Unsupervised Keyword Extraction

YAKE!

Yet Another Keyword Extractor

A light-weight unsupervised automatic keyword extraction method using text statistical features. No training required. Language-independent. Single-document processing.

25+
Languages
500+
Citations
8M+
Downloads
Open
Source

Why Choose YAKE?

Powerful features that make keyword extraction simple and effective

Unsupervised

No training required. Works out-of-the-box without the need for labeled data or external corpora.

Language Independent

Supports 25+ languages without language-specific configurations or dictionaries.

Single Document

Extracts keywords from individual documents. No corpus needed for comparison.

Quick Start

Get started with YAKE in seconds

Installation
Install via pip:
pip install git+https://github.com/INESCTEC/yake