Development build. This page describes main, not a released package. The latest published Lodestar.Text is 0.7.0 — read its documentation.

Home › Text › Stemming

PorterStemmer

English stemming by Martin Porter's original 1980 algorithm.

public static class PorterStemmer

Example — the textbook plural.

using Lodestar.Text.Stemming;

string stem = PorterStemmer.Stem("caresses");  // => caress

Remarks — reach for this only to match an index that already exists. Porter himself replaced it with Porter2, published as Snowball and available here as EnglishSnowballStemmer, which is his own correction of it. The index page lists all six words the two disagree about.

The algorithm is five steps of suffix rules, gated on a measure of vowel-consonant alternation rather than on named regions — the mechanism Porter2 replaced with R1 and R2. It knows only ASCII letters, so an accented word is not something it was built to receive.

Two forms of the algorithm are published, and PorterStemmerMode picks one: the 1980 paper as printed, the default, or the departures Martin Porter's own reference implementation makes from it. Reference behaviour is nltk.stem.porter.PorterStemmer in the matching mode — ORIGINAL_ALGORITHM or MARTIN_EXTENSIONS — matched in both over every entry of a 73,456-word English dictionary.

Applies to — net10.0, netstandard2.0.

See also — PorterStemmerMode, EnglishSnowballStemmer, the stemming index.

Members

Member	What it does
PorterStemmer.Stem	The Porter stem of one English word, by the paper or by Martin's extensions.