Akar Htut Kaung

Akar Htut Kaung

He/Him/His

Obsessed with getting robots to handle the chaos of the real world: dodging humans, adapting on the fly, and making decisions without hand-holding. I build intelligent systems that learn, plan, and move with purpose.

Core NLP XML to Brat Ann Converter

View on GitHub

Overview

This Python script automates the conversion of CoreNLP XML files into the Brat annotation format, specifically handling Enhanced++ Dependencies. It extracts Part-of-Speech (POS) tags and syntactic dependency relations to generate annotation and visualization configuration files compatible with the Brat tool, streamlining linguistic annotation workflows.

Key Features

  • Parses CoreNLP XML structure using Python’s built-in xml.etree.ElementTree library.
  • Extracts tokens, POS tags, and dependency relations with Enhanced++ Dependencies support.
  • Generates Brat-compatible output files: .ann (annotations), annotation.conf, and visual.conf.
  • Designed for ease of use: run from the command line with a single XML input file.

Usage

Run the script in your terminal by specifying the input XML file:

python3 converter.py [fileToConvert.xml]

This produces three output files in the current directory:

  • fileToConvert.ann — Brat annotation file
  • annotation.conf — Configuration for annotation labels
  • visual.conf — Visualization settings

Important Notes

  • The script relies solely on Python Standard Library modules, requiring no external dependencies.
  • Input XML files must follow CoreNLP’s structure with <document>, <sentences>, <tokens>, and <dependencies> elements for proper parsing.
  • This tool processes one XML file at a time and does not support batch conversion.

Development Details

  • Language: Python
  • Libraries: xml.etree.ElementTree (Standard Library)
  • Focus: Natural Language Processing (NLP) annotation automation