Skip to main content

An intelligent guide to support productive online dialogue

Objective

One of the most important trends in CSCL/W is the proliferation of tools to support e-discussion due to their value in learning, negotiating, and other social applications. Discussions, however, do not necessarily lead to desirable results but often turn out to be ineffective or chaotic when no moderator/tutor is present.

The characteristics and demands of the role of the moderator/tutor have not been fully studied, especially in synchronous discussions using environments displaying graphical argumentation maps. Moreover, appropriate tools to assist moderators in these duties have not been developed to any satisfactory level. ARGUNAUT aims at unifying awareness and feedback mechanisms for working in e-discussion environments. The mechanisms will work on two existing (and potentially more) platforms: Digalo and Cool Modes. The feedback is primarily directed to a human moderator facilitating the interaction but may also help students.

At the heart of the AI components proposed, an off-line analysis mechanism based on machine learning techniques - the "deep loop" - takes selected human-annotated examples and generalizes them to indicators that can in turn be used for awareness feedback. Some semantic indicators will also be included. The evaluation scenarios will rely on the analysis of discourse arising from this setting. The impact: A big change in education and training programs at all levels, which will build to a larger extent on argumentative practices capitalizing on the proposed tools.

Using our tools, we expect to increase the learners per moderator by a factor of 2-3, although it is not our intention to focus on such a quantitative measure but to improve the quality of discussions and enhance their value for learning. Experimentation will be done in secondary, higher, professional and vocational education environments. The project, planned for 33 months, will be carried out by 7 partners: 3 univ., 2 R&D institutes and 2 industrial companies in 5 countries.

Funding Scheme

STREP - Specific Targeted Research Project

Coordinator

THE HEBREW UNIVERSITY OF JERUSALEM
Address
Mt. Scopus
91905 Jerusalem
Israel

Participants (6)

DEUTSCHES FORSCHUNGSZENTRUM FUER KUENSTLICHE INTELLIGENZ GMBH
Germany
Address
Erwin-schroedinger-strasse 57
67663 Kaiserslautern
FRAUNHOFER IAF
Germany
Address
Tullastr. 72
79108 München
POULIADIS ASSOCIATES CORP
Greece
Address
Sigrou Avenue 142
17671 Athens
SILOGIC S.A.
France
Address
6 Rue Roger Camboulives
31100 Toulouse
UNIVERSITAET DUISBURG-ESSEN - STANDORT DUISBURG
Germany
Address
Universitatsstrasse 2
47117 Essen
UNIVERSITY OF EXETER
United Kingdom
Address
Northcote House, The Queens Drive
EX4 4QJ Exeter