Hizking21

The Hizking21 project, run between 2002 and 2004 by the same consortium as the one forming Ber2Tek, was the first joint experience among the most important agents in the Basque environment in the areas of language technologies for voice and multimedia. Its aims were mainly as follows:

  • To put Basque on the same level as other languages in terms of voice language technologies.
  • To improve the life quality of citizens by facilitating the interaction of different systems, communication and the rendering and receiving of services, as well as by contributing towards ensuring the universalisation of information access by Basque speakers without the need for them to renounce their own language while at the same time encouraging the use of the language among citizens.
  • To set up a "Network of Excellence" in Language Engineering based on a network of research, development and marketing around speech and language technologies that will enable it to successfully compete and collaborate with other European networks.
  • To bring the solutions offered by Language Engineering closer to SMEs by simplifying platforms and cutting costs.
  • To enable the setting up of new technology-based firms (NTBFs) geared towards the development of advanced applications in the area of Language Engineering on the basis of the results of the research.

The following results obtained and the technologies developed in the Hizking21 project are among those worth highlighting.

  • Basic tools for analysing and tagging text in Basque (lemmatizing, syntactic analysis, recognition of entities and multi-word units, etc.)
  • Lexical resources (e-dictionaries, lexical databases, etc.)
  • Corpus creation and tagging tools
  • Science and Technology Corpus
  • Monolingual and bilingual terminology extraction tools
  • WordNet ontology
  • Spanish-Basque machine translation
  • Plugins for office automation, PDA tools, etc.
  • Various voice corpora
  • Phonetic transcriptor
  • Voice processing tools
  • Text-to-voice conversion tools
  • Avatars