages.

generateContentMap

public static Map generateContentMap(String phrase)

Returns text.NLGGenerator.phrase.Content in all available languages.

generateMapForPhrase

public static Map generateMapForPhrase(String phrase)

Returns phrase in all available languages.