A Framework for Developing Adaptive Personalized eLearning
نویسندگان
چکیده
Developing adaptive personalized eLearning has proven to be quite expensive regarding time and cost. Typical adaptive course creation requires input from subject matter experts, pedagogical experts and technical experts. This paper describes a framework for developing adaptive personalized pedagogically sound courses in a developer-supported environment. The paper describes the Adaptive Course Construction Toolkit (ACCT) which was created to provide the course developer with tools to design, test and deploy adaptive personalised eLearning. The ACCT offers support based on pedagogy, instructional design principles, knowledge domain ontology descriptions and learning resource selection. Introduction Current tools are involved with developing multimedia and the construction of online classes, i.e. class and course management. These systems are useful but their effectiveness in a constructivist-oriented learning environment is limited. The ability to create adaptive personalized eLearning in a flexible and dynamic way is needed. From our experience in adaptive personalized eLearning, a number of key principles have been noted. It is critically important that the teacher/tutor be empowered within the learning experience and not disenfranchised. As a blended teaching approach, the adaptive and personalized eLearning does not replace the teacher. It transforms and enriches their role in the learning experience. Adaptive personalized eLearning is a tool which the teacher/tutor can use to increase the potential educational effectiveness of the entire learning experience. Adaptive eLearning is not just about adaptive content retrieval and construction. It is a mixture of pedagogy, domain knowledge (subject matter expertise) and adaptivity. The production of personalized courses is realized at the concept level not content level. Specific content selection is an aspect of learning experience development that should not concern the teacher/tutor. This paper will give a brief overview of the state of art of adaptive hypermedia (AH), AH authoring and online pedagogical development tools. The paper will introduce the Adaptive Course Construction Toolkit (ACCT). It will describe the requirements and specifications identified during the ACCT initial development process. The paper will then describe the models, interface components and services of the ACCT. The paper will then present the evaluation process and results of the initial ACCT trialing period. This paper will conclude with a brief overview of the future work in adaptive personalized eLearning development. 1 State of the Art in Developing Adaptive Hypermedia and Adaptive Personalized eLearning Overview There has been significant work in the past decade on Intelligent Tutoring Systems (ITS). Theses systems employ Artificial Intelligence techniques to aid in the teaching of a domain of information. Early ITS however failed to provide the learner with significant control over their learning. They often, contrary to theory of constructivism, imposed a strict learning path. This restriction in learning path flexibility, due to a “system is always correct” mentality, is not pedagogically viable and can be quite frustrating. In recent years Adaptive Hypermedia Systems (AH) have added some learner characteristics to the system architecture through explicit learner modeling. Through this the AHS can customize certain aspects of the course to the learner’s characteristics. Early AHS offer a level of personalization to online courses in the form of simple sequencing or “Adaptivity Lite”. It allows multiple learning paths to be created, then based on the learner requirements the most “appropriate” learning path is selected. Some later AHS expand this approach by allowing adaptivity to occur on more axes. These axes of adaptivity included techniques such as link hiding, link sorting, link annotation, logical object inclusion/hiding and stretch-text. Adaptive information retrieval systems like HyperDoc, HypAdapter and Adaptive learning systems like InterBook and AHA! have proven successful in the past [Millard et. al.(2003), Hohl et. al. (1996), Eklund, Brusilovsky (1999), De Bra et. al. (2003)]. The deficiencies with some of these educationally-oriented systems originate with their inflexible adaptivity descriptions. Most AHS combine the intelligence for delivering the learning resource adaptively, with the learning resource itself. This semantic inflexibility inhibits the reusability of the constituent learning resource outside the context of the adaptive course in which it currently exists. From a purely educational perspective there are a range of tools that aid in the construction of online “pedagogy”. For example, the REDEEM system [Ainsworth et. al. (1999)] allows the teacher to create pedagogical online courses by describing the structure and flow of the content of the course and also the sequencing of the content. It allows the teacher to divide the course in sections and describe the content that the course will use. From an active learning perspective the LAMS system [Dalziel (2003)], which is built upon the emergent Learning Design standard (Previously Educational Markup Language EML), allows the teacher to create, describe and sequence learning activities. Due to the complex and dynamic process of authoring Adaptive Hypermedia, the need for author support in creating adaptive and non-adaptive pedagogically sound personalized eLearning is evident [De Bra et. al. (2003), Brusilovsky et. al. (2002)]. The reach and effectiveness of adaptive personalized eLearning systems is also limited due to the cost of application development. 