KWJA: Kyoto-Waseda Japanese Analyzer †

KWJA is a Japanese language analyzer based on pre-trained language models. KWJA performs many language analysis tasks, including:

• Typo correction
• Tokenization
• Word normalization
• Morphological analysis
• Named entity recognition
• Word feature tagging
• Dependency parsing
• PAS analysis
• Bridging reference resolution
• Coreference resolution
• Discourse relation analysis

For more details, see the GitHub repository (https://github.com/ku-nlp/kwja).

Usage †

Analyze a text

 $ kwja --text "KWJAは日本語の統合解析ツールです。様々な言語解析を統一的な方法で解いています。"

Analyze a text file

 $ echo "KWJAは日本語の統合解析ツールです。様々な言語解析を統一的な方法で解いています。" > path/to/file.txt
 $ kwja --file path/to/file.txt

Requirements †

• Python 3.8 or later

Installation †

 $ pip install kwja

Change Log †

• 1.1.0 - 2022/10/11 Add an interactive mode. Support Python 3.8.
• 1.0.0 - 2022/9/28　Release