The main C-ORAL-ROM objective is to permit the validation of HLT based on spoken language interface with challenging LRs which represent spontaneous speech. A further objective is to stimulate the development of multilingual speech technology in the Romance area.

C-ORAL-ROM develops a highly innovative Multilingual LR simultaneously ensuring:

1. representativeness of spontaneous speech for each language
2. comparability throughout the four Romance languages corpora
3. simultaneous access to acoustic and textual information
4. tagging with respect to prosodic parsing & action values of the all textual information
5. utterance based text / speech alignment

The multimedia output enables the direct use of C-ORAL-ROM, in order to validate both automatic voice recognition (sound-to-text interface) & voice synthesis (text-to-sound interface).