The goals of our research on task ontology are to exemplify the benefits of task ontology through the development of an ontology based authoring tool for Computer Based Training CBT systems. In this paper, we will discuss the basic issues on the concept of task ontology and then describe the design principle of an ontology-based authoring tool for Computer Based Training CBT systems. To the list of all publications. The target task of SmartTrainer is mainly to recover the accidents of substations in the electric power system. When an accident happens, the electric power transmission will be interrupted, and the operators should recover it as quickly as possible.
The operators should find the spot of the accident, continue to supply the electric power to some special places such as hospital, police station at once by borrowing some power from the other substations, find the causes of the accident and recover it within the limited time. The goal of the training oriented by SmartTrainer is to improve capability of not only skill-based or rule-based reasoning but also knowledge-based reasoning.
The set of the scenarios incorporated into SmartTrainer has been designed by the experienced trainers. In order to let the trainee master the principled knowledge, SmartTrainer let them do practice first and then teach them the first principle behind it adaptively to their mistakes, and finally, check their learning result by practice training again. Here we want to emphasize that the training we give to the trainee has the time-limitation just like in real accidents.
Multi-media technique has been widely used in SmartTrainer to attain high fidelity, including the sound processor to create mock buzzer when an accident happens, the movie display to show the accident scene when the repairing man needs, the picture processor to create the static graphics of the various equipment, and so on. SmartTrainer is composed of five parts, those are human interface, authoring, training model based on the training ontology, teaching materials model based on the teaching materials ontology and simulator.
Here we will discuss the designing of authoring environment based on task ontology in SmartTrainer mainly. It is based on the idea that knowledge should not be simply transferred from a teacher to a learner but be built in a learner's head while interacting with each other in group activities. Although collaborative learning is not a new concept which has been carried out in classrooms for long time, it is expected to be a new promising paradigm in AI in ED community.
Although all of them are equally important, computers have nothing new to do with the first one. The last issue is interesting, but it is very ambitious, since it requires almost complete natural language understanding capability of computers in order to build an operational system. According to this observation, we take the second one as a target to realize in our research. Our research is mainly concerned with the following three major goals:.
In order to achieve these goals, we do need a sophisticated vocabulary in terms of which we can describe objectives and modes of communication, decision models, knowledge for decision making, etc. This implies we first design ontology for CSCL. Needless to say, AI techniques based on symbolism need primitives or a set of basic vocabulary for representing knowledge and objects. They reflect conceptualization of systems under consideration.
One might think that ontological issues must be far into domain and hence it is domain-specific and loses generality. By ontology, however, we mean a system of basic vocabulary usable across various domain knowledge, that is, "generic task". Working hypothesis of this research is that we can find a good ontology for CSCL task by looking at the task carefully from generic task point of view. What should be notice is to design a good ontology to represent the domain knowledge, the communication model, and the learning process model from the educational point of view. We are currently engaged in developing the intelligent support system for collaborative learning of physics domain.
A large number of useful techniques which contribute to realization of educational activities have been proposed for the development of ITS. Those achievements contributed not only to the promotion of ITS research itself but also to the establishment of various fundamental techniques in artificial intelligence.
It is only in recent years, however, that there have been discussions on the general problems in the design of ITS. The major objective of this research is the enumeration of the computational agents needed for implementing reactive behavior of tutoring systems. The design and development of a domain-independent framework for ITS posing a number of questions such as "What kinds of inference schema are needed in ITS? To answer these questions, it is necessary to view the problem in a top-down way based on the concept of gentric task. By generic task, we mean a system of domain-independent but task-dependent vocabulary, which is defined in both knowledge level and symbol level.
The symbol level constructs of the generic task are referred to as building blocks. In the current implementation, FITS , which stands for Framework for ITS, is composed of six building blocks, each of which covers an essential task for teaching. Narrative Concepts are concepts that are utilized within the narrative description process.
This activity may use resources and tools that are simulation-based and collaboration-based. This flexibility allows the course developer to rapidly build adaptive courses which contain both simple and complex storylines plots. Narrative Attributes Narrative Attributes consist of adaptive axes, adaptive techniques, associated descriptions and usage guidelines as illustrated in figure 4. Adaptive Axes are high-level descriptions of learner and learning environment characteristics to which narrative concepts can be adapted. Adaptive Techniques are the low-level mechanisms which adaptive axes can use to perform an adaptive task.
Narrative Concepts are used to create the custom teaching structure for a non-adaptive online course. To make an online course adaptive, the course developer must choose which sections, concepts or learning activities they wish to be adapted to the learner. Narrative Attributes can be used to describe the behaviour of a Narrative Concept. A narrative attribute may, for example, be used to describe some adaptive context in which the Narrative Concept will exist. Figure 4. Narrative Attributes are key elements in the conversion of a non-adaptive online course to a personalized adaptive online course.
Figure 4 illustrates the logical hierarchy of Narrative Attributes. The Adaptive Techniques reference a set of potential learning resource candidate selectors that may be used. The selectors are functionally exposed through a service-based architecture. Selectors are passed a list of parameters to reason across, for example, the return type of the selector, the ontological elements to reason across and potentially infinite other parameters.
