Work Package Deliverables
WP1: Robust Learning in Elementary Mathematics
D1.1
State-of-the-art for intelligent tutoring systems and exploratory learning environments
Submission date: 31 October 2013
This deliverable presents a key building block in the iTalk2Learn project by preparing the ground for the domain-specific and pedagogical aspects that facilitate the development of the exploratory learning environment and its tasks as well as the intervention models for the learning platform. The emphasis is on definitions, to provide a common ground for partners when preparing for the work around adaptive intelligence (WP2), intuitive interaction interfaces (WP3), planning evaluation activities (WP5) and to a lesser extent deployment and integration of the learning platform (WP4). After a general introduction (Section 1), Section 2.1 defines procedural and conceptual knowledge – the two components students require for robust mathematical learning. It explains how, through the lenses of a cognitive science and a mathematics education perspective, procedural knowledge (knowledge about and application of procedures) and conceptual knowledge (implicit or explicit understanding about underlying principles and structures of a domain) develop iteratively through different pedagogical approaches to create robust mathematical knowledge.
D1.2
Report on learning tasks and cognitive models
Submission date: 31 January 2014
This deliverable reports on the design of the exploratory tasks that will be used in the learning platform and explains how existing content from Whizz and Fractions Tutor are being reused and interweaved with appropriate exploratory activities on the project (T1.2). It also reports on the related task of the project of the identification and operationalisation of relevant topics, learning objectives and problem solving strategies for elementary mathematics (T1.3).
D1.3
Pedagogical intervention model that unifies structured and exploratory learning approaches
Submission date: 30 April 2014
This deliverable describes the pedagogical intervention model (cf. Mazziotti, Loibl, & Rummel, in press) underlying the technical development within the iTalk2Learn project (cf. D2.2.1 for sequencing and support, D3.1 for speech recognition, and D4.2.1 for the platform development), and thus accomplishes milestone MS4 (pedagogical interventions). With our intervention model we aim to help students develop robust knowledge about fractions. In order to develop robust knowledge, students need to gain procedural and conceptual knowledge, which evolve differently and require different types of instructional support (cf. D1.1). Procedural knowledge can be acquired through repeated practice and deepening of problem-solving procedures. In contrast, conceptual knowledge can be facilitated by providing students with exploratory learning activities and encouraging reflection. Therefore, students need to be provided with structured and exploratory learning activities for developing complete robust knowledge.
WP2: Adaptive Intelligence for Robust Learning Support
D2.1
Report on state-of-the-art of and requirements for machine learning methods for intelligent tutoring systems
Submission date: 31 October 2013
This deliverable reports on the state-of-the-art in machine learning methods for intelligent tutoring systems, and discusses requirements for these systems. The state-of-the-art section is divided into two parts. The first part describes approaches in machine learning for tutoring support in Intelligent Tutoring Systems. The second part describes approaches to intelligent support in exploratory learning environments. Following this, the requirements for adaptive intelligence in robust learning support are outlined. This last section covers the requirements relating to (1) target tasks to support; (2) transaction on data collection; and (3) background data collection.
D2.2.1
Initial report on methods and prototype for adaptive intelligence for robust learning support
Submission date: 30 April 2014
In this deliverable we address the problems performance prediction, sequencing and adaptive support. For performance prediction we explain the machine learning method Matrix Factorization and how it can be applied to Intelligent Tutoring Systems, especially to the Maths-Whizz data set. The main advantages of Matrix Factorization for performance prediction are its domain independence, small computation costs and no need of an exploratory corpus. Our proposed sequencing approach uses this performance prediction method in addition to a policy based on the theory of Vygotsky’s Zone of Proximal Development.
D2.2.2
Final report on methods and prototype for adaptive intelligence for robust learning support
Submission date: 31 October 2015
WP3: Intuitive Interaction Interfaces for Elementary Mathematics
D3.1
Requirements and state-of-the-art of speech production and voice recognition for ITS
Submission date: 30 April 2013
Automatic speech recognition for children is used in a growing number of commercial and academic systems, ranging from entertainment, like toys and games, to educational systems. This deliverable presents an analysis of the state-of-the-art of speech recognition systems applied to tutoring systems for children. There is a reasonable number of such systems in the area of helping children to improve their language skills, like the various reading tutor systems. A consensus among researchers seems to be that speech recognition for children is still a hard task.
D3.2
Requirements and technical state-of-the-art on intuitive interaction interfaces for robust learning
Submission date: 31 July 2013
Work Package 3 (WP3) envisions to provide the technical means by which more exploratory, conceptually-oriented learning activities can be integrated with the rest of the platform and enable research on the role of the different modalities of speech and direct manipulation as well as multiple and alternative representations in learning elementary mathematics (and particularly the domain of fractions as selected by the consortium).
D3.3.1
Initial report on methods and prototype for advanced speech recognition models for young learners
Submission date: 30 April 2014
This deliverable presents information on the intended use and integration of automatic speech recognition (ASR) within iTalk2Learn. The particular target domain – mathematical tutoring in the domain of fractions – as well as the target users – young learners in England and Germany – present specific requirements and challenges for ASR-operation, model training and evaluation. The report provides background information on the methods and principles of speech recognition model training and evaluation and presents initial results of a first prototype created for the fraction tutoring domain. Furthermore, data collection efforts and acquired data resources are presented and the next steps regarding model training and evaluation as well as the final mechanics of the system are outlined.
