NVL

This position paper describes a long-term Technology-Enhanced Learning initiative at the Leeds Institute of Medical Education in which a personalised adaptive learning mentor will be deployed for all MBChB students enrolled in the course. The system, myPAL, is enriching the existing TEL programs embedded in the curriculum and will be leveraging recent advances in Learning Analytics and Open Learner Modelling. The paper presents the context of the project and the opportunities that deployment settings will offer, and highlights the research and development strands that will underpin it.

This position paper presents a forward-looking view on addressing a long standing professional learning challenge faced by higher educational institutions, namely assisting students to make meaning from work-based experience and develop as reflexive professionals. We suggest that a synergetic approach, building on existing research in professional lifelong learning and intelligent learning environments and taking advantage of new opportunities provided by emerging technologies, will underpin a new breed of intelligent mentoring systems for professional learning. They will foster the learners’ meaning making process, as well as assist tutors in their roles as coaches/mentors.

The proposal seeks to understand the personal behaviours, journeys and access points of Multiple Sclerosis (MS) citizens, in order to build out an eco-system for a Personal Data Store (PDS) and elicit issues around personal control over personal data. Research and recent reports highlight the urgent need for more integrated person-centred services as a means of delivering better patient outcomes, better clinical outcomes and better economic outcomes. Different implementation scenarios carry different configurations of cost, risks and benefits for different stakeholding gro ups, and the implementation of digital services has suffered in the past from lack of co-production or consultation with people and stakeholders on the ground before implementation. The proposed project will enable a group of citizen participants (plus organisations and their representatives) to interact in person-centred scenarios. These individuals may have long termconditions or professional interests with such condition – we have identified Multiple Sclerosis (MS) as a potential starting point – and we will identify needs, barriers, benefits and co-produce implementation scenarios.

Background Previous studies of frequency discrimination training (FDT) for tinnitus used repetitive task-based training programmes relying on extrinsic factors to motivate participation. Studies reported limited improvement in tinnitus symptoms. Purpose To evaluate FDT exploiting intrinsic motivations by integrating training with computer-gameplay. Methods Sixty participants were randomly assigned to train on either a conventional task-based training, or one of two interactive game-based training platforms over six weeks. Outcomes included assessment of motivation, tinnitus handicap, and performance on tests of attention. Results Participants reported greater intrinsic motivation to train on the interactive game-based platforms, yet compliance of all three groups was similar (∼70%) and changes in self-reported tinnitus severity were not significant. There was no difference between groups in terms of change in tinnitus severity or performance on measures of attention. Conclusion FDT can be integrated within an intrinsically motivating game. Whilst this may improve participant experience, in this instance it did not translate to additional compliance or therapeutic benefit. Trial Registration ClinicalTrials.gov NCT02095262

This paper reports on an application that delivers automated formative feedback designed to help university students improve their assignments. The aim of the system is to improve the confidence and skills of the user by promoting selfdirected learning through metacognition. The system focuses on the content of an essay by using automatic summarisation techniques, automatic structure recognition, diagrams, animations, and interactive exercises that promote reflection. The system is currently undergoing initial exploratory rounds of testing by ex-student volunteers and will be the subject of two full-scale empirical evaluations starting in September 2013. The main claims of this paper are the application and adaptation of graph-based key word and key sentence ranking methods for a novel purpose, and ensuing observations concerning the suitability of two different centrality algorithms for the purposes of key word extraction.

This paper reports on progress on the design of OpenEssayist, a web application that aims at supporting students in writing essays. The system uses techniques from Natural Language Processing to automatically extract summaries from free-text essays, such as key words and key sentences, and carries out essay structure recognition. The current design approach described in this paper has led to a more “explore and discover” environment, where several external representations of these summarization elements would be presented to students, allowing them to freely explore the feedback, discover issues that might have been overlooked and reflect on their writing. Proposals for more interactive, reflective activities to structure such exploration are currently being tested.

This paper presents observations that were made about a corpus of 135 graded student essays by analysing them with a computer program that we are designing to provide automated formative feedback on draft essays. In order to provide individualised feedback to help students to improve their essays, the program carries out automatic essay structure recognition and uses domain-independent graph-based ranking techniques to derive extractive summaries. These procedures generate data concerning an essay’s organisational structure and its discourse structure. We have selected 27 attributes from the data and used them in a comparative analysis of all the essays with a view to informing further development of the feedback program. The results of this analysis suggest that some characteristics of students’ essays that our domain-independent feedback program is measuring may be related to the grades that tutors assign to their essays.

