stanfordnlp/CoreNLP

What it solves

Stanford CoreNLP is a set of natural language analysis tools that transforms raw human language text into structured data. It provides the foundational building blocks for higher-level text understanding applications by automating the extraction of linguistic features from text.

How it works

It is an integrated framework written in Java that uses a combination of rule-based systems, probabilistic machine learning, and deep learning components. The system can be configured to run a sequence of language analysis tools (annotators) on a piece of text to extract various linguistic properties.

Who it’s for

It is widely used by researchers in academia, industry, and government for natural language processing tasks across multiple languages, including English, Arabic, Chinese, French, German, Hungarian, Italian, and Spanish.

Highlights

• Comprehensive Analysis: Extracts base forms of words, parts of speech, named entities (companies, people, etc.), and syntactic structures (phrases or dependencies).
• Normalization: Interprets and normalizes dates, times, and numeric quantities.
• Coreference Resolution: Identifies which noun phrases refer to the same entities.
• Multilingual Support: Provides varying levels of support for multiple major world languages.
• Easy Integration: Allows users to run a full suite of analysis tools with just two lines of code.