JoBimText: A framework for Distributional Semantics
This tutorial presents JoBimText, an open source framework for Distributional Semantics developed by the LT group at TU Darmstadt. The framework comprises tools to compute distributional thesauri, word sense induction, pattern matching and word sense labeling. The computation scales to (almost) arbitrary large data and mostly relies on Hadoop’s MapReduce.
The tutorial is split in three parts: First we will provide a theoretical background for the methods in the framework. The second part will be hands-on and we will explain how to compute models and how to use them using our API. We will also introduce JoBimViz, which is a visualization platform for our models, and let the audience use our API and access the JoBimText models from Java programs. Here, you can get an insight to different models that we have pre-computed with a description of how to compute such models on your data. The third part will discuss examples where JoBimText has already been successfully applied and an outlook in which NLP applications it might be useful.
For more information visit: https://sites.google.com/site/jobimtexttutorial/
Wissensorganisation ‘15: Lexical Resources for Knowledge Organization
Wissensorganisation is a series of bi-annual conferences/workshops with a long tradition organized by the German chapter of the International Society of Knowledge Organisation (ISKO). A list of previous conferences and workshops can be found on the website of the German ISKO (http://isko-de.org/konferenzen/). WissOrg’15 as a full day workshop with the topic “Lexical Resources for Knowlgede Organizsation”.
For more information visit: http://isko-de.org/konferenzen/
NLP applications: completing the puzzle
The scope of the workshop is bringing together approaches trying to solve several NLP task at the same time and mutually using the information among the specific subtasks to reach a good overall solution. Other interesting research topics targeted by the workshop are the use of external knowledge resources (such as DBpedia, Wikipedia or the Web), in order to extract background and real–world information that could be used to understand texts and solve NLP problems.
For more information visit: http://wordpress.let.vupr.nl/nlpapplications/