Persian-Keyword-Extraction

A hybrid syntactic-statistical method for extracting keywords from Persian texts.

Methodology

1. Preprocessing: Normalize text (e.g., unify spellings).
2. Processing:
   • POS tagging using Hazm.
   • Generate noun/adjective combinations.
   • Stemming and frequency counting.
3. Selection: Pick top 5 combinations by frequency.

Usage

1. Install dependencies:

   pip install -r requirements.txt
   Run the pipeline

from preprocessing import normalize_persian_text from processing import generate_combinations from selection import select_keywords

text = open("data/sample_input.txt", encoding="utf-8").read() normalized = normalize_persian_text(text) combinations = generate_combinations(normalized) keywords = select_keywords(combinations) print(keywords)