DESCRIPTION OF TASKS
There are two different groups of tasks this year: two challenge tasks and two vintage or classic tasks. Both groups of tasks are in the domain of travel and tourism.The Challenge tasks involve two pairs of languages: Chinese-English, and Italian-English. These two tasks resemble the challenge from the 2006 edition of the IWSLT workshop which involved translating spontaneous speech, with a slight difference in the data available for the Italian-English pair.
The second group of tasks, named "Classic Vintage", resembles the tasks offered in the 2005 edition of the IWSLT workshop which required translation of read speech. These tasks involve the language pairs of Japanese-English, and Arabic-English.
Challenge Task Details:
Two Classic Vintage BTEC** Tasks:
For details about the data for each of these tasks, please see the data specifications.
* The SITAL corpus contains dialogs between a travel agency's operator and a person seeking travel information or to book a ticket, a hotel room or a flight.
** The BTEC corpus contains tourism-related sentences similar to those that are usually found in phrasebooks for tourist going abroad.