Narrative Structures Instructional Design Principles, Pedagogical and Andragogical theory formalize and describe learning and teaching strategies. Narrative Structures are a model-based representation of theses descriptions. The models can be used as templates when constructing an online course and the descriptions can be presented as usage guidelines for the strategy. The combination of guideline and model can be used during reconciliation and validation of the online course. Narrative Structures are used to provide the course developer with a solid foundation, based on sound pedagogical and instructional design principles, from which to build their online course.
These models are interpreted to produce real-time support for the course developer. This support forms a framework for the online course based on the selected narrative structure s. The use of Narrative Structures allows the course developer to produce online learning based on single or multiple instructional design principles.
The general course structure may follow a didactic approach, however within the scope of this course their may be lessons that are best taught using different pedagogical approaches, e. One key challenge of online learning is to facilitate the reuse of all learning resources within a knowledge domain. Narrative Structures are formalized metadata models outlining the general narrative concepts and the flow of narrative concepts outlined by a particular instructional design strategy. They can be used in whole or as part of a customized teaching strategy. They offer guideline support to the course developer by offering usability information.
Narrative structures can then be used by course developers to share their particular teaching strategy for a domain of information. The desired effects from each and modelling principals are quite different yet both are equally important to the learning experience. The role of the learner is fundamental to an active learning pedagogy which specifies a learner-centric, constructivist learning environment.
The tutor is fundamentally involved with forming the scope of, providing guidance to and defining the learning objectives of the learning experience. The illustration in figure 5 shows the input types, based on learner and teacher involvement, that influence the learning experience. The learner model captures information about the prior knowledge, competencies, goals and capabilities of the learner while the teacher model captures information about preferred teaching strategies and learning goals.
Both of these models are queried during the composition of the learning experience. Influential factors in the Learning Experience Learner Constructivism involves the learner becoming active and interactive within their own learning experiences to develop their own understanding of the knowledge domain [Jonassen ]. One key goal of the multi-model approach to personalized eLearning taken at Trinity College Dublin involves the empowerment of the learner.
The learner should be in control of their learning experience and should have the capability to modify and abstract their personal learning path. Through learner empowerment [Bajraktarevic et. The Learner Model LM is defined as a schema representing the characteristics of a learner that must be modelled. The schema will define the structuring of the LM to provide a mechanism for cross-session interoperability and consistency.
Since the LM is only consulted during the decision making phase of the candidate selection process, the main influence of the attributes of the LM will be the narrative concept space since it is here that the adaptive axes are applied to the narrative concepts. Teacher Through the ACCT the ability to empower the teacher within the learning experience can be realized using a teacher model schema.
The Teacher model can be used to scope the course towards a group of learners or the curriculum of the domain ontology. It allows the course developer to specify semantic boundaries around the information space. The Teacher model will also influence the learner modelling instrument. Based on recommendations made by the Teacher, the pre-course questionnaire can be dynamically generated in line with the tutor restrictions.
The Teacher model will also feed into the candidate selection process, i. The learner model would then reflect the curriculumized decisions of the teacher. The teacher model schema can be automatically generated using the ACCT. The ACCT creates a teacher model schema by creating a translated view of the graphical Narrative Model representing the aspects of the adaptive course that be influenced by the teacher.
The teacher model schema provides the foundation and structure for the teacher model allowing the course developer to place curriculumized guidelines on the adaptive course structure. Learning Activities With the growth in online learning, distance learning and adaptive learning, the paradigms of instructional design are evolving [Reigeluth ].
In order for the learner to acquire higher order cognition skills analysis, synthesis and evaluation , the need for instructional design which facilitates, promotes and supports activity based learning must be realized. Through online learning and eLearning we can provide a more active learning experience, promote active learner involvement and encourage self motivation. Learning Activities typically consist of some form of task s , associated tools which could be used to perform the task s , and appropriate learning content.
Typically Learning Activities require some intuitive sequencing of operations. This sequencing describes the flow between the sub-activities within the Learning Activity. Applying this approach, Learning Activities can be structurally modelled to provide reusable, scaleable and customizable units of instruction. Figure 6. Learning Activity workflow within the Narrative Model In order to flexibly incorporate Learning Activities into the personalized course composition process it was important to design a flexible and descriptive Learning Activity model.
The model contains a description of the Learning Activity, the type of the Learning Activity atomic or composite , the types of outcomes it can provide and the types of communications tools available. These activities can take the form of an atomic activity e. Associated with an atomic activity is a description of the types of communication tools available, for example, email, chat, instant messaging, forum, etc.
This flexible modelling approach increases potential for reusability, accessibility and interoperability of Learning Activities. This means that a composite Learning Activity can only be used as a whole unit. Within the coming months the flexibility of the Learning Activity model and potential for learning resource reuse will be realized with the creation of a Learning Activity composition workspace within the ACCT.
From current work in adaptive hypermedia and personalized eLearning it is evident that there are two areas of research which need future development, the design of pedagogically sound courses and the support offered to the course developer during the composition of pedagogically sound courses. The ACCT is a design-time tool which allows the course developer to create adaptive and non-adaptive activity-oriented courses based on sound pedagogical strategies in a developer-supported environment. The architecture of the ACCT is built upon a reusability-focused, developer-supported and service-oriented architecture.