The SAFeSEA project (Supportive Automated Feedback for Short Essay Answers) aims to develop an automated feedback system to support university students as they write summative essays. Empirical studies carried out in the initial phase of the system’s development illuminated students’ approaches to and understandings of the essay-writing process. Findings from these studies suggested that, regardless of their experience of higher education, students consider essay writing as: 1) a sequential set of activities, 2) a process that is enhanced through particular sources of support and 3) a skill that requires the development of personal strategies. Further data collected from tutors offered insight into the feedback and reflection stages of essay writing. These perspectives offer important considerations for the ongoing, iterative development of this automated feedback system and indeed, for any institution developing tools to support students’ writing.

OpenEssayist is a system which is currently under development. It aims to provide an effective automated interactive feedback system that yields an acceptable level of support for university students writing summative essays. The principal natural language processing technique currently employed is extractive summarisation using graph-based ranking algorithms. OpenEssayist will be piloted in September 2013 with Open University UK students following a Master’s course of study.

Many (2D) Dynamic Geometry Systems (DGSs) are able to export numeric coordinates and equations with numeric coefficients to Computer Algebra Systems (CASs). Moreover, different approaches and systems that link (2D) DGSs with CASs, so that symbolic coordinates and equations with symbolic coefficients can be exported from the DGS to the CAS, already exist. Although the 3D DGS Calques3D can export numeric coordinates and equations with numeric coefficients to Maple and Mathematica, it cannot export symbolic coordinates and equations with symbolic coefficients. A connection between the 3D DGS Calques3D and the CAS Maple, that can handle symbolic coordinates and equations with symbolic coefficients, is presented here. Its main interest is to provide a convenient time-saving way to explore problems and directly obtain both algebraic and numeric data when dealing with a 3D extension of "ruler and compass geometry". This link has not only educational purposes but mathematical ones, like mechanical theorem proving in geometry, geometric discovery (hypotheses completion), geometric loci finding... As far as we know, there is no comparable "symbolic" link in the 3D case, except the prototype 3D-LD (restricted to determining algebraic surfaces as geometric loci).

Difficulties communicating are common in everyday life. It is frustrating when you cannot understand someone at the pub or on a bad mobile phone connection. The education of children is hampered when they cannot understand the teacher because the classroom is noisy. The frequency and severity of these communication difficulties are increased for individuals with hearing impairments. Auditory assistive devices (e.g. hearing aids or cochlear implants) reduce some of these difficulties. Unfortunately, it can take many months of continuous use before patients achieve the maximum benefits. During this initial familiarization stage, many users grow frustrated and discontinue using their assistive devices. Our aim is to investigate how auditory perceptual learning, educational technologies and game design can be further combined into an approach of training that is suitable for use by individuals outside the laboratory, e.g. on home computers or mobile devices. Projects are underway to develop casual games for training on basic auditory tasks (e.g. discriminating between two frequencies or identifying the location of a sound source) and on more “realistic” listening tasks and social settings (e.g. speech intelligibility in “cocktail party” settings). The design methodology will be based on user-centric approaches, including participatory design, rapid and incremental prototyping, usability studies and formative evaluation. The efficacy of the design approach will be compared both from an auditory learning point of view (e.g. changes in performance) and from a user engagement point of view (e.g. flow experience).

The L4All system allows learners to record and share learning pathways through educational offerings, with the aim of facilitating progression from Secondary Education, through to Further Education and on to Higher Education. It provides facilities for learners to create, maintain and share their own 'timeline' (a chronological record of their learning, work and personal episodes) with other users, in order to foster collaborative elaboration of future goals and aspirations. This paper describes the design of the system's personalised mechanism for searching for 'people like me', presents the results from an evaluation session held with a group of mature learners, and identifies recommendations arising from this evaluation which have led to further development of the system.

The L4All project that preceded the MyPlan project developed a prototype system, the L4All system, which provided facilities for visualising, planning and reflecting on lifelong learning. It offered an online space providing information on learning opportunities in London, and a forum for learners to share information about each others’ learning and career timelines. The MyPlan project aimed to increase the value of the L4All pilot by researching, developing, deploying and evaluating (i) user models for lifelong learners, (ii) personalised functionalities for the creation, search and recommendation of learning pathways, and (iii) a game-based application to support learners in exploring the range of educational and career possibilities. This document reports on the second phase of evaluation of the new personalised functionalities of (ii) above. This evaluation phase started following the completion of Version 2 of the personalised system, which is described in Deliverable D4.3 of the MyPlan project. This present document (which is Deliverable 5.2 of Workpackage 5) presents our findings from two evaluation sessions that took place at the London Knowledge Lab on 29th July 2008 and at the College of North East London on 11th November 2008.