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. al. (2003), Kay et. al. (2002)] in personalized eLearning it is evident that there are two areas of research which need future development, the design of pedagogically sound adaptive courses and the support offered to the course developer during the process of developing pedagogically sound adaptive courses. Pedagogy can be supported by specifying a requirementsbased 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. 2 Adaptive Course Construction Toolkit (ACCT) The development of the Adaptive Course Construction Toolkit (ACCT) has resulted from the need for a pedagogical and adaptive course developer support framework. The ACCT is a design-time tool which allows the course developer to create adaptive and non-adaptive activity-oriented course narratives based on sound pedagogical strategies in a developer-supported environment. The ACCT supports the course developer by providing tools to create and describe the knowledge domain that they are trying to teach, to create a customized teaching strategy through this knowledge domain, to define curriculumized scoping of the domain ontology, to search and select learning resources from multiple remote repositories, to build and distribute ACCTbased course packages and to allow for real-time adaptive course verification by rendering through an Adaptive Personalized eLearning Service (APeLS). The architecture of the ACCT is built upon a reusability-focused, developer-supported and serviceoriented 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 ACCT uses an abstraction-based approach, similar to the multi-model metadata-driven approach, in its course construction methodology. Different levels of abstraction can be used during the creation of the adaptive or non-adaptive course to describe the course pedagogy, sections, sub-sections, activities, available tools and concepts. This abstraction mechanism allows the course developer to define their teaching strategies and information 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. Pedagogical and instructional design principles were studied and modeled to form reusable and scaleable design guidelines. These principles were analyzed to produce guidelines for writing narratives supported by the selected principles. 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 ACCT can interact with, interpret and create the MultipleModels that are used by the Architecture for multi-Model Metadata-driven Adaptivity within APeLS. 2.1 Requirements and Specifications During the process of creating adaptive and non-adaptive courses using a support-based framework there are a number of requirements that must be realized. These requirements span teacher/tutor, learner/student, educational and technical needs. For example, the teacher/tutor may be concerned with the representation of the knowledge space, the learner/student may be concerned with their learning goals/objectives/learning modality, from an educational perspective the produced course should be pedagogically sound, and from a technical perspective the course construction and delivery process should be aided by automatic model generation and reconciliation. These example requirements may feed into the development process at different chronological points to produce different affects. The following sections will outline the initial requirements and specifications of developing and constructing adaptive personalized and non-adaptive courses with the Adaptive Course Construction Toolkit (ACCT). 2.1.1 Teacher Requirements Several teacher requirements need to be realized during the creation of the Adaptive Course Construction Toolkit (ACCT) to support and aid the teacher during the process of creating pedagogically sound adaptive and non-adaptive courses. The teacher/tutor must have the ability to graphically represent the domain of knowledge within which the adaptive/non-adaptive course will exist. This graphical representation of the “Concept Space” allows the subject matter expert to define an ontology of subject matter concepts. This ability to create a “space” of concept definitions, relationships and interrelationships provides the subject matter expert with a mechanism to describe the knowledge domain in logical terms. The teacher must have the ability to search for and select appropriate learning resources to populate the adaptive/non-adaptive course. They must have the ability to open and search these repositories based on key semantic and pedagogical terms. Learning resources can be pedagogical/instructional design models, narrative models, adaptive axes, content, etc. The teacher must also have the ability to develop adaptive/non-adaptive course narratives. The course narrative represents the “conceptual flow” of the developed course. This process can exist independent of the content that will be used to render the course. This provides the course developer with the flexibility to abstractly describe and develop adaptive/non-adaptive courses. The teacher should be supported during the Narrative development process with the availability of sample pedagogical models. These sample pedagogical models will offer the course developer support during the creation of pedagogically sound learning experiences. For example, models to represent didactic, web-quest and case-based teaching will be provided. This requirement arises from because most online learning experiences will employ a blend of pedagogical strategies (e.g. certain sections of a didactic course should be taught in a case-based way) to fulfill its goals and objectives. Learner empowerment must coexist with teacher empowerment. The teacher must be supported during the Narrative development process with usability and usage guidelines for the provided pedagogical models. They must have the ability to define semantic conceptual scope within the course. This will provide the course developer with the ability to shrink and grow the course scope with respect to learning sessions. This scope definition mechanism gels with a curriculumized learning environment whereby for example in lecture week 1, I want the student’s course to be scoped to match the goals of lecture week 1. In this way the teacher is empowered to define and mould the learner experience to their teaching goals and objectives. 