Projects

The aim of this project was to lay the foundations for a new generation of systems that enable fast, efficient and robust natural language processing and are still sufficiently general. Based on the assumption that particular aspects of performance are grammaticalized, we pursued a novel approach to grammar where performance and competence aspects are already interleaved within the grammar model. In particular, we aimed at modeling the interaction of generativity which is the distinctive feature of competence, and lexicalization which is a feature of language usage. To achieve this goal, the influence of lexicalization on generativity was studied within the phenomenon of collocations. The interaction of lexical and structural information was modeled by means of corpus-based statistical techniques.

The objective of FASTY was the creation of a system for increasing the text generation rate of disabled persons by Predictive Typing and dedicated advanced input devices. The system was be developed for the German, French, Dutch and Swedish language, the concept, however, is useable for most European languages. FASTY assists motor, speech, learning and language impaired persons to produce texts faster, with less physical/cognitive load and with better spelling and grammar. FASTY is configurable for different types of disabilities, different communication settings and different European languages. It allows easier access to PC based office systems, to modern forms of IT communication and a faster usage of text to speech synthesisers for voice communication. Thus the FASTY system contributes to ensure equal access to Information Society for all Citizens.

The project studied the potential of animated agents for eLearning. In its first phase, a survey of current research and development in this area was undertaken, resulting in a pedagogically based categorization of animated pedagogical agents. The project then concentrated on those agents that carry most potential for pedagogical innovation, i.e. agents in non-pedagogical roles. These roles include agents as learning companions, teachable agents, and agents as actors in pedagogical role-plays. A demonstrator was built in the domain of job interview training for long-term unemployed people. The interactive role-play was implemented on the basis of the "Socialite" application and with the tools built in the framework of the European NECA project, which is coordinated by OFAI.

While in everyday life, humans communicate with their environment by language, spoken as well as written, supported by signs and gestures, human-computer interaction is still far behind. Currently, graphical user interfaces (GUI) are a de-facto standard; the logical next step is to move toward an even richer and more natural interaction by integrating communication via language. A prominent example for this need is demonstrated by the World Wide Web (WWW) which is of growing importance in everyday life. While developers of WWW pages may use any combination of text, audio, image, and video in their presentation to address the user - thus fully exploiting the multimedia possibilities of the web - the users' reaction is much more limited in being restricted mainly to point-and-click operations.This project aimed at showing new ways of integrating speech and language with classical access methods to the World Wide Web. Shortcomings and advantages of different access method combinations were investigated. As a testbed application, a system for providing access to German language newspapers available on-line was developed.

This project aims at developing a new, more flexible model of Data Mining, and an operational software architecture embodying this model. The problems that motivate this research project are two-fold: (1) a lack of flexibility of current methods and tools on the one hand, and (2) a lack of guidance through the highly complex and interactive process of data mining on the other hand.

SolEuNet aimed to develop a Web-based infrastructure for virtual enterprise services in data mining and decision support.

EUROMAP was a European initiative designed to conduct a current situation analysis of national and EU-wide Language Engineering (LE) activity and marketplace requirements and to make recommendations (based on the analysis of information gathered) regarding the focus and direction of future LE efforts and investments. Based on the findings of the first phase, the second phase (EUROMAP II) was devoted to strengthen contacts between the constituencies, to form a European-wide network for LE-issues, and to intensify information dissemination at EU and national level.

The task is to adapt the existing prototype to the needs of professional users, to introduce it into their knowledge management processes and to evaluate it, based on this adaptation and introduction. The goal is to learn how to further develop the application and to produce a marketable version. The main issues for this project will be usability, further adaptation of the artificial-intelligence-based tool, ontology handling and cross-lingual search. The prototype will be evaluated to see if the presumptions of the development until now where correct or if some adaptations need to be made. This includes the features, the usability of the 3D interfaces, the qualities of results, which are represented, the amount of data to be included in the system, and the working with ontologies. Part of the proposal will be the development of new benchmarking indicators for validation of the system in cooperation with the users.

Meta-learning is concerned with the selection of a suitable learning tool for a given application. In the MetaL project, a meta-learning assistant was developed for providing user support in machine learning and data mining.