The MyPlan project aimed to contribute to the JISC e-Learning programme by developing, deploying and evaluating new techniques and tools that allow personalised planning of lifelong learning. The project brought together stakeholders from a broad range of institutions all of whom are committed to providing lifelong learning opportunities which enhance career development and widen participation. These stakeholders contributed to the formulation of user and technical requirements, and the evaluation of the tools developed by the project. The project had three major aims: (i) development and evaluation of learner models and an ontology for learner modelling in a lifelong learning context; (ii) development, deployment and evaluation of personalised functionalities for the creation, searching and recommendation of learning pathways; (iii) development and integration of a game-based application into the system, to give learners better understanding of the possible implications of different career decisions and educational choices. The project ran for 27 months, starting on 1st September 2006 and ending on 30th November 2008. It produced two successive versions of new personalised functionalities for lifelong learners. The software developed is in the form of components and services that extend the existing L4All system. A key feature of the project were our frequent engagements with users (several student groups from FE/HE institutions) and user stakeholders (the tutors of these groups, the project Advisory Group and project Partners, the JISC programme manager). These engagements were invaluable in informing the design of the system, and the aims and methodology of the evaluation sessions.

Version 1 of the L4All personalisation engine was described in detail in Deliverable D4.1 of the MyPlan project, and focussed on the first three issues above: search for people like me, customisation of the system and recommendation (only partially described in that deliverable). This document describes Version 2 of the personalisation engine and also aims at giving a general description of the current state of the L4All system. This document is organised as follows. First, the current architecture of the L4All system is described. Next, the overall GUI of the system is presented, highlighting the significant changes since Version 1. Third, the implementation of the personalisation engine, Version 2, is described, focusing on the new approach that is now being used for providing a personalised recommendation mechanism.

The use of kinaesthetic modes of learning is unquestioned in primary education, yet these are largely underused in secondary education despite considerable evidence of their importance for conceptual understanding for mathematics. There are rich realms of geometry to be learned in three-dimensional Euclidean space, and various “non-Euclidean” spaces, such as the surface of a sphere. It is also an area rich in potential for physical exploration with our bodies, based on kinesthetic thinking, by which we mean thinking that is mediated by real or imagined bodily motion and manipulation of objects. A core problem is that as the abstract conceptual content of education increases, the intellectual distance from physical activity increases, and it becomes a significant challenge to maintain the connections for learners. Inspiration for our approach is drawn from the 2-dimensional “Turtle Geometry” microworld that was developed in the 1970s-80s as part of the Logo programming environment. A turtle is a physical or virtual device that has a definite position in the 2-D plane, and a heading, and it carries a “pen” to mark the trace of its movement. One of the great insights in the design of Turtle Geometry was the recognition of “body syntonic” learning (Papert, 1980): that young children learn through the use of their bodies, and therefore a computational learning environment could simultaneously allow children to use body syntonicity to negotiate with the learning environment and promote them towards thinking about and through systematic, symbolic (programmed) structures. The interface device of the turtle provides the basic body syntonic interaction, and there are systematic geometrical principles “embedded” in the microworld which children discover/encounter through play and exploration (Noss & Hoyles, 1996).

The L4All system provides an environment for the lifelong learner to access information about courses, personal development plans, recommendation of learning pathways, personalised support for planning of learning, and reflecting on learning. Designed as a web-based application, it offers lifelong learners the possibility to define and share their own timeline (a chronological record of their relevant life episodes) in order to foster collaborative elaboration of future goals and aspirations. A keystone for delivering such functionalities is the possibility for learner to search for ‘people like me’. Addressing the fact that such a definition of ‘people like me’ is ambiguous and subjective, this paper explores the use of similarity metrics as a flexible mechanism for comparing and ranking lifelong learners’ timelines.

This deliverable covers workpackage WP4 (Development and deployment of personalised functionalities for planning of lifelong learning) and reports on the current state of the (re)design of the L4All system. Τhe developed personalisation engine will provide: (i) personalised search of timelines from "people like me"; (ii) personalised recommendation of which course(s) to study next; (iii) customisation of the delivery and presentation of contents. The design of the system is now reaching completion and is under internal testing before being transferred to the server that will be used for the online evaluation and user testing. This report was initially planned at month 12 but has been postponed to month 14. That was necessary in order to cover the extra work (2 months) required for the redesign of the user interface, which was not included in our original project plan. The main activities performed to achieve our targets in this workpackage included: · Redesign of the GUI of the L4All system, using DHTML/javascript for the front-end and JSP/servlet for the back-end. · Redesign of several aspects of the ontology underlying the L4All system to accommodate for the new functionalities: different categories of user (learner, expert, institution), a two-axis taxonomy of events, etc. · Design and implementation of a similarity measure engine for the comparison of learner’s timelines. The mechanism is based on converting timelines into string of comparable tokens and on using string metrics for ranking them. · Design and implementation of a recommendation engine, using the timeline formalism to represents requirements/recommendations and the similarity engine for proposing matches. · Design and deployment of several customisation procedures (colour/shapes used in the timeline visualisation, bookmarks for interesting timelines, etc.)