For example, the ACCT allows the course developer to interact with the learning resource repository, searching for candidates based on keywords and contextual prior use, through a web-service interface. The abstraction mechanisms employed by the ACCT allow the course developer to define their teaching strategies and subject matter domains in a reusable and collaboratively supported way. This active promotion of reusability not only at the asset level but also the pedagogical, instructional design, concept and activity level will aid in the rapid construction of pedagogically sound online adaptive learning experiences.
Figure 7. The guidelines will identify and describe the abstract logic and reasoning behind the conceptual layout of the course. The guidelines are also represented in model form whereby the course developer can see and interact with the model structure during the creation of their customized course narrative. The developed model guidelines, or schema, will be translated into the model support framework for the adaptive hypermedia authoring architecture of the ACCT.
The architecture of the ACCT, as illustrated in figure 7, follows a web services paradigm. This provides access to modelled pedagogy, subject matter domain, learning activities, content and adaptivity. The course developer can then use the available modelled information to compose an adaptive course narrative. The sample methodology in figure 1 outlines an adaptive course construction process whereby the course goals and objectives are initially identified, a pedagogical strategy s for the course is chosen, the subject matter domain is modelled and applied to the chosen pedagogy s , the learning resources are selected, the adaptivity is applied to the pedagogically-based course structure and the course semantics are tested.
This rapid course prototyping approach can be achieved with the ACCT as depicted in figure 7. The ACCT actively supports the course developer during the creation of the SMCS through facilitating addition, deletion and modification of subject matter concepts. The relationships are provided as a set of guidelines that the course developer can utilize to created relationship definitions. These relationships however can be customized. The ACCT allows the course developer to create and define new customized relationships, hence offering more control to the course developer during the course construction process.
Figure 8. It illustrates that the concepts within the space can be graphically and logically grouped with associated defined relationships. The concepts are listed on the left hand side and the logical layout is assembled on the right hand side. Customized Narrative Model Creation The custom narrative model builder is used by the course developer to describe the course structure in pedagogically-supported narrative terms.
The course developer is supported with a drag and drop interface providing tools built from sample pedagogical models, pedagogical narrative concepts, narrative attributes, previously defined subject matter concept space model, learning activities and collaboration paradigms. A learning resource repository interaction service is provided allowing the course developer to search for learning resources. A Narrative Structure consists of a collection of Narrative Concepts.
The Narrative Concepts allow the course developer to apply aspects of pedagogical strategies to certain parts of the adaptive course. By representing different pedagogical approaches as a workflow of concepts and learning activities the models provided by the ACCT become fully customizable and can be used to create hybrid pedagogies by blending different flavours of different pedagogies.
As depicted in figure 9, the ACCT pedagogically supports and guides the course developer during the design of the custom course narrative by providing a palette of fully customizable sample pedagogical models. The sample models provided are used to from the basis for the customized course narrative.
Narrative Structures have been created to represent pedagogical strategies such as case-based, didactic and web-quest teaching. This approach implies that the course developer has the flexibility to apply a blend of pedagogical strategies. This flexibility empowers the course developer with a tool that is capable of creating complex, and realistic, pedagogically- sound adaptive course offerings. The course developer will be offered guidance on how to best use such Narrative concepts within the scope of the sample pedagogical model. Based on course developer preference, all or part of the supplied sample pedagogical model can be used.
While constructing a course narrative the previously defined subject matter concept space is always available in the tools palette as seen in figure 9. By dragging a subject matter concept into the graphical narrative model, its associated metadata descriptions and relationship information are made available to the Narrative. This information can be then used by any applied adaptivity.
The current version of the ACCT provides support for learning activities in unit form. Each learning activity is viewed as an atomic unit which it own internal concept descriptions, communication requirements and workflow. This atomic unit can be adapted in the same way as any other element of course. With version 2 of the ACCT, the flexibility offered by the learning activity model will be realized with the provision for a learning activity builder supporting the course developer to fully customize the provided learning activities and create new activities.
The Narrative Structures allow the course developer to build a non-adaptive narrative model based on sound pedagogical strategies. To make the narrative model adaptive the course developer must select Narrative Attributes from the available palette as illustrated in figure 9. The course developer will associate the Narrative Attribute with the Narrative Concept to which they want the adaptivity to be applied. Narrative Attributes are defined to facilitate adaptivity on axes such as prior knowledge and learning objectives, learning context, preferred learning modalities and delivery device.
The course developer is supported during this process through guideline information and sample implementation domains. The course developer can view examples and best practice information based on the current selected Narrative Attribute. Building a custom Narrative using the ACCT The ACCT has a plug-in service that allows the course developer to search across multiple remote learning resource repositories to identify and select appropriate learning resources based on keywords and prior usage information. As shown in figure 9, the ACCT actively promotes the reuse of learning resources by empowering the course developer to select learning resources from a shared repository.
