Welcome to BatteryDataExtractor!ΒΆ

BatteryDataExtractor is a battery-aware text-mining software embedded with BERT models for automatically extracting chemical information from scientific literature.

Features:

  • Open-source battery-specific literature-mining toolkit

  • BERT-based token-classification models: abbreviation detection, part-of-speech tagging, chemical-named-entity recognition

  • Double-turn question-answering model for the data extraction of materials and properties

  • State-of-the-art performance on downstream evaluation data sets

  • Updated NLP plugins: new web scrapers, document readers, and tokenizers

  • New options: database auto-saving, original text-saving, and device-selection

Citing:

BatteryDataExtractor: battery-aware text-mining software embedded with BERT models