BLLIP Parser – statistical constituency parser

BLLIP Parser is a statistical constituency parser, also known as the Charniak-Johnson reranking parser. It analyses sentences by predicting phrase-structure trees, providing a way to study grammatical constituents and their relationships within a sentence.

The system uses two main stages. A generative parser proposes candidate parse trees, after which a discriminative reranker evaluates the candidates to choose a final analysis. This architecture was influential in statistical parsing research before neural parsers became common.

BLLIP Parser has been used in evaluation work involving newswire, web text and biomedical publications. Different trained models suit different writing styles, making the choice of model important when reproducing historical experiments. It remains valuable for research replication and existing integrations but should not be portrayed as a recently developed parser.

This is free and open source software.

Key Features

• Generates constituency parse trees for input sentences.
• Uses a two-stage generative parser and reranker.
• Applies a maximum-entropy discriminative reranking model.
• Can generate candidate parse lists for reranking.
• Provides trained parsing models for different text domains.
• Includes models derived from Wall Street Journal-style treebank data.
• Offers models tailored to web and biomedical text.
• Exposes Python interfaces and compatibility with NLTK workflows.
• Supports evaluation and reproduction of established parsing experiments.
• Includes scripts for parser and reranker experiments.
• Allows first-stage parse candidates to be assessed by a separate reranker.
• Can evaluate a parser and a parser-reranker combination independently.
• Provides separate models for formal newswire and more informal writing.
• Exposes output suitable for analysing syntactic constituents.
• Includes evaluation-oriented tools for comparing parsing configurations.

Website: github.com/BLLIP/bllip-parser
Support:
Developer: Eugene Charniak, Mark Johnson and contributors
License: Apache License 2.0

Related Software

Read our verdict in the software roundup.

LinuxLinks helps you discover the best free and open source software for Linux. Explore our extensive collection of recommended applications, including alternatives to popular proprietary software. We also publish software reviews, Linux tutorials, hardware reviews, free programming books and articles covering the wider open source community. Browse our Linux resources to discover more. Know a useful free and open source Linux application we haven't covered? Suggest a program for inclusion. Has your software been featured on LinuxLinks? Display one of our Featured on LinuxLinks badges on your website or GitHub README.