matra 0.3.0

NLP library: UDPipe structured parse, base text metrics, summarization, keyphrase extraction. Rust core with Python bindings.
Documentation
cff-version: 1.2.0
title: matra
message: >-
  If you use matra in work you publish, cite the software with the entry
  below, and separately cite the publication behind each measure you
  report. matra implements published methods with documented departures,
  so the software citation identifies the implementation and the method
  citation identifies the method. Both are needed. The per-measure
  references are listed in the methodology page of the documentation.
type: software
authors:
  - name: mox labs
    email: yza.v@moxlabs.org
    website: 'https://github.com/mox-labs'
abstract: >-
  An NLP library that parses text into a typed structure tree, measures
  it, and ranks what is in it. It provides a UDPipe structured parse in
  full CoNLL-U, base text metrics covering readability, lexical density,
  compression, vocabulary type-token ratio, nominalization and passive
  ratio, extractive summarization by TF-IDF and TextRank, and keyphrase
  extraction by RAKE and YAKE. The core is Rust, with Python bindings and
  a command line interface. It reports structure and numbers and makes no
  judgment about the text.
version: 0.3.0
date-released: '2026-11-06'
license: MIT
repository-code: 'https://github.com/mox-labs/matra'
url: 'https://mox-labs.github.io/matra/'
keywords:
  - nlp
  - udpipe
  - dependency parsing
  - summarization
  - keyphrase extraction
  - readability