ATuF

The objectives of the project ATuF are the development and the empirical evaluation of adaptive and tutorial feedback components (ATuF components) and strategies. In the first phase of ATuF, apart from domain analysis and representation, we have investigated parts of the user modelling and of the diagnosis of errors. Furthermore we have developed ATuF components and examined their effects within exercises with typical errors while conducting first experiments.

In the second phase we are investigating the following empirical and technological requirements for the adaptation of feedback and first forms of the real adaptation:

• (a) Investigation of the effects of ATuF components and strategies depending on the characteristics of the exercises and the learners
• (b) manual and automatic analysis/ data mining of the log files from the experiments
• (c) empirically sustained and extended user modelling as basis for the adaptation of feedback
• (d) extension of the competence, diagnosis as well as diagnosis of motivational and other variables
• (e) empirical investigations of various dimensions of the feedback adaptation
• (f) adaptation of the feedback strategies in relation to the characteristics of the exercises and the learners
• (g) evaluation of the developed adaptation dimensions
• (h) empirical investigation and technical requirements for the feedback and its adaptation within various types of exercises

Working on these issues is done using an interdisciplinary approach and the learning platform ActiveMath within ATuF.

Math-Bridge

Math-Bridge project page

In most European countries there is a high demand for tailored remedial teaching materials for mathematics enabling the transition of students from schools to higher education, in particular engineering students. As a rule, existing content for remedial mathematics is available in a single language only, rarely online and badly accessible. Moreover, it is represented in multiple formats, in various notations, and cannot be tailored to the learners' needs.

Math-Bridge aims at changing this situation to the better and helps to bridge the gap between schools and higher education in Europe. It will provide multi-lingual and multi-cultural semantic access (e.g. search and course generation) to remedial mathematics content which adapts to the requirements of a learner and his/her subject of study. It will bring together content from different European sources and offer it in a unified way. This access will be provided through a sustainable Pan-European learning service for remedial mathematics, which will be built by collecting appropriate learning resources, extending them in terms of structure and multi-linguality and making them useful and easy-to-find. The extended formats of the content will make a wider use of standards and, hence, will make this content re-usable and “transferable” between different learning environments. In order to achieve its goals Math-Bridge will study the (target) competencies required for target subjects of study, adapt existing semantic and multi-lingual search software, tailor assessment tools and methodologies, and adjust the cutting-edge ActiveMath learning environment to the remedy-scenario which includes specific diagnostic means and decisions for the transition from school to higher education. The service will be able to adapt to the level of learner competences and interests.

Moreover, Math-Bridge will enable collaborative authoring of the content on the basis of Creative Commons’ licenses and improve instrumental support to collaborative authoring. This will stimulate collaborative production and assembly of educational content, which, we believe, is a future must. The results will be usable way beyond mathematics.

Intelligent Support for Authoring Semantic Learning Content

Content authoring for adaptive educational systems is a tedious and expensive procedure. It requires from the author to have a high level of expertise in the subject domain, knowledge engineering and metadata standards. Over the last decade considerable efforts have been put into improving authoring practices through the development of authoring tools and proposal of metadata standards. Yet, to a large degree, the existing collections of learning objects suffer from a number of problems. Lack of metadata limits the use of learning content in adaptive systems and impedes its discovery, sharing and reuse. Another set of problems results from the design of existing authoring tools. While providing sophisticated functionality they have limited support of the author herself. Intelligent authoring support will allow a wider population of authors to participate in semantic learning content creation. Most of the existing authoring tool are developed to be used by a single author. This paradigm disregards the success of collaborative authoring patterns introduced by Social Web applications.

This project will address the problem of semantic learning content authoring from several perspectives and will employ a set of technologies to provide comprehensive and intelligent support of the authoring process:

• collaborative authoring support;
• intelligent support of metadata creation;
• open-corpus content discovery.

AdaptErrEx

AdaptErrEx project page

Exploring the Learning Benefits of Erroneous Examples and their dynamic adaptations within the context of middle school mathematics, the AdaptErrEx project, funded by the U.S. Department of Education, has developed materials to support learning of decimals through erroneous examples. The main goals are to support deeper, more robust learning and to improve error-detection skills.

All the materials developed on the AdaptErrEx project will eventually be freely available through MathTutor, a website that helps middle school students learn math.