Speech, Language and Communication Technology
Ailbhe Ní Chasaide, Christer Gobl, Elaine Uí Dhonnchadha, Martine Smith, Lorraine Leeson, Breffni O’ Rourke
Speech and Language Technology for Irish
Over the last years, we have developed the first Irish speech synthesis system, along with the prerequisite linguistic resources: both the synthesis and resources are available at www.abair.ie. A modular design allow for extending the current provision (Donegal and Connemara) to all the dialects of Irish. Further research is geared towards personalisation of the synthetic voices, to ensure that the user can choose not only the dialect, but the quality of the voice. An interface to facilitate screenreading for the visually impaired has also been developed. The provision of personalised, speech-based communication aids for those with severe speech impairments is a future objective, as is the development of language pedagogy applications.
We have also developing a suite of Natural Language Processing Tool for Irish, which include a tokenizer, morphological analyser, partial dependency parser and chunker for the automatic processing of Irish in natural language applications. The tokenizer, morphological analyser have been used for part-of-speech tagging of the National Corpus of Irish and in the Irish Speech Synhesisor. Work is ongoing on the dependency parser and chunker.
See also the Irish Speech and Language Technology Research Centre page ITUT
College’s Strategic plan: The technology applications tie in with College’s Strategic emphasis on innovation and entrepreneurship. This research will also involve intensive collaboration within EU