The L4All project has developed a prototype system, the L4All system, for visualising, planning and reflecting on personal learning and lifelong learning. It offers an online space that provides information on learning opportunities and career development pathways, and a forum for learners to share information and collaborate with peers and tutors. The MyPlan project aims to increase the value of the L4All pilot by researching, developing, deploying and evaluating (i) user models for lifelong learners, (ii) personalised functionalities for the creation, search and recommendation of learning pathways, and (iii) a game-based application to support learners in exploring the range of educational and career possibilities. This document reports on the first phase of the evaluation of the enhanced L4All system, that started upon the delivery of Version 1 of the system, as described in Deliverable D4.1. A second evaluation phase is planned for the summer of 2008, after delivery of Version 2 of the enhanced system. This document (which is Deliverable 5.1 of Workpackage 5) it presents our findings from two evaluation sessions that took place at Birkbeck on the 19th of February 2008 and at Community College Hackney on the 13th of March 2008.

L4All is a system that records and shares learning trails through educational offerings with the aim of facilitating progression of lifelong learners from Secondary Education, through to Further Education and on to Higher Education (HE). The focus is on helping those post-16 learners who have traditionally not participated in HE. L4All allows learners to access information and resources registered with the system by their providers, to plan their own learning, and to maintain a record of their learning. Tutors are able to register recommended learning pathways through courses and modules, thereby encouraging progression into HE. The system allows learners to share their learning plans and experiences with other learners (if they wish) in order to encourage collaborative formulation of future learning goals and aspirations.

Learner models, understood as digital representations of learners, have been at the core of intelligent tutoring systems from their original inception. Learner models facilitate the knowledge about the learner necessary for achieving any personalisation through adaptation, while most intelligent tutoring systems have been designed to support the learning modelling process. Learner modelling is a necessary process to achieve the adaptability, personalisation and efficacy of intelligent tutoring systems. This chapter provides an analysis of the migration of open learner modelling technology to common e-learning settings, the implications for modern e-learning systems in terms of adaptations to support the open learner modelling process, and the expected functionality of a new generation of intelligent learning environments. This analysis is grounded on the authors’ recent experience on an e-learning environment called LeActiveMath, aimed at developing a web-based learning environment for Mathematics in the state of the art.

Opening a model of the learner is a potentially complex operation. There are many aspects of the learner that can be modelled, and many of these aspects may need to be opened in different ways. In addition, there may be complicated interactions between these aspects which raise questions both about the accuracy of the underlying model and the methods for representing a holistic view of the model. There can also be complex processes involved in inferring the learner's state, and opening up views onto these processes - which leads to the issues that are the main focus of this paper: namely, how can we open up the process of interpreting the learner's behaviour in such a manner that the learner can both understand the process and challenge the interpretation in a meaningful manner. The paper provides a description of the design and implementation of an open learner model (termed the xOLM) which features an approach to breaking free from the limitations of "black box" interpretation. This approach is based on a Toulmin-like argumentation structure together with a form of data fusion based on an adaptation of Dempster-Shafer. However, the approach is not without its problems. The paper ends with a discussion of the possible ways in which open learner models might open up the interpretation process even more effectively.

We analyse how a learner modelling engine that uses belief functions for evidence and belief representation, called xLM, reacts to different input information about the learner in terms of changes in the state of its beliefs and the decisions that it derives from them. The paper covers xLM induction of evidence with different strengths from the qualitative and quantitative properties of the input, the amount of indirect evidence derived from direct evidence, and differences in beliefs and decisions that result from interpreting different sequences of events simulating learners evolving in different directions. The results here presented substantiate our vision of xLM is a proof of existence for a generic and potentially comprehensive learner modelling subsystem that explicitly represents uncertainty, conflict and ignorance in beliefs. These are key properties of learner modelling engines in the bizarre world of open Web-based learning environments that rely on the content+metadata paradigm.

