Home

ZemberekDotNet

No Java. No JVM. No sidecar. Pure .NET Standard 2.1 Turkish NLP — morphology, tokenization, NER, classification, normalization, language ID and more.

ZemberekDotNet is a C#/.NET port of Zemberek-NLP that has evolved into an actively maintained library. It is no longer a strict port — correctness fixes and .NET-specific improvements are introduced where needed.

See the repository README for NuGet quick-start, module table, and code samples.

---

Get started

I want to…	Go to
Install a package and write first code	README — Quick Start
Migrate from Java Zemberek	Java to .NET Migration Quickstart
Compare Java and .NET API names	Java vs .NET Side-by-Side
Build and test from source	Developer Guide
Run CLI tools (train, evaluate, preprocess)	Apps CLI Guide

---

Module reference

Module	Wiki page
Morphological analysis and disambiguation	Morphology Notes
ZemberekDotNet-specific API additions and release notes	Additions and Release Notes
Morpheme inventory	Morphemes Reference
Text classification (fastText)	Classification Training Guide
Noisy text normalization / spell check	Normalization Guide
Proper nouns and named entity recognition	Proper Nouns and Named Entities
Dictionary and lexicon rules	Text Dictionary Rules

---

Reference

• FAQ — common questions, licensing, performance, .NET Framework limits
• Java Wiki Porting Checklist — tracks parity between original Java wiki and this repo

---

Scope notes

• gRPC module documentation is deferred — the gRPC project itself is not yet in scope.
• Apps CLI covers classification, morphology, corpus preprocessing, and NER. Additional commands may be added over time.