The course developer can then associate learning resources with the concepts of their narrative model. Multiple resources can be associated with multiple concepts. It is the role of the candidate selector to choose the appropriate candidates during the execution of the customized Narrative Model. Note that the learning resources do not necessarily have to exist. One of the features of the ACCT is to act as a content specification tool whereby the course developer can describe the concepts of the course and their context in a content-independent way.
This implies that the content need not exist during the building of the ACCT courses. Course Verification One of the key challenges of authoring adaptive and non-adaptive courses is the ability to test the output of the course. The ACCT offers the course developer a mechanism to test, evaluate and re-develop their course through a multi-Model Metadata-driven Adaptive Engine service that can interact with and interpret the course and models produced by the ACCT. The ACCT allows the course developer to publish their course in the form of a content package. The content package is then used during the runtime execution and reconciliation of the course allowing the course developer to test the pedagogical coherence of their adaptive course.
The evaluation process included pedagogical and instructional design experts from the Centre for Learning Technologies at Trinity College Dublin and technology experts from the Knowledge and Data Engineering Group at Trinity College Dublin. In a workshop-based test environment, a demo of how to use the ACCT was given and a detailed explanation of the models involved in the adaptive course construction process was provided. The workshop attendees were provided with a customizable sample Concept Space, providing the subject area, in which to develop their short adaptive course.
The course developers felt empowered by the ability to efficiently create, test and deploy their short adaptive courses with the ACCT. The course developers were extremely satisfied and comfortable with making a non- adaptive course adaptive using the supplied palette of Narrative Attributes. They felt that the provided Narrative Structures modelled pedagogy formed a solid basis to build pedagogically sound course offerings. The ability to rapidly search for and select learning resources from multiple remote repositories promoted the reuse of the learning resources.
The courses produced by the ACCT proved as technically effective as the existing hand-developed courses. The main noticeable difference was the course development timeline. Related Works In order to evaluate this research, a brief review of the state of the art is presented here which illustrates the similarities and the differences between the ACCT and the reviewed systems.
Current Adaptive Hypermedia AH systems and authoring tools for AH, in the educational domain, concentrate on developing and providing adaptive content retrieval and display capabilities. For educationally effective adaptive eLearning however, the pedagogy must be the focus of development. Once the pedagogy has been customized i. Currently, there are a range of tools available to create online pedagogy. It allows the teacher to divide the course into sections and describe the content that the course will use. REDEEM has been quite successful in construction courses however it supports no elements of adaptivity and dynamic personalization.
However, LAMS likewise provides no support for adaptivity of pedagogical structure and content selection. Adaptive Hypermedia authoring tools are a novel research area specifically in the domain of adaptive educational systems. Similarities that exist between ACCT and LAOS are the domain model knowledge domain representation , and the adaptation model both use hierarchical relationships between adaptive axes and adaptive techniques.
The ACCT differs though by explicitly making the pedagogical model Narrative the primary focus of the course development process. Certain Pedagogical elements may be implemented in LAOS through the goals and constraints model, although they would be more focused on curriculum or course scoping. Due to the complex and dynamic process of authoring Adaptive Hypermedia, the need for author support in creating adaptive pedagogically sound personalized eLearning is evident [De Bra et. The large initial setup cost of adaptive hypermedia is too high for the mass adoption of AHS in education.
From current work in adaptive hypermedia [Aroyo et. Pedagogy can be supported by specifying a requirements-based framework in which pedagogy can be described, used, reused and distributed in an effort to actively promote the cost reduction of adaptive course creation. The course developer can be supported by offering structural support and guideline support during the process of creating adaptive and non- adaptive courses. Based on the state of the art in adaptive hypermedia and online pedagogy authoring, the ACCT will support and provide innovative ways of applying adaptivity to pedagogy to produce personalized eLearning.
Conclusion The main goals of the research were three fold. Firstly we aimed at reducing the complexities of composing an adaptive course, i. Secondly we tried to increase the efficiency of the course composition process, both in terms of the time and effort taken to compose an adaptive course and also the time taken to actually understand how to compose an adaptive course, i. Thirdly we aimed at reducing the costs associated with composing an adaptive course i. These goals are being addressed in several different ways.
Initial indications illustrate that steps towards the realization of these goals have been successful. The complexities of composing an adaptive course have potentially been reduced by facilitating the specification and representation of different compositional models such as instructional strategies, adaptivity, learning activities and subject matter representations.
Through the ACCT these models can now easily be created, used, reused, shared and stored. The Subject Matter Concept Space builder greatly reduces the complexities of creating a domain ontology. The Custom Narrative Builder significantly reduces the complexities of creating adaptive course narratives by providing a palette of modelled components to use during the composition process. Through the course verification service, the complexities associated with testing course semantics have been greatly reduced. This reduction in complexity, inevitably leads to a decrease in the cost associated with composing an adaptive course.
However, the ability to produce more efficient and more effective adaptive learning experiences using the ACCT has not yet been evaluated. To identify the potential benefits of this research we have established a programme of trials over a two year period. Firstly, there was a small scale trial consisting of subject matter experts and instructional design experts from the Centre for Learning Technologies and the Knowledge and Data Engineering Group at Trinity College Dublin.
