Corpus and interactions


OLKi is a federated data sharing and communication platform that implements the ActivityPub protocol.

Contact: Christophe CERISARA



GenI is a surface realiser which generates from a flat semantic representation, the sentence(s) verbalising this semantics. GenI uses Feature-Based Lexicalised Tree Adjoining Grammar equipped with a unification based compositional semantics to compute the sentences verbalising the semantic input.

GenI was developed by Carlos Areces and Eric Kow under the supervision of Claire Gardent. It is now maintained by Eric Kow.

To download and install GenI, please visit this site.

Contact: Claire GARDENT


Jeni is a port of GenI in Java realised by Alexandre Denis and Claire Gardent.

Contact: Claire GARDENT



XMG is a so-called MetaGrammar compiler written in Oz. It compiles a grammar description written in the XMG language into an actual electronic grammar. The compiled grammar can then be used within an NLP application, e.g. a symbolic parser, text generator, or else to generate synthetic data for Machine Learning systems. To learn more about XMG and XMG2 (its extension towards modular grammar development), you can check XMG2’s wiki.

Contact: Yannick PARMENTIER



J-Safran (Java Syntaxico-semantic French Analyser) is a 100%-Java free open-source software for manual, semi-automatic and automatic syntactic dependency parsing in French. It supports other languages as well, but does only include French models.

Demo : JSafran live demo

Contact: Christophe CERISARA

Sentiment analysis


SATI API (Sentiment Analysis from Textual Information) is a multilingual Web API to analyze emotions and sentiments conveyed by text. Given a text, it returns the sentiment (positive, negative, neutral) or the emotion formatted in EmotionML. See the specification and the Web API access at this url.

Contact: Samuel CRUZ-LARA

Speech processing


JTrans is a software that is primarily designed to bring speech alignment (and by extension speech recognition) technology right to the user, and ready to use, in a GUI-rich 100% Java software.

More details on the jtrans page.

How to cite:

  author =   {Cerisara, C. and Mella, O. and Fohr, D.},
  title =    {JTrans, an open-source software for semi-automatic text-to-speech alignment},
  booktitle =    {Proc. of INTERSPEECH},
  year =     {2009},
  address =      {Brighton, UK},
  month = sep,

Contact: Christophe CERISARA