Termontological platform

This section describes some of the tools that we developed to support the creation and maintenance of multilingual termontological resources for different purposes (see our projects). These tools fulfill a specific role in the overall Termonto Platform.

Basic NLP unit

This unit is composed of a tagger and shallow parser. It currently contains modules for English, Dutch and French. It can be extended to analyse other languages.

Interlexor

Interlexor

This is a tool for term extraction for any domain. It currently works with Dutch, English and French modules but it can be easily extended to process other languages. The tool helps in term extraction, term variation detection, definition and collocation extraction.

There are a series of modules that can be used independently:

  • tagger
  • shallow parser
  • sense disambiguator
  • collocation collector
  • definition finder 
  • term extractor

Below, a basic automatic definition extraction running. The definition rules can be adapted to meet different conditions and in that way, optimize precision and recall measures.

Definition extraction


TermontoGrapher

This tool picks up the information coming from the term extraction and an ontology and lets users edit and model further linguistic and semantic information about any given domain. 

Termontotographer

The image below shows a semantic view of the termontographic data about Brussels' cultural events from the OSCB project. The Termontographer allows users to nagivate through the semantic or linguistic data and edit and enrich it.

A part of the semantic view

TermontoSpider

This is a Wikipedia crawler that analyses the textual and metadata of Wikipedia to provide an optimal selection of Wikipedia pages relevant to a certain domain, together with additional information about term variation and conceptual issues that have been identified thanks to the connectivity, textual and non-textual hints within Wikipedia.

Termontospider

Reference: Domínguez Burgos, A., Kerremans, K., Temmerman, R., 2012. Strategies in automatic traversal of Wikipedia articles for mining multilingual resources, in: Proceedings of  the Workshop on Challenges to Knowledge Representation in Multilingual Contexts (TKE 2012 conference). Presented at the Challenges to Knowledge Representation in Multilingual Contexts, Madrid.