The primary focus of this trial was the usability of the tool. The Second phase of evaluation is due to start in the middle of November at Intel Ireland. Again this trial will focus primarily on the usability of the tool and the understanding of the adaptive course compositional process and models involved.
The third phase of evaluation will involve the development of short adaptive courses that will be tested by select student groups. This phase of evaluation will focus on the effectiveness of the learning experiences produced with the ACCT measured against a control set of hand-written adaptive courses. Resulting from the initial trial phases, several key updates were made to both the functionality and the feature set of the ACCT.
During the next phase of research and development several different aspects of adaptive course composition will be addressed. This will be achieved by interpreting an adaptive course narrative as a content package with simple sequencing. Currently, our research into the development of novel composition environments is looking at modularized composition components, for example, SVG-based composition environments.
A view of taking the ACCT to the open source community is currently being researched, allowing potential course developers to customize and personalized their adaptive course design environment while improving and extending core functionality. From this research, we have created an environment where educators can adopt personalized eLearning systems as an educational tool. This has enabled a totally different type of course developer, one that does not need to be a technology expert or an instructional design expert.
References Ainsworth, S. Apted, T. Automatic Construction of Learning Ontologies. Aroyo, L. Bajraktarevic, N. Incorporating learning styles in hypermedia environment: Empirical evaluation. Adaptive hypermedia. Adaptive educational systems on the worldwide web: A review of available technologies.
Clarke, L. Conklin, J. Hypertext: An Introduction and Survey. IEEE Computer, 20 9 , Conlan, O. McNaught Eds. Dagger, D. Dalziel, J. De Bra, P. Eklund, J. Universe Science News, 12, 8- Frankola, K. Why online learners dropout. Workforce, 10, Gilbert, J. Grunst, G. Adaptive hypermedia for support systems. Schneider-Hufschmidt, T.
Malinowski Eds. Hockemeyer, C. Jonassen, D. Designing Constructivist Learning Environments. Reigeluth Ed. Kaplan, C.
Adaptive hypertext navigation based on user goals and context. User Modelling and User-Adapted Interaction, 3 3 , Kayama, M. A mechanism for knowledge-navigation in hyperspace with neural networks to support exploring activities. Laurillard, D. Meister, J. Milosavljevic, M. Augmenting the user's knowledge via comparison. Jameson, C. Tasso Eds. HyperTutor: From hypermedia to intelligent adaptive hypermedia.
Reigeluth, C. Specht, M. The New Review of Hypermedia and Multimedia, 4, Vassileva, J. A task-centered approach for user modeling in a hypermedia office documentation system. User Modelling and User-Adapted Interaction, 6 , They also offer adaptive sequencing navigation over the learning content based on rules that stem from the user model requirements and the instructional strategies. EAHA are gaining the focus of the research community as a means of alleviating a number of user problems related to hypermedia. However, the difficulty and complexity of developing such applications and systems have been identified as possible reasons for the low diffusion of Adaptive Hypermedia in web-based education.
Experience from traditional Software Engineering as well as Hypermedia Engineering suggests that a model-driven design approach is appropriate for developing applications where such requirements and constraints occur. This paper presents on a model-driven design process of EAHA. This process accords to the principles of hypermedia engineering and its innovation is the use of a formally specified object oriented design model. Introduction An Educational Adaptive Hypermedia Application EAHA is a dynamic web-based application, which provides a tailored learning environment to its users, by adapting both the presentation and the navigation through the learning content.
Such an application is comprised of learning resources that have specific learning objectives and they are interrelated in order to facilitate the learning process. The learning resources are designed based on pedagogical rules or teaching rules that combine the domain model of the content with the user model and the instructional strategies. EAHAs are currently a research topic of particular interest in the broader field of adaptive hypermedia applications and several EAHA systems have been built during the past years De Bra et al.
The design and implementation of EAHA are complex, if not overwhelming, tasks. This is due to the fact that it involves people from diverse backgrounds, such as software developers, web application experts, content developers, domain experts, instructional designers, user modeling experts and pedagogues, to name just a few. Moreover, these systems have presentational, behavioral, pedagogical and architectural aspects that need to be taken into account. Therefore, systematic and disciplined approaches must be devised in order to overcome the complexity and assortment of EAHA and achieve overall product quality within specific time and budget limits.
One such approach is the use of a systematic design method to support the whole design process. Two candidate approaches exist in this direction, software engineering and hypermedia engineering design methods. Software Engineering methods fail to deal with the particular requirements of hypermedia applications, their user interface intensive nature and their complex node-and-link structure.
Although the discipline of Hypermedia Engineering Garzotto et al. The authors and the forum jointly retain the 26 copyright of the articles. Since educational applications deal with learning, the specification of such applications is a planned set of carefully designed activities and tasks, assessment procedures, selection of proper resources that will support these activities and procedures, that is, the outcome of instructional analysis.
Existing hypermedia engineering methods do not provide adequate constructs for capturing this specification since they pertain to applications that provide views over highly structured data and front ends for transactions resulting to consistent changes of these data. Thus, modeling elements suggested by these methods for capturing the domain concepts are abstractions and representations this kind of data usually stored in databases.
