Development build. This page describes main, not a released package. The latest published Lodestar.Text is 0.6.0 — read its documentation.

Home › Text › Stemming

FrenchSnowballStemmer

French stemming by the Snowball algorithm.

public static class FrenchSnowballStemmer

Example — a masculine, a feminine and an adverb, all one key.

using Lodestar.Text.Stemming;

string masculine = FrenchSnowballStemmer.Stem("heureux");  // => heureux
string feminine = FrenchSnowballStemmer.Stem("heureuse");  // => heureux
string adverb = FrenchSnowballStemmer.Stem("heureusement");  // => heureux

Remarks — French inflects heavily at the end of the word, which is exactly what a suffix stemmer is good at: national, nationale and nationaux all reduce to national, and the -issait/-issant forms of second-group verbs reduce with their infinitive.

The algorithm uses RV alongside R1 and R2 — a region defined from the start of the word rather than from a suffix, which is what lets it protect short stems that the R1/R2 pair alone would eat. Six steps run in order, and which of them run at all depends on whether the previous one changed the word.

Input is normalised to NFC before the rules see it, so é written as a combining accent behaves like é written as one character.

Reference behaviour is snowballstemmer.stemmer("french") 3.1.1, the Snowball project's own implementation, matched over the 571-word corpus and over all 346,244 words of a French dictionary. nltk's SnowballStemmer("french") implements an older reading and gives another stem for 278 of those words: indicatrice is ind there and indiqu here, bijoux stays bijoux, and l'avion keeps its elision (decision 0006).

Applies to — net10.0, netstandard2.0.

See also — the stemming index, ItalianSnowballStemmer and SpanishSnowballStemmer, which share the Romance region scheme.

Members

Member	What it does
FrenchSnowballStemmer.Stem	The Snowball stem of one French word.