2.4. By the Way: What is QALL-ME?


You can safely skip this section about the roots of the QALL-ME Framework. The section is not technically relevant for using or understanding the framework.

In the previous sections you should have gotten a pretty comprehensive technical overview of the QALL-ME Framework. Perhaps you may have wondered, though, about the name “QALL-ME”? This section will provide some background about the roots of the framework – and about the name “QALL-ME”.

The QALL-ME Framework is rooted in the EU funded QALL-ME project (http://qallme.fbk.eu/). Started in autumn 2006 the project’s general objective was to establish a shared infrastructure for multilingual and multimodal question answering (QA) for mobile devices. From this objective the project has also gotten its name: Question Answering Learning technologies in a multiLingual and Multimodal Environment – or just QALL-ME (pronounced as “call me”). The QALL-ME Framework represents the multilingual QA part of the project’s general objective. The machine learning part of the project name is just one possibility of implementing parts of the framework. Multimodal interaction with mobile devices (e.g., voice input or TTS) is outside the immediate scope of the framework: the QALL-ME Framework itself is agnostic to different interaction modalities; nonetheless, the QALL-ME project could show that the framework is well usable in applications with different input and output modalities.

The QALL-ME project consortium consists of four academic and three industrial partners. The academic partners are all renowned research institues in the field of language technology. They are mainly responsible for the QA part of the QALL-ME project, i.e., for the QALL-ME Framework. The three companies in the project make sure that the outcome of the project is also industrially exploitable and evaluate the integration of the developed QA system in real-world mobile application scenarios. These are the members of the QALL-ME consortium:

Fondazione Bruno Kessler (FBK), Human Language Technology research unit – Trento (Italy)
German Research Center for Artificial Intelligence (DFKI), Language Technology lab – Saarbrücken (Germany)
University of Alicante, Department of Software and Computing Systems – Alicante (Spain)
University of Wolverhampton, Research Group in Computational Linguistics – Wolverhampton (United Kingdom)
Comdata S.p.A. – Turin (Italy)
UbiEst S.p.A. – Treviso (Italy)
Waycom Srl – Albisola Superiore (Italy)

Since its beginning, the project has developed the QALL-ME Framework and several showcases that demonstrate the usability of the framework in different domains: the first QALL-ME prototype system has paved the way to a Service Oriented Architecture (SOA) for QA systems and integrated the QA core system into a scenario with mobile access to the system. In the second year of the project the framework has been slightly reworked and polished. The workflow and component architecture have been improved which has eventually led to the second QALL-ME prototype system. The framework which has been in use in this prototype system now comprises the first stable release of the QALL-ME Framework. In the first QALL-ME prototype the domain that the demo system could handle was restricted to movie events, i.e., questions about movies and schedules of local cinemas. For the second QALL-ME prototype the QA domain was augmented by accommodation data; now it was possible to ask questions about local accommodations, such as questions about prices, facilities or room availability. By changing the domain of the QA system we could show that the QALL-ME Framework has been designed generic enough to be widely usable and easily applicable to different domains.