Objective
Nowadays, voice portals provide users with a broad variety of information (timetables, traffic circulation, weather forecast, cultural events, etc.), and are experiencing an exponential increment of popularity. Most of the times, the common factor of the sought after information is its dynamism: the world rapidly changes over the time, users ask (and pay) for completely fresh information. Gathering and maintaining all the relevant information, and providing users with complete and updated contents are undoubtedly a bottleneck and an expensive work for voice portals and web-based information services. Doing it better can mean the difference between their success and failure.
Open Domain Question Answering (QA), the core technology behind the QALL-ME project, provides at the same time both a research and technological framework to address the above-mentioned issues and a challenging applicative perspective for real use scenarios. In contrast to the technologies behind today's web search engines, the goal of QA is not to return a list of documents, but the actual sequence of data (i.e. words) which constitutes the answer. The scientific and technological objectives of the project address three crucial directions: multilingual open domain QA, user-driven and context-aware QA, learning technologies for QA.
The potential of open domain QA will be experimented and evaluated in the context of mobile applications for information seeking, a multimodal scenario which includes spontaneous speech as input and the integration of textual answers with maps as for output. The selected domain is represented by local events in a town, usually available either through specialized web sites or local newspapers and publications.