2.1.2 Technical Requirements Through a study of the process of developing adaptive/non-adaptive courses, several technical requirements of adaptive/nonadaptive course construction must be realized by the ACCT including user interface design, human-computer interaction and model generation. The user interface must intuitively support the teacher/tutor during the course development process. It will support the course development process by providing interface tools to design and develop an adaptive/non-adaptive course. The current model creation process is complex and error prone. A technical requirement of adaptive course construction is the automatic generation and reconciliation of multiple models based on the course developer interaction with the user interface. The ACCT will be responsible for generating, editing, storing and retrieving all models of the course development process. Another technical requirement of the ACCT is the generation of a multi-session workspace and course package generation mechanism. The ACCT must maintain a virtual workspace to store the models of the adaptive/non-adaptive course. This workspace is used during “course-publication” whereby the ACCT can create a course package. The course package will contain all relevant information and models that are required by the adaptive/non-adaptive course. With the growth in popularity of service-oriented architectures, the ACCT is required to provide a service-oriented interface to both repository interaction and realtime course verification. The ACCT will communicate with multiple repositories through a generic and common service interface which will define the functionality that the service can provide during interaction with the repository. An important requirement of the ACCT is the ability to rapidly test/verify the course output based on the custom designed narrative. This realization of this requirement relies on the ability of the ACCT to communicate with an external Adaptive Engine (AE) service. The ACCT produced course must be readable and interpretable by the AE service in order to function appropriately. 2.1.3 Educational Requirements One of key educational goals of the ACCT is to support the course developer during the process of producing pedagogically sound adaptive/non-adaptive courses. The ACCT must have the ability to represent, interact with and interpret teaching strategies based on proven pedagogical and instructional design strategies. It is responsible for providing the course developer with the pedagogical support required to develop educationally effective courses. Through this functionality the ACCT must be able to interact with and interpret custom pedagogical models. The ACCT is required to produce educationally beneficial personalized eLearning experiences. The learner must feel in control of their learning experiences when using courses developed with the ACCT. This identifies a requirement that involves the ability of the developed learning experiences to adapt to the learner through the ability to understand the learner and produce learner model schemas based on the custom course narrative. A key challenge arising from this requirement is the development of test and evaluation processes and metrics for measuring course effectiveness. A key educational requirement of the ACCT is to take a learning activity based approach. Currently most Adaptive Hypermedia Systems (AHS) in the educational domain concentrate on information retrieval and the adaptivity on the content level. The ACCT is required to support the development of adaptively delivered learning activities and services. Learning Activities, although somewhat contradictory of the traditional classroom-based learning environment where the teacher is the sole disseminator of the learning experience, involves the next-level of learner empowerment. This allows the learner to control the pace and focus of their learning while the teacher/tutor is available to provide classroom, group and individualized support. Although the LAMS [Dalziel (2003)] system allows from for the creation and sequencing of learning activities it does not support the creation of adaptive narrative based learning activity delivery. The ACCT is required to have the ability to produce such rich learning experiences. 2.2 Design and Implementation (Adaptive Course Construction Toolkit) Based on the requirements specification process, a number of models and interface aspects were designed and developed. The following sections will describe the models and the interface that was developed based on the previously defined requirements. 