Neural Pipeline

In this section, we introduce in more detail the options of Stanza’s neural pipeline, each processor in it, as well as the data objects that it produces. We also introduce how you can convert data between common formats used in NLP and Stanza data objects, as well as how to build your own processors to use in Stanza’s pipeline.

Table of contents

• Pipeline and Processors
• Data Objects and Annotations
• Data Conversion
• Tokenization & Sentence Segmentation
• Multi-Word Token (MWT) Expansion
• Part-of-Speech & Morphological Features
• Lemmatization
• Dependency Parsing
• Constituency Parser
• Named Entity Recognition
• Sentiment Analysis
• Language Identification
• Coreference