Skip to main content
European Commission logo print header

COMPANIONS: persistent multi-modal interfaces to the Internet

Project description

Multimodal Interfaces
The COMPANIONS project has developed virtual Companions for conversation to change the way people think about the relationships of people to computers and the Internet.

The COMPANIONS vision is that of a personalized, conversational interface to the Internet, one that knows its owner, on a range of platforms, indoor and nomadic, based on integrated high-quality research in multimodal human-computer interfaces, intelligent agents, and human language technology. This is a research agenda bound by a metaphor, that of the COMPANION, an artefact for the large groups of EU citizens whom the Internet simply does not serve as well as it does the academics who invented it. Among those groups are those who live alone: up to 40% of the population of advanced societies will live alone within 20 years, only half of them old. They will be in need of interactive company, conversation and diversion, over and above their need to use the Internet to get information or use chatrooms, or, increasingly, to make sense of their own digital repositories (texts, videos, images) as well as information held about themselves. These needs have not been well met by current interfaces based on browsing, profiling and adaptation, and there is good evidence that a more directly personal language-using persistent interface will be more acceptable. Our case in this proposal is that the technical basis for a Companion agent to the Internet is now in view: and what is needed is a fresh research effort to bring stronger methods of machine learning to bear on these three modalities and their integration. Within 4 years a new integration of existing and novel techniques can be achieved that will together provide a striking advance over anything currently demonstrable, and that enough can be shown after an eighteen month effort to attract the attention of major industrial players. The Consortium contains centres with outstanding world expertise in the application of Machine Learning (particularly Bayesian methods) to problems of speech and language and these will be extended to language-vision correlations; it will function in at least one new EU language (Czech).

Call for proposal

See other projects for this call


EU contribution
€ 2 242 233,00
211 Portobello Street
S1 4DP Sheffield
United Kingdom

See on map

Activity type
Higher or Secondary Education Establishments
Total cost
No data

Participants (14)