Creating SCORM Content
This is the reason why these modeling primitives typically derive from entity-relationship and object-oriented modeling approaches. However, these primitives do not integrate well with the semi-structured nature of the instruction solutions met in educational applications. Moreover, this pure integration makes it difficult to assign to modeling structures of existing methods proper semantics, which is crucial if it is desired for models to drive the generation of actual applications. Furthermore, although navigation structures proposed by existing methods are generic enough to be used in the domain of educational applications, a more specialized approach would be more eloquent and efficient to use.
This is the main reason why a new approach, specialized in the educational domain is needed. The steps and the outcomes of each step of the design method are presented next. In the following section an approach for the automatic generation of educational applications is discussed. A review of the related literature is presented afterwards and the paper ends with some concluding remarks.
These phases are requirements capturing, design, implementation and evaluation. Furthermore, the product model can form the basis for the description of existing applications, provide the blueprints that depict knowledge and common understanding for particular applications, either completed or under development, much in the way that the blueprints of a building can both drive its development and depict its form, structure and function. This profile is specified by the extension of basic UML elements, the definition of additional semantics for the new elements and as well as the definition of syntactic constraints for the interconnection of these elements, beyond these defined in the specification of the UML itself.
The design model can be decomposed into three sub-models: conceptual, navigational and user interface presentation models. The Conceptual Model The Conceptual Model defines the learning activities that students will be engaged in during the instructional process of a specific subject, together with the semantic interrelationships between these activities.
The learning activities are applied to the various thematic concepts-topics of the domain. The thematic topics should be considered as the Ontology of the subject domain to be learned by the students. The Conceptual Model provides a didactic solution over the objective definition of the knowledge subject. This definition is provided by the authors of the educational application who are considered as subject matter experts.
For each set of topics a set of learning objectives is defined. Extract of the Conceptual Model of a course on Digital Signal Processing Each learning activity is related to particular learning objectives, notions and terms to be taught, etc, according to the syllabus. The activities are hierarchically structured, since composite activities can encapsulate simpler ones.
Apart from their hierarchical organization, activities can be associated with each other with specific interrelationships thus forming a semantic network that provides an abstract representation of the solution of the problem of instruction of a specific topic. This particular view can be reused per se, thus promoting the reusability of educational applications at an abstract level, apart from navigation and presentation issues. In this way, the proposed method incorporates the principle of separation of concerns and promotes reusability.
The activities are associated with specific learning resources. The resources align with the notion of Learning Object. These resources are physical, reusable, binary entities, either static fragments of digital content, e. An example of a Conceptual model is illustrated in Figure 1, which concerns a hypermedia course on digital signal processing.
The hierarchical structure of the Conceptual Model diagrams defines an implicit ordering of activities. The children of an activity are visited after their parent, from left to right. The elements of this sub-model are expressed as stereotyped UML classes and they are actually attribute-value pairs connected with proper association relationships.
The concepts are mapped to specific learning resources. This is the top-level element in the hierarchy of activities that compose the conceptual view of the application. This defines a simple activity which is an atomic one. This activity may contain specific attributes. Predefined attributes are the title and the type of the activity information, assessment, etc. This element defines a composite activity, which contains others, either atomic or concept, thus forming a hierarchy of activities into the educational application. This refers to the association between two activities, atomic or composite.
This element defines the resources associated to specific activities. These two sub-models are presented in the following subsections. The Navigation Structure Model This model defines the structure of the EAHA and defines the actual web pages and the resources contained into these pages. An example of this model is shown in Figure 2. The chapters and subtopics in which an electronic tutorial or book are organized are examples of composite entities.
An access structure contains structural links, that facilitate the navigation into the hypertext space of the application. Content, Composite and Nodes are associated with Concept elements, or directly with Resources, in the Conceptual Model. Fragments correspond to Resource elements in the Conceptual Model. Note that these links are associative links Garzotto et al. They are not structural links denoting, for example, the transition from a page in the learning content to the next one.
The Navigation Structure Model defines, also an implied sequence of nodes. Again, this model has a hierarchical structure. However, this default behavior can be altered by alternative sequencing which is specified in the Navigation Behavior Model, described later. An extract of the Navigational Structure Model Figure 2 illustrates an extract of a Navigation Structure Model, though not all elements of this model are used in it. In this figure the hierarchy of composite and simple nodes are displayed. As shown a node is associated to one or more learning activities from the Conceptual Model.
For example, a designer could use a hypermedia node to incorporate both the presentation of a theoretical part of the subject domain along with an assessment task. Another designer might want to separate those two learning activities. This fact allows the reusability of design models and the separation of concerns.
This behavior overrides the default run-time behavior, implicit in the previous Navigation Structure Model, which is elicited by the structuring of the hypermedia nodes. Earlier research attempts, such as De Oliveira et al. The Navigation Behavior model uses statecharts, as they are incorporated in the UML in order to specify the dynamic transitions of the hypertext structures as the user interacts with the EAHA.
Every containing element of the Navigation Structure Model Content, or Composite is associated to a composite state in the Navigation Behavior Model, while every Node corresponds to a simple state. Thus, the hierarchy of the navigational elements defined in the Navigation Structure Model corresponds to the hierarchy of nested states in the Navigation Behavior Model. The events that fire the transitions in the Navigation Behavior Model correspond to structure links into the Nodes: next, previous, up level, etc.