D3.3.2
Final report on methods and prototype for advance speech recognition models for young learners
Submission date: 31 October 2015
D3.4.1
Initial report on methods and prototype for intuitive interaction interfaces for robust learning
Submission date: 30 April 2014
Work Package 3 (WP3) envisions to provide the technical means by which more exploratory, conceptually-oriented learning activities can be integrated with the rest of the platform and enable research on the role of the different modalities of speech and direct manipulation as well as multiple and alternative representations in learning elementary mathematics (and particularly the domain of fractions as selected by the consortium).This deliverable reports on methods and technologies for creating intuitive interfaces for robust learning with a focus on following activities: (1) behavioural interaction interpretation for enhanced learner understanding; (2) speech production; (3) GUI framework for exploratory learning activities in elementary mathematics; (4) harmonisation of GUIs and enhancements on usability of the iTalk2Learn platform.
D3.4.2
Final report on methods and prototype for intuitive interaction interfaces for robust learning
Submission date: 31 October 2015
WP4: Deployment and Integration
D4.1
Technical description of the first prototype of the iTalk2Learn platform
Submission date: 30 June 2013
iTalk2Learn will provide a strong platform with high level integration (scalability) with new technologies and services; and aims at simple maintenance for future development. It will be based on the Service Oriented Architecture (SOA) paradigm to achieve our goals. It will have an interface developed in HTML5 at the presentation layer, to make it available to most popular devices nowadays.
D4.2.1
Initial technical report, end user manual and prototype of the iTalk2Learn system
Submission date: 30 April 2014
This deliverable describes the integration platform that is responsible of bringing together all components in the iTalk2Learn system as well as the two different kinds of learning content (structured vs exploratory). The integration platform integrates both kinds of content, developed using different technologies (Flash, HTML5, Unity) to provide a seamless learning experience to students.
D4.2.2
Final technical report, and user manual and prototype of the iTalk2Learn systems
Submission date: 31 October 2015
WP5: Data Collection and Evaluation
D5.1
Report on Evaluation Plan
Submission date: 31 October 2013
This deliverable reports on the iTalk2learn evaluation plans. We first define our evaluation goals. Subsequently we describe two interlinked plans: the formative evaluation plan and summative evaluation plan. The section on the formative plan describes the iterative design and test process involved in developing adaptive sequencing and support mechanism, exploratory learning activities, and infant speech recognition. The summative evaluation plan defines appropriate methodologies to test the new technology in two experiments, using quantitative measures.
D5.2
Report on Formative Evaluation Results
Submission date: 31 October 2014
The formative evaluation aims at evaluating the effectiveness and usability of the newly-developed components of the iTalk2Learn platform, namely exploratory learning environment (WP1 & WP3), speech recognition for young learners (WP3), and automatic adaptivity concerning sequencing the tasks (WP2), switching between exploratory and structured tasks, and support functionalities (WP1). Simultaneously the deployment and integration of the components so far (WP4) was tested. Deliverable 5.2 reports on the formative evaluation activities conducted in Y2. The results of the formative evaluation informed the iterative development process, and form the basis for the summative evaluation in Y3.
D5.3
Report on Formative Evaluation Results
Submission date: 31 October 2015
The results of the summative evaluation clearly demonstrate that the combination of structured and exploratory tasks promotes learning more than structured tasks alone. This result was replicated in both educational contexts which underlines the effectiveness of iTalk2Learn to foster robust fractions knowledge in students. The results also showed that the role of speech is more complex than previously thought: in the United Kingdom, the version with speech adaptivity fostered learning more than the version without speech adaptivity, albeit not significantly so. In Germany, the version without speech adaptivity fostered learning more than the version with speech adaptivity. One possible explanation is that the benefits of speech depend on the prior knowledge of students: they may be more pronounced for students with low prior knowledge. This and other explanations will be investigated in further analyses.
WP6: Dissemination and Exploitation
D6.3.1
Report on Dissemination and Contribution to Standards, Year 1
Submission date: 31 October 2013
This deliverable provides an update on D6.2 focusing on the main dissemination activities carried out as of M12 of the project and some of those planned to be accomplished in the near future. It also provides a plan for the contribution the project is making with respect to related standards.
D6.3.2
Report on Dissemination and Contribution to Standards, Year 2
Submission date: 31 October 2014
This deliverable provides the second dissemination report focusing on activities carried out between M13 and M24 of the project. It also reports on the use of standards and the contribution that the project is making in this area, and identifies some key dissemination steps for the final year. Dissemination activities included events such as conferences and workshops, publications, and links with other projects. In addition, our formative evaluation work in schools (see D5.2) throughout Y2 also functioned as an on-going dissemination opportunity. Finally, we also undertook a major overhaul of the website in order to reflect key project outcomes and to make the project as accessible as possible to the widest audience (including academics, developers and teachers worldwide).
D6.3.3.
Report on Dissemination and Contributions to Standards, Year 3
Submission date: 31 October 2015
WP7: Project Management
D7.2
Report on Project Evaluation Plan
Submission date: 31 October 2013
In this deliverable we will discuss metrics of success for the project. We will discuss the internal and external metrics, together with the impact that the project will have on the market. Direct connections are made to WP5 and WP6 deliverables that contain more detailed descriptions of the success pursuing activities that have been completed and planned by the consortium.