Lexeme

Namespace: Nlp4Net.NlpLib Assembly: NlpLib.dll

public class Lexeme : IUserData, ICloneable

Lexeme is a string of characters. There are three types Lexeme.LexType of Lexemes.

Lexemes with syntax and semantic information contain Words.

Note: lexical ambiguity

There may be several syntactically different Words associated with the same Lexeme. For example the same lexeme "code" has two Words: noun and verb; it plays different syntax roles and carries different semantic in the following utterances: "We code the project. The code is complex." Which Word is used can be determined only during higher levels of processing.

Words may belong to different languages. Currently NlpLib supports only en-US language.

You can use lexical ambiguity in OCR or speech recognition when Lexeme is not clearly recognized. Instead of processing different lexemes, overload the same Lexeme with possible Words and let syntax parser to make a choice.

Constructors	Description
Lexeme()	creates an empty Lexeme
Lexeme(Lexeme other)	copy constructor
Lexeme(string text)	initializes Lexeme with a string

Methods	Description
object ICloneable.Clone()	creates a copy of Lexeme

Properties	Description
bool HasWords	true if Lexeme has at least one Word
int IndexInUtterance	index in Utterance.Lexemes array
Lexeme.LexType LexemeType	Lexeme can be one of three types. See Lexeme.LexType
long Position	Position in chars in a text stream since last NLParser.Flush()
string Text	string of characters comprising the Lexeme
object IUserData.UserData	Your arbitrary object
Word[] Words	An array of Words associated with the Lexeme