This paper describes the design of the learner modelling component of the LeActiveMath system, which was conceived to integrate modelling of learners’ competencies in a subject domain, motivational and affective dispositions and meta-cognition. This goal has been achieved by organising learner models as stacks, with the subject domain as ground layer and competency, motivation, affect and meta-cognition as upper layers. A concept map per layer defines each layer’s elements and internal structure, and beliefs are associated to the applications of elements in upper-layers to elements in lower-layers. Beliefs are represented using belief functions and organised in a network constructed as the composition of all layers’ concept maps, which is used for propagation of evidence.

We describe XLM, the learner modelling subsystem of LEACTIVEMATH, from the viewpoint of how it makes use of technologies associated with the Semantic Web. We discuss how a better usage of these technologies could make of XLM a more generic learner modelling engine to serve a variety of elearning systems. We try to foresee important issues to be addressed and difficult problems to be solved in the way to this goal.

This report represents the Student Model Dialogue deliverable (D32) and describes in detail the improvements that have been made to the dialogue component of the Open Learner Model (OLM) since its original description in D29. In this document, the term “dialogue” in the Open Learner Model should not be understood in the broad sense of open-ended conversation, but rather as the verbalisation, in ordinary language, of the interaction between learners and the OLM, including the medium through which such interactions are assisted, such as the Graphical User Interface. The work on the prototype described in this document is the continuation – and conclusion – of two previous deliverables: the Student Model Specification (Deliverable D10 [6]) and the Open Student Model (Deliverable D29 [5]). For the readers’ convenience, this document includes a summary of the current state of the OLM. The core of this document is a description of two parallel lines of improvements that have been pursued since D29: augmentation of the Graphical User Interface of the OLM to better support interaction between the system and the learner, and the elaboration of the template-based verbalisation of various aspects of this interaction. The former follows the recommendations outlined in Deliverable 38 [7], Evaluation Student Model .

This document provides a description of the implementation of the Open Learner Model (OLM) as required by the workplan for LeActiveMath. The OLM is one of the component of the Extended Learner Model (xLM), the prototype learner modelling subsystem of LeActiveMath. Its aim it to provide LeActiveMath with an interface for the learners to access, explore and challenge the judgements that the xLM is holding about them. It supports several external representations for displaying the various sources of information that the xLM is using to establish its judgement (belief, evidence, etc.) and a mechanism - loosely based on Toulmin Argumentation Pattern - to control the exploration and negotiation of the beliefs. The OLM is implemented as a SWING applet, using the powerful resources of this Java GUI library in order to provide a suitable and usable interface for the learners. The implementation of the OLM conforms to the requirements generated by the project, as in Deliverable D5 [11], and to the specification of the Extended Learner Model, as in Deliverable D10 [17]. Justifications are provided for any divergence with both documents. For the reader’s convenience, relevant extracts from the Description of Work and from the requirements are summarised in section 1. The document is organised in four parts: • Section 2 describes the general principles of the OLM, i.e. which pieces of information – extracted from the Learner Model– are presented to the learner and how they are organised. • Section 3 describes the architecture of the OLM and its communication with LeActive- Math. For the reader’s convenience, information about the integration of the OLM with the front-end of LeActiveMath is also provided in this document. Such information will be replicated, together with similar description of the Learner Model, Situational Model and Learner History components, in Deliverable D31 (Integration of the Student Model in LeActiveMath ) for an overview of the Extended Learner Model. • Section 4 describes the Graphical User Interface (GUI) of the OLM, focusing on the external representations used to present the LM to learners, the mechanism used to control the interaction with the model and the mechanism used to support natural language. • Section 5 concludes the document by highlighting some of the important issues raised by the implementation and deployment of the OLM. The document contains also a transcription of a user-testing session (section A) and a short description of a tool that has been developed for debugging and testing the Extended Learner Model (the EventGenerator, section B).

This report provides a description of the integration of the Extended Learner Model (xLM) into LEACTIVEMATH, replacing the earlier ACTIVEMATH learner model component. It includes a brief description of xLM architecture in relation to LEACTIVEMATH and how information is exchanged between them via event messaging and procedure calls, followed by details of how each one of xLM components interact with LEACTIVEMATH. xLM simple interface allows other LEACTIVEMATH components to request information about beliefs held in learner and situational models, including the evidence supporting them. They can also request decisions on what is the more likely status of learner states or dispositions, suggestions on how much autonomy and approval to give to learners and details of the learner history. For more content-oriented components, xLM is able to provide beliefs and decisions about learner capabilities and dispositions in relation to individual content items, such as an explanation or an exercise. Most of xLM is integrated into LEACTIVEMATH even at the level of source code, meaning that most of xLM source code is included in LEACTIVEMATH source and it is compiled and deployed using the same mechanisms. The exception to the rule is the Situational Model component, whose source code is currently distributed, compiled and deployed independently.

