Lemmatizer API

Overview

Dictionary-based lemmatizer for Nepali words. Maps inflected forms to their canonical lemma (dictionary form).

Classes

NepaliLemmatizer

from nepalikit.lemmatizer import NepaliLemmatizer

lemmatizer = NepaliLemmatizer()

# Single word
lemma = lemmatizer.lemmatize("गर्छन्")
print(lemma)  # "गर्नु"

# Full text
text = lemmatizer.lemmatize_text("मैले किताहरू पढेको छु")
print(text)  # "म किता पढ्नु छु"

# List of words
words = lemmatizer.lemmatize_words(["मैले", "किताहरू", "पढेको"])
print(words)  # ["म", "किता", "पढ्नु"]

Constructor: - NepaliLemmatizer(dictionary_path=None, min_residue=2) — Optional custom dictionary path and minimum stem length.

Functions

lemmatize(word)

Lemmatize a single word using the default lemmatizer.

from nepalikit.lemmatizer import lemmatize

print(lemmatize("घरले"))     # "घर"
print(lemmatize("गर्छन्"))    # "गर्नु"
print(lemmatize("मैले"))     # "म"

lemmatize_text(text)

Lemmatize all words in a text string.

from nepalikit.lemmatizer import lemmatize_text

print(lemmatize_text("मैले किताहरू पढेको छु"))  # "म किता पढ्नु छु"

Methods

NepaliLemmatizer.lemmatize(word)

Find the lemma for a single word. Checks dictionary first, falls back to rule-based suffix stripping.

Parameters: - word (str): A Nepali word

Returns: str — The lemma of the word

NepaliLemmatizer.lemmatize_text(text)

Lemmatize all words in a text string.

Parameters: - text (str): Nepali text

Returns: str — Text with each word replaced by its lemma

NepaliLemmatizer.lemmatize_words(words)

Lemmatize a list of words.

Parameters: - words (list): List of Nepali word strings

Returns: list — List of lemma strings