In addition, guard conditions in these transitions can define alternative navigational transitions, which correspond to conditional behavior of the EAHA, thus implementing content sequencing and adaptive navigation. An example of such a design model is shown in Figure 3. Such rules can be defined by the designer implementing specific Instructional Design strategies.
In particular, each Node in the Navigation Model and its resources are associated with a presentation model element. Note that a multitude of navigation elements can be associated with the same presentation specification, thus promoting uniformity and ease of maintenance of the user interface. Example of the basic templates that have been used in the DSP courseware Elements of the User Interface model are associated to particular nodes of the Navigation Model thus assigning specific presentation attributes to these nodes, as well to their children in the navigation structure hierarchy.
This means that the formatting defined in this template is applied to all pages of the DSP tutorial. XML parsers. This process is illustrated in Figure 5. This is a UML activity diagram, where the various artifacts are represented as rectangles objects and the activities or software tools in this case, are represented as activities rounded rectangles. The produced web pages are a composition of the provided resources wrapped with the proper XHTML code.
These pages are accompanied by a description of their structure in the form of a XML manifest file. In Figure 6 a part of the manifest file for the DSP courseware is depicted. An extract of a Content Packaging Manifest file This XML manifest file accompanied by the learning resources can be deployed to any Learning Content Management Systems that support the IMS Content Packaging specification in order for the educational application to be delivered to its users.
A screenshot of this course is depicted in Figure 7. At the moment, the CGA tool cannot make use of the information about the dynamic navigational behaviour, i. The designer creates distinct design models per user type users with different stereotype. Thus, we create educational hypermedia applications that provides a variety of personalised views of the domain per user type focusing on composition and structural relationships between the learning activities and the respective nodes. For each view, the designer can associate templates in order to specify the look and feel of the nodes.
In this way the generated applications are not adaptive but, in a way, adaptable. A screenshot of the DSP tutorial Related Work This work aspires the bridging of the gap between the conceptual description and the implementation of educational web applications as it is also suggested in Aroyo et al.
The depiction of the instructional design solution as a set of interrelated learning activities and associated resources is also found in the ACCT toolkit methodology Dagger et al. This scheme of activities and resources is also Like other approaches for the design of generic web applications such as WebML Ceri et al. The use of XMI and the focus of the CADMOS-D model on the specific domain of education, which sets certain constraints in the structure of applications makes it different from the aforementioned methods.
It incorporates the principle of separation of concerns in the design of hypermedia applications, dividing the design of the application in three stages: conceptual, navigational and presentational. We also claim that this separation of concerns aligns with the three types of adaptation, navigation and presentation.
Beyond the design model, the development of open, portable, and maintainable EAHA can be facilitated with the adoption of learning technology standards. The lack of dynamic navigational structure is a limitation of our approach, but not an unsolved problem. This schema defines rules for personalised sequencing of navigation into the educational content.
Acknowledgements Many acknowledgements to Mrs. Siskos who produced the learning resources for the DSP courses. References Aroyo, L. Ontological Support for Web Courseware Authoring. In Ceri, S. Booch, G. In Westbomke, J. Methods and techniques of adaptive hypermedia. User Modeling and User Adapted Interaction, 6, Adaptive and Intelligent Technologies for Web-based Education.
Kunstliche Intelligenz, 4, From adaptive hypermedia to the adaptive web. Communications of the ACM, 45 5 , 30— Ceri, S. De Oliveira, M. Statechart-Based Model for Hypermedia Applications. Dolog, P. In Stevens, P. Fischer, S. Garzotto, F. Gomez, J. Conceptual modeling of device-independent Web applications. IEEE Multimedia, 8 2 , 26— Hennicker, R. In Siau, K. IMS a. IMS b. Lowe, D. OMG UML Version 1. Rossi, G. Schwabe, D. Communications of the ACM, 38 8 , SCORM W3C Buffi 13, Lugano, Switzerland jacopo.
Among the reasons, there is the difficult task of designing and authoring an interactive adaptive course, especially for non-technical group of educators. The authoring tool is tailored to the community of non- technical instructors. It is pedagogic-neutral, allowing to define several different instructional strategies. The basic set of adaptive techniques available is complete enough to support several different application scenarios.
Keywords Educational adaptive hypermedia systems, Authoring tools, Visual programming, Web-based education, Macromedia Flash Introduction Adaptive technologies in the field of education have proven so far their effectiveness only in small lab experiments, thus they are still waiting for being presented to the large community of educators.
First of all, as pointed out by some recent studies Brusilovsky, , educational adaptive hypermedia systems EHAS are difficult to design, set-up, and implement, due to the high technical competencies they require to master them. In particular, all of the few existing general purpose educational adaptive systems have a steep learning curve, that forbids a non technical teacher to autonomously create a course.go to site
Homepage - Adapt Learning
More generally speaking, the main issues that hinder the spread of many of the available adaptive systems in community of educators are: 1. High technical competencies to set up an adaptive course i. Difficulty in specifying in the system language the interactions that must occur between the user and the system i. Lack of ready-to-use patterns that exploit frequent adaptive teaching strategies. Despite this situation, different researchers point out the importance of adaptivity in the definition of effective learning scenarios. The Authoring Process of EAHS Since an Educational Adaptive Hypermedia System is a hypermedia system, firstly we can refer to the well developed corpus of literature on Hypermedia Design to extract general features of the design process of hypermedia applications.