This report is part of deliverable D30 Report and Prototype of Diagnostic Functionalities and provides a description of the implementation of the diagnostic functionalities of the Extended Learner Model (xLM), the prototype learner modelling subsystem of LEACTIVEMATH that replaces the earlier ACTIVEMATH learner model. It includes also an explanation of the rationale behind the design and implementation of xLM, and a discussion of outstanding issues. The diagnostic functionalities of xLM are built on top of a concept map that explicitly tries to represent the semantic structure of the subject domain and LEACTIVEMATH content. The formal description of the underlying structure of a mathematical domain is complemented with the operationalisation of a new standard for supporting and evaluating mathematical learning, based on the notions of mathematical competency and competency level. In addition, xLM is able to diagnose and model a variety of emotional, attitudinal, motivational and situational aspects of learning. Furthermore, by integrating the Open Learner Model delivered in D29 Report and Prototype of Open Student Model, xLM is able also to diagnose some aspects of the meta-cognitive skills of learners. The inferencing capabilities of xLM are developed on top of Bayesian inference and a variation of Dempster-Shafer Theory known as the Transferable Belief Model. The former has a short but solid tradition in learner modelling, and the latter offers the opportunity to model a broader sense of uncertainty about the learner and deal better with conflicting evidence. xLM simple interface allows other LEACTIVEMATH components to request information about beliefs held in learner models, including the evidence supporting them. They can also request decisions on what is the more likely status of learner states or dispositions, suggestions on how much autonomy and approval to give to learners and details of the learner history. For more content-oriented components, xLM is able to provide beliefs and decisions about learner capabilities and dispositions in relation to individual content items, such as an explanation or an exercise.

The terms dynamic representation and animation are often used as if they are synonymous, but in this paper we argue that there are multiple ways to represent phenomena that change over time. Time-persistent representations show a range of values over time. Time-implicit representations also show a range of values but not the specific times when the values occur. Time-singular representations show only a single point of time. In this paper, we examine the use of dynamic representations in instructional simulations. We argue that the three types of dynamic representations have distinct advantages compared to static representations. We also suggest there are specific cognitive tasks associated with their use. Furthermore, dynamic representations of different form are often displayed simultaneously. We conclude that to understand learning with multiple dynamic representations, it is crucial to consider the way in which time is displayed.

This document provides a specification of the Student Model as required by the workplan for LEACTIVEMATH. The specification relatesWork Package 4 Student Model (WP4) to the other LEACTIVEMATH components and describes its subcomponents. The Work Package main deliverable, the student modelling subsystem of LEACTIVEMATH, is renamed as the “Extended Learner Model” subsystem (xLM) and is composed of four subcomponents named Learner Model (LM), Learner History (LH), Situation Model (SM) and Open Learner Model (OLM). This specification conforms to the requirements generated by the project, as in Deliverable D5 (LeActiveMath Partners, 2004c). For the benefit of the reader, all requirements referenced in this document can be found in Appendix C. The project used a Claims Analytic approach to generating Deliverable D5 Requirements, and the WP4 team is continuing to work in this way. Several new claims for the Extended Learner Model are included in this specification in order to support design and implementation. This claims generation process will continue and will be extended along low-level design and implementation of the subsystem. In addition to addressing the issue of conformance of the specification with the requirements, this specification provides a view of the research background, the challenges to learner modelling in LEACTIVEMATH and the methodology employed to design the Extended Learner Model, as well as descriptions of the functionality and architecture for xLM and its various subcomponents. The work of WP4 is a combined effort by researchers from the University of Glasgow (formerly from Northumbria University), the University of Edinburgh and the University of Saarland. This document is constructed out of sub-teams working across the boundaries between these three Universities.

Many researchers have begun to call for environments to be developed which specifically recognize the distinctive contribution that multiple external representations (MERs) can bring to learning. In this paper, we present the conceptual framework that forms the basis for our approach to understanding learning with MERs and describe how we have embedded in the design of an instructional simulation that uses dynamic visualizations. We discuss briefly the results of an initial evaluation of the system and its implications for testing the framework.

DEMIST is a multi-representational simulation environment that supports understanding of the representations and concepts of population dynamics. We report on a study with 18 subjects with little prior knowledge that explored if DEMIST could support their learning and asked what decisions learners would make about how to use the many representations that DEMIST provides. Analysis revealed that using DEMIST for one hour significantly improved learners ’ understanding of population dynamics though their knowledge of the relation between representations remained weak. It showed that learners used many of DEMIST’s features. For example, they investigated the majority of the representational space, used dynalinking to explore the relation between representations and had preferences for representations with different computational properties. It also revealed that decisions made by designers impacted upon what is intended to be a free discovery environment.

