Hazm - Persian NLP Toolkit

Hazm is a python library to perform natural language processing tasks on Persian text. It offers various features for analyzing, processing, and understanding Persian text. You can use Hazm to normalize text, tokenize sentences and words, lemmatize words, assign part-of-speech tags, identify dependency relations, create word and sentence embeddings, or read popular Persian corpora.

Features

• Normalization: Converts text to a standard form (diacritics removal, ZWNJ correction, etc).
• Tokenization: Splits text into sentences and words.
• Lemmatization: Reduces words to their base forms.
• POS tagging: Assigns a part of speech to each word.
• Dependency parsing: Identifies the syntactic relations between words.
• Embedding: Creates vector representations of words and sentences.
• Hugging Face Integration: Automatically download and cache pretrained models from the Hub.
• Persian corpora reading: Easily read popular Persian corpora with ready-made scripts.

Installation

To install the latest version of Hazm (requires Python 3.12+), run:

pip install hazm

To use the pretrained models from Hugging Face, ensure you have the huggingface-hub package:

pip install huggingface-hub

Pretrained-Models

Hazm supports automatic downloading of pretrained models. You can find all available models (POS Tagger, Chunker, Embeddings, etc.) on our official Hugging Face page:

👉 Roshan Research on Hugging Face

When using Hazm, simply provide the repo_id and model_filename as shown in the examples below, and the library will handle the rest.

Usage

pythonCopy
from hazm import *

# ===============================
# Stemming
# ===============================
stemmer = Stemmer()
stem = stemmer.stem('کتاب‌ها')
print(stem) # کتاب

# ===============================
# Normalizing
# ===============================
normalizer = Normalizer()
normalized_text = normalizer.normalize('من کتاب های زیــــادی دارم .')
print(normalized_text) # من کتاب‌های زیادی دارم.

# ===============================
# Lemmatizing
# ===============================
lemmatizer = Lemmatizer()
lem = lemmatizer.lemmatize('می‌نویسیم')
print(lem) # نوشت#نویس

# ===============================
# Sentence tokenizing
# ===============================
sentence_tokenizer = SentenceTokenizer()
sent_tokens = sentence_tokenizer.tokenize('ما کتاب می‌خوانیم. یادگیری خوب است.')
print(sent_tokens) # ['ما کتاب می\u200cخوانیم.', 'یادگیری خوب است.']

# ===============================
# Word tokenizing
# ===============================
word_tokenizer = WordTokenizer()
word_tokens = word_tokenizer.tokenize('ما کتاب می‌خوانیم')
print(word_tokens) # ['ما', 'کتاب', 'می\u200cخوانیم']

# ===============================
# Part of speech tagging
# ===============================
tagger = POSTagger(repo_id="roshan-research/hazm-postagger", model_filename="pos_tagger.model")
tagged_words = tagger.tag(word_tokens)
print(tagged_words) # [('ما', 'PRON'), ('کتاب', 'NOUN'), ('می\u200cخوانیم', 'VERB')]

# ===============================
# Chunking
# ===============================
chunker = Chunker(repo_id="roshan-research/hazm-chunker", model_filename="chunker.model")
chunked_tree = tree2brackets(chunker.parse(tagged_words))
print(chunked_tree) # [ما NP] [کتاب NP] [می‌خوانیم VP]

# ===============================
# Word embedding
# ===============================
word_embedding = WordEmbedding.load(repo_id='roshan-research/hazm-word-embedding', model_filename='fasttext_skipgram_300.bin', model_type='fasttext')
odd_word = word_embedding.doesnt_match(['کتاب', 'دفتر', 'قلم', 'پنجره'])
print(odd_word) # پنجره

# ===============================
# Sentence embedding
# ===============================
sent_embedding = SentEmbedding.load(repo_id='roshan-research/hazm-sent-embedding', model_filename='sent2vec-naab.model')
sentence_similarity = sent_embedding.similarity('او شیر میخورد','شیر غذا می‌خورد')
print(sentence_similarity) # 0.4643607437610626

# ===============================
# Dependency parsing
# ===============================
parser = DependencyParser(tagger=tagger, lemmatizer=lemmatizer, repo_id="roshan-research/hazm-dependency-parser", model_filename="langModel.mco")
dependency_graph = parser.parse(word_tokens)
print(dependency_graph)
"""
{0:  {'address': 0,
      'ctag': 'TOP',
      'deps': defaultdict(<class 'list'>, {'root': [3]}),
      'feats': None,
      'head': None,
      'lemma': None,
      'rel': None,
      'tag': 'TOP',
      'word': None},
  1: {'address': 1,
      'ctag': 'PRON',
      'deps': defaultdict(<class 'list'>, {}),
      'feats': '_',
      'head': 3,
      'lemma': 'ما',
      'rel': 'SBJ',
      'tag': 'PRON',
      'word': 'ما'},
  2: {'address': 2,
      'ctag': 'NOUN',
      'deps': defaultdict(<class 'list'>, {}),
      'feats': '_',
      'head': 3,
      'lemma': 'کتاب',
      'rel': 'OBJ',
      'tag': 'NOUN',
      'word': 'کتاب'},
  3: {'address': 3,
      'ctag': 'VERB',
      'deps': defaultdict(<class 'list'>, {'SBJ': [1], 'OBJ': [2]}),
      'feats': '_',
      'head': 0,
      'lemma': 'خواند#خوان',
      'rel': 'root',
      'tag': 'VERB',
      'word': 'می\u200cخوانیم'}})

"""

Documentation

Visit https://roshan-ai.ir/hazm to view the full documentation.

Evaluation

Module name
DependencyParser	85.6%
POSTagger	98.8%
Chunker	93.4%
Lemmatizer	89.9%

	Metric	Value
SpacyPOSTagger	Precision	0.99250
	Recall	0.99249
	F1-Score	0.99249
EZ Detection in SpacyPOSTagger	Precision	0.99301
	Recall	0.99297
	F1-Score	0.99298
SpacyChunker	Accuracy	96.53%
	F-Measure	95.00%
	Recall	95.17%
	Precision	94.83%
SpacyDependencyParser	TOK Accuracy	99.06
	UAS	92.30
	LAS	89.15
	SENT Precision	98.84
	SENT Recall	99.38
	SENT F-Measure	99.11