In this field several conceptual models and methodologies have been developed. Just to mention a few: RMM Isakowitz et al. Although adopting different and incompatible name conventions, all of them agree on the fact that essentially a hypermedia application is composed of Coda et al. The authors and the forum jointly retain the 36 copyright of the articles. To the proposed list we can add a fifth element that is context, and we obtain a generic adaptive hypermedia model as the AHAM model, cf. De Bra et al. Therefore each element of this list must be taken into account during the design process of virtually any AHS.
Interestingly the researchers also sketched a list of authoring tasks for any AHS: 1. Write concepts and concept hierarchy 2. Define concept attributes define main and extra attributes 3. Fill concept attributes write contents 4. Add UM related features 6. Decide among adaptation strategies, write in adaptation language medium-level adaptation rules or give the complete set of low level rules such as condition-action or IF-THEN rules. Define format presentation means-related; define chapters 8.
Add adaptive features regarding presentation means define variable page lengths, variables for figure display, formats, synchronizations points, etc. The main benefit of this model is its generality, because it fits well with virtually any adaptive hypermedia system. Unfortunately it is too complex for a non-expert of the field, and it requires a deep understanding of each layer and its sub-layers as well to master it.
Besides, it is too heavy for small educational contexts, thus scalability is a problem. A Teaching Task can be atomic or composed, the latter being composed by other TTs. This way a simple hierarchy of TTs can be defined. Composed TTs can be presented with two sequencing strategies: a sequenced and or an alternative or strategy. The former imposes to complete all of the sub-tasks to complete the whole task, while the latter allows completing any subtask. This authoring model is very simple and straightforward. E2ML, in Botturi , as the closest example. On the other hand this system does not allow to creatively extend the adaptive behavior over the given few sequencing methods.
Moreover its application is limited to the education field only. Some insights on the theoretical authoring process of EAHS come from a recent study about authoring learning styles in adaptive hypermedia Stash et al. In this research the authors experimented with authoring adaptive courseware using two different perspectives: an adaptive engine pull, and an authoring push.
The former method deals with manually writing low-level adaptation rules to mimic the intended interaction with the user in order to achieve a specific instructional goal, namely in the example a learning style matching goal. The latter relates to choosing an instructional strategy among a dictionary of predefined strategies and letting the system define and perform automatically the necessary interactions with the user. The authors also report the importance of the granularity of the design method: some approaches work at the instance level of the course namely single pages, and rules , whilst others are related to the schema level general concepts.
Both of them seem promising, yet from our experience it seems that non-technical people seem more keen reasoning in terms of instances thus locally , than in general terms or at the schema level. Each stage involves several tasks as shown in Table 1. With respect to the list of authoring tasks presented by Cristea et al. Yet, both of them strive for the diffusion of a highly structured design method which contrasts with the expected end-users non technical instructors that should benefit from them.
Therefore the best solution seems to be: pruning out some elements from these heavy methods, leaving only the most basic and relevant steps that may be helpful in practice for the large community of educators. Design and authoring steps in the process of creating adaptive hypermedia systems from Brusilovsky Stage Task Design Design and structure the knowledge space Design a generic user model Design a set of learning goals Design and structure the hyperspace of educational material Design and structure the hyperspace of educational material Design connections between the knowledge space and the hyperspace of educational material Authoring Create page content Define links between pages Create some description of each knowledge element Define links between knowledge elements Define links between knowledge elements and pages with educational material From our experience with the design and authoring of adaptive courses, we came across with similar results.
MAID method concerns with the following steps: 1. Interaction Model: define the application behavior; 2. Domain Model: define the content structure; 3. User Model: define the relevant information about the user; 4. Interface Model: map the behavior into user interface elements; 5. With respect to the former methods, MAID takes into consideration all of the aspects of the design process ranging from requirement analysis to implementation and testing.
Moreover it is a practical method that can be taught in a whole day session. Unfortunately, although the method is quite simple to learn, the tasks involved are still complex and often too structured to be executed properly by a non-expert without assistance from technical experts. Therefore an interesting question is whether we can prune some parts of these methods to reduce the most complex design and authoring tasks, without loosing too much flexibility.
Since the degree of maturity of the research on general principles for the design and authoring of EAHS is still low, we can count only a bunch of proposals for authoring tools. In the last decade, some domain-independent Educational Adaptive Hypermedia Systems EAHS , with different degrees of authoring capabilities were released, some important examples being: Interbook Brusiolvsky et al. Yet, these tools have been used only within the research group who developed them with very rare exceptions, for example the Author of this paper experimented with AHA! More recently some research projects have directly addressed the authoring problem, by developing tools to support the process.
Among the most relevant examples we can list: Schoolbook Kupka et al.