Designing an effective Multiple External Representations (MERs) environment remains a difficult task, as research that has evaluated how they support learning has produced mixed results. Different approaches can be used to complete this effectiveness objective. One such approach, defined as Domain-Informed, may consist in starting with and fitting in pedagogical analysis of the domain, which required the active collaboration of experts in the design process itself. Such a participatory design was initiated for the conception of Calques 3D, a three-dimensional dynamic geometry microworld. This paper describes the external representations available in the software, their pedagogical purpose and the decisions and problems that have occurred during the participatory design process.

Learning environments use multiple external representations (MERs) in the hope that learners can benefit from the properties of each representation and ultimately achieve a deeper understanding of the subject being taught. Research on whether MERs do confer these additional advantages has shown that learning can be facilitated but only if learners can manage the complex tasks associated with their use. Our approach examines how the design of learning environments influences the cognitive task demands required of the learner with the longer term goal of using these findings to develop more adaptive and supportive multi-representational environments. In this paper, we begin by summarising the key features of the DeFT framework and then illustrate how such a framework can be used to classify existing systems. The main body of the paper describes the architecture of an instructional simulation that embodies DeFT. Finally, we conclude by illustrating the research questions we hope that experiments with this system can answer.

This PhD thesis takes place in the context of designing Interactive Learning Environment (ILE) where one of the diffculties is that the proposed solutions are hardly accepted by teachers because they are often not adapted to their own approach of teaching. Thus, our objective is to take in account the teachers needs directly in the design process, by identifying and categorising the different pedagogical needs, in order to provide the 'user-teacher' (intended to use the software) with an open environment on which he will be able to choose or define an adapted configuration. The application field we chose is spatial geometry learning, with the design of a microworld, Calques 3D. In order to reach our objectives, we collaborated, during the design process, with two groups of 'author-teachers' who enriched the prototype with their wide-ranging approaches of the same learning field. In order to ease and conduct the design of the software, we proposed a framework, named "utilisation contexts of educational software", that author-teachers are intended to use for expressing their choices on knowledge presentation and for describing activities they plan to organise around them. This framework allowed us in turn to extract the implicit knowledge that teachers have about "how and why they teach concepts in a particular way" in order to implement it and choose an adapted internal representation. So, it served as a negotiation support between teachers and software engineers in order to come to an agreement on the pedagogical content, in particular on the domain concepts and their external presentations.

This paper emphasizes the importance of capturing, i.e. making explicit, the knowledge that teachers implicitly use in teaching - the content as well as pedagogy. We describe a process for obtaining information from geometry teachers that will help us to better understand how they teach the spatial properties inherent in 3D geometry. This in turn enables us to improve the design of special software we have built (Calques 3D) based on their requests to have software that will help them bridge the gap between their current ways of teaching and the objective of having their students better prepared for the world of Computer Assisted Design. We describe the tools (forms) we use to capture this information and the results we experience - both the advantages and problems.

Calques 3D is a microworld designed for constructing, observing and manipulating geometrical figures. It provides the user with functions and visual feedback for seeing and understanding objects in the third dimension. It also allows him to dynamically construct geometrical figures from elementary objects and construction operations and, finally, to explore and discover the properties of the figure by deforming it.

Our major research aim is to analyse the teachers' needs and to propose a design process taking into account teachers as users. The main characteristics of this project are: 1) the development of Calques 3D, a microworld for spatial geometry learning, 2) a pluri-disciplinary work, 3) a framework to lead the negotiation between the different actors, and 4) a production cycle based on a step-by-step upgrading of a prototype. The ILE design has underlined the need of negotiation between the involved partners, since it may result in didactic or computational transpositions of the teaching domain. On the one hand, teachers have often to adapt the domain they teach (e.g. filter, simplification, metaphor, etc.). On the other hand, the software engineers have to restrict the required implementation due to development constraints or ergonomic requirements. This work allows us to focus on three design rules induced by the need for maintaining teachers' agreement. First, it appears that any pedagogical choices, even accepted during the design process, could be mistrusted by user-teachers. It implies that an adequate set of environments parameters should be implemented in order to allow the teacher to adapt the software according to his usage. Secondly, we consider that each constraint or property added to the teaching domain by the software engineers, for computational reasons, have to be expressed explicitly (e.g. by enhancing graphical interface rules). Finally, adding software features requested by teachers could increase the environment complexity and thus reduce its conviviality. So it appears essential to simplify user-interface either permanently or during a particular task (e.g. by curbing ILEs).

