OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Developer Tools & SDKs

Mitie

MITIE is a dlib-based information-extraction library and toolset, written in C++, for named-entity extraction, binary relation detection and training custom extractors and relation detectors.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Mitie — OpenViST project profile in AI Developer Tools & SDKs.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use MITIE when an application needs entity or binary-relation extraction from text, or when a domain-specific corpus calls for custom extractors.

Who it is for

  • Developers integrating information extraction into text-processing applications.
  • Teams training extractors or relation detectors for specialized corpora.
  • Researchers and developers working with named entities or binary relations in text.

Documented uses

  • Identify named entities in text using a pre-trained model.
  • Detect binary relations in text.
  • Train custom named-entity extractors and relation detectors.
  • Build a streaming command-line tool to annotate text line by line.
Before you choose

Requirements & limits

The repository is public and describes MITIE as open source under the Boost Software License 1.0; its README says the license permits commercial use subject to its terms.

  • MITIE is built on dlib, and its core is written in C++.
  • The primary API is C; repository examples and bindings cover C, C++, Java, R and Python 2.7.
  • Pre-trained models provide varying levels of support for English, Spanish and German; Spanish and German models are separate downloads.
  • The provided examples require trained model files.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Build the streaming named-entity tool

Compile MITIE's command-line tool for processing text files line by line.

Before you start

  • A local MITIE repository.
  • CMake and a C++ build environment, or make.

Steps

  1. Download the trained model files with the command below, or download and extract the model archive into the MITIE folder.

    Documented command
    make MITIE-models

  2. Build ner_stream with make in the top-level folder or in tools/ner_stream, or use CMake from that tool's folder.

  3. Pass text to the executable with the English NER model, following the README's cat sample_text.txt | ./ner_stream MITIE-models/english/ner_model.dat example.

Expected outcome

The tool processes each input line independently and outputs marked-up text.

Guide limitations

  • The recipe uses the English named-entity model shown in the README; Spanish and German models have separate downloads.
Documentation 1 ↗
Source-based practical guide

Build a MITIE shared library with CMake

Compile and install the shared library for use by an application.

Before you start

  • A local MITIE repository.
  • CMake and a C++ build environment.

Steps

  1. Change to the mitielib directory.

  2. Create and enter a build directory.

  3. Run cmake ...

  4. Run cmake --build . --config Release --target install.

Expected outcome

The README says this creates a MITIE shared library in the `mitielib` folder.

Guide limitations

  • The README's CMake instructions note that locating BLAS libraries outside standard locations may require setting CMAKE_LIBRARY_PATH.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
17 October 2026
Original website title
mit-nlp/MITIE · Official homepage title · retained in its source language
How we review sourcesSuggest a correction