2.2.1 Models Deficiencies of current approaches to Adaptive Hypermedia System (AHS) development involve the combining of content and intelligence that describes the adaptive delivery of the content [De Bra et. al. (2003), Brusilovsky et. al. (2002)]. The multi-Model Metadata-driven approach separates the content from the rules that govern its adaptive delivery [Conlan, et. al. (2002)]. This separation of concerns increases the potential for reuse, not only of the content but also the intelligence (pedagogy, ConceptSpace and narrative structure) behind how to adaptively deliver such content. To facilitate flexibility and reusability in the design and implementation of new course offerings the multi-model approach was designed to include an abstraction layer. This layer, extended through the research by the ACCT, enables design-time peercollaboration during the development of an adaptive course. For example, the abstraction enables the course author (knowledge domain expert) to develop a teaching strategy describing the course sequencing not in terms of the learning resources to be added, but in terms of the concepts or learning activities to be learned, i.e. content-independent course development. The models used will extend the standardized models of eLearning, namely IMS LRM, IEEE LOM and ADL SCORM. One of the key attributes of creating educationally and adaptively effective metadata is not to record the similarities of the educational or adaptive values of the learning resource being modeled but the differences. The de facto standards are extended to accommodate educationally and adaptively differentiating descriptions that can be reconciled during the decision making of the narrative execution process. This is achieved through the use of the “adaptivity_type” element. It provides pedagogical and adaptable information, about the learning resource being modeled, which the narrative can interpret at runtime. An example of which can be seen in fig.??? # $ ! % " ! ! & $ ' ( *) ,+ "-. /%/ 0 1 % 2+*3 3! ! 4 + )! 5 6 6 0 ,' 1 '! % /%/ Fig. 1 2.2.1.1 Subject Matter Concept Space (SMCS) One of the key challenges of the adaptive course construction process is to identify and represent the abstract domain of information within which the adaptive course will exist. The SMCS is the embodiment of this scaffolding of the domain knowledge. Through the use of the subject matter concept space the subject matter expert can describe the information domain to be taught as a collection of high level concept descriptions, relationships and interrelationships in a content-independent way. This ability to describe the information domain in a content-independent way provides the course developer with a mechanism to rapidly prototype ACCT-developed adaptive/non-adaptive course structures. 7 8:96; 8 <:=!> ? @ ACB D B >FEG= < ACB 3 'IH& JK L * B M N ? ; O PFQ RL<6S ACB D B T 7 ; U V(F? 96; N U ; X ACB BCT 0 )! 5 6 -. ^ 3! 6 % ) 1 $ " _ Z '! 2[! " & 7 \ @ F? 96; T 7 RL<6N U >F? 96; ]GQ ? = ] T 7 RL<6N U >F? 96; ]GQ ? =5M N ? ; O PFQ RL<6S ACB ` BaT 7 RL<6N U >F? 96; ]GQ ? = b > EG= < M N ? ; O PFQ RL<6S ACB ced fYg g fih,j ZekYl B6\ T 7 RL<6N U >F? 96; ]GQ ? = b 8 U RL@ ? ; U N ? >FE M N ? ; O P Q RL<6S ACB m!n$m B&\!T 7 RL<6N U >F? 96; ]GQ ? = b @ F? 96;WM N ? ; O P Q RL<6S AYB BoT [! 6 6 C *3 p 3 & % ) "3 ( q 6 % ) sr:[ 4 ZYket
منابع مشابه
A conceptual model for developing a personalized adaptive elearning system in a business environment
In this paper we present and recommend a conceptual model for designing and developing an adaptive elearning system that main application is the identification of educational needs of employees from formal and semi-structured data and the choice of the appropriate personalized learning strategy. The human resource department has the responsibility for the all process. We use available internati...
متن کاملAgent-oriented Middleware Supporting Delivery of Mobile Elearning Services
Within project Distributed eLearning Center (DeLC) we are developing a system for distance and eLearning, which offers fixed and mobile access to electronic content and services. Mobile access is based on InfoStation architecture, which provides Bluetooth and WiFi connectivity. On InfoStation network we are developing multi-agent middleware that provides context-aware, adaptive and personalized...
متن کاملCognitive Architecture for Affective eLearning Cognitive Architecture for Affective eLearning
Learning is a psychological experience one receives to make behavioral modifications to improve his/her capabilities while gaining knowledge or skills. Even though the traditional learning is a teachercentered pedagogical instruction, the modern learning framework is centered under a student-centered LMS and thus it is weak in cognitive, emotional, behavioral and social measures. Meanwhile, mor...
متن کاملEvaluation of APeLS - An Adaptive eLearning Service Based on the Multi-model, Metadata-Driven Approach
The evaluation of learner and tutor feedback is essential in the production of high quality personalized eLearning services. There are few evaluations available in the Adaptive Hypermedia domain relative to the amount of research interest this domain is attracting. Many of the papers in this domain focus on the technological design of systems without justifying the designs through the lessons l...
متن کاملEvaluation of Adaptive Course Construction Toolkit (ACCT)
The ability to create pedagogically driven, activity oriented personalized eLearning is a key factor in the success of future eLearning. The ability to empower course developers with intuitive environments which provide pedagogic, subject matter, activity and personalization support during the process of creating a personalized course is fundamental to the adoption of adaptive educational hyper...
متن کاملMulti-model, Metadata Driven Approach to Adaptive Hypermedia Services for Personalized eLearning
One of the major obstacles in developing quality eLearning content is the substantial development costs involved and development time required [12]. Educational providers, such as those in the university sector and corporate learning, are under increasing pressure to enhance the pedagogical quality and technical richness of their course offerings while at the same time achieving improved return...
متن کامل