A Computer-Based Learning Environment (CBLE) needs to be adapted to several teaching styles, since this is a condition for acceptance and effective use in school. In this paper we propose to provide teachers with an opportunity for describing the learning sequences they plan to perform within the environment. Then, from these descriptions, a specific instance of the environment could be built and made available to the learners and teachers. To allow such learning sequences descriptions we need a common agreement on data, concepts and basic reasoning criteria that could be used. We describe such a process and the results we have obtained in the framework of a spatial geometry learning environment.

Un environnement informatique d'apprentissage doit pouvoir s'adapter à différents styles d'enseignement pour être effectivement intégré dans les classes et accepté par les enseignants. Nous pensons que les difficultés d'une telle adaptation proviennent de la diversité des points de vue sur chaque connaissance du domaine. Dans cet article, nous décrivons plusieurs exemples de cette diversité à différents niveaux de connaissances pour l'environnement de géométrie dynamique 3D que nous avons développé. A la suite d'une collaboration avec des enseignants, nous nous sommes particulièrement intéressés au cadre d'expression des informations collectées auprès d'eux, favorisant l'adaptabilité du logiciel aux diverses activités qu'ils souhaitent réaliser. La méthodologie de conception, décrite dans cet article, est basée sur la construction d'une ontologie de l'enseignement de la géométrie, déduite de contextes d'utilisation. Nous montrerons l’intérêt de cette spécification pour paramétrer un environnement selon des objectifs pédagogiques.

Nos travaux se placent dans le cadre du tuteur TALC dont l'objectif est de diagnostiquer la correction d'une figure géométrique construite par un élève vis-à-vis d'une spécification fournie par le professeur. Dans le cas où des constructions sont incorrectes, il parait très utile pour produire une bonne explication à l’élève de disposer d'un modèle de ses croyances en géométrie. Dans ce contexte des EIAO (Environnements Interactifs d’Apprentissage avec Ordinateur), il est ainsi évidemment intéressant de générer ce modèle automatiquement plutôt que d'en appeler à l'enseignant. La démarche que nous présentons consiste à utiliser comme «boite noire» des systèmes de Programmation Logique Inductive disponibles et de les expérimenter sur des conception erronées déjà analysées par des didacticiens.

Using the TALC geometric tutor, a student can obtain a diagnosis about the correctness of his construction of a figure with respect to a teacher's specification. In the case his construction is incor¬rect, our aim is to improve the explanations given by TALC using a student (mis)conceptions model. Since the knowledge representation in TALC uses first-order logic, more precisely Horn clauses, our pur¬pose is to define and to exper¬iment with the use of Inductive Logic Programming (ILP) systems as a 'black box' in order to generate such geometric student model, using a corpus from didactic research about per-pendicular symmetry.

Notre travail se place dans la cadre de la conception et de la réalisation du Tuteur d'Aide Logique pour la Construction de figures géométriques TALC. L'objectif de ce tuteur est de diagnostiquer la correction d'une figure construite par un élève vis-à-vis d'une spécification fournie par le professeur. TALC fournit actuellement trois types de diagnostics : la figure est correcte, la figure ne contient pas tous les objets et/ou propriétés spécifiées, la figure contient des propriétés non déductibles de la spécification (figure particulière). Dans le cas d'une construction incorrecte, la difficulté consiste à fournir à l'élève des explications pertinentes sur les raisons de cette inadéquation. L'approche présentée dans ce document est basée sur l'hypothèse didactique de la cohérence des raisonnements de l'élève et consiste à mettre en évidence les conceptions de la géométrie propres à l'élève. Du fait de la représentation des connaissances géométriques de référence dans TALC - et donc de la représentation des connaissances de l'élève - sous forme logique de clauses de Horn, cette modélisation des croyances de l'apprenant revient à trouver un modèle logique cohérent avec les productions de l'élève en vue de fournir une base d'explication et de négociation entre l'apprenant, le professeur et le système. Pour automatiser la recherche d'un tel modèle, l'utilisation de l'apprentissage automatique à partir d'exemples s'avère approprié. La représentation logique des connaissances nous a amené plus particulièrement à nous intéresser à un sous-domaine de recherche de l'apprentissage automatique: la Programmation Logique Inductive (PLI). La méthode utilisée pour ce travail est l'étude de la réutilisation d'un système de PLI aisément disponible pour l'inférence d'un modèle des croyances de l'apprenant.