EML - Educational Environment Modeling Language

EML is a modeling language spe cifically developed for the design of educational environments. It models the goals, requirement and design of the teaching and learning activity. Its core parts are a verbal statement and visual mapping of goals and a visual UML-like representation of the learning activity. EML can be profitably integrated with Instructional Design models in order to enhance the communication within design team and the elicitation of requirements, and to provide a successful integration of new media in the learning experience through a well-shaped and documented design process. Introduction: New Challenges in the Design of Education E-learning has brought a cultural shift in the way we conceive teaching and learning, and in the everyday practice of instructors. New media and networking have made available several tools for education, and these in turn make possible a number of new educational strategies based on distributed team working, on asynchronous online communication, etc. Instructional Design (ID) is the discipline concerned with research and theory about instructional strategies and the process for developing and implementing those strategies (Smith & Ragan 1993). ID models are tools that can be used for structuring a correct design process and for checking the quality of the final product. During the last years, they have been revised and updated in order to fit to e-learning design, but still the design of education remains a hard task in any context. In this paper we present EML, a semi-formal representation language for educational environments that can be profitably exploited and integrated with other models in instructional design. The first paragraph will provide a short state of the art in ID, while the following ones will introduce EML with definitions and examples. Finally, conclusions and outlooks are reported. Instructional Design: Where Are We? Gustafson and Branch have categorized ID models in three main categories (Gustafson & Branch 1991, Gustafson & Branch 1997). 1. Classroom-oriented models represent the greatest part of ID models. They are used for designing one or a few hours of instruction, mainly in the traditional classroom setting. They assume that one instructor is at work, the exemplary situation being that of a schoolteacher preparing a lesso n or an activity. These models serve as checklist for designers, guiding them from the requirement analysis to the final evaluation of results before redesign, like a specialized project management red-thread. The basic model is ADDIE (Analyze, Design, Develop, Implement, Evaluate). Other models are Gagne’s Nine Events of Instruction (Gagné 1987; Gagné et Al. 1992) and ASSURE (Heinrich et Al. 1983). 2. Product-oriented models are aimed at designing effective learning materials. These models have found new relevance and new challenges with the introduction of new technologies, of multimedia and distance communication means. Like other hypermedia design models, they are concerned with a team at work. One of the most interesting models is CADMOS -D, a specialized UML notation for educational applications (Retalis et Al. 2002, Psaromiligkos & Retalis 2002) 3. System-oriented models are developed for a team designing a great amount of instructional units, such as a course or a curriculum. They are usually more technical than classroom-oriented models and also guide the design team through a well-shaped process to the achievement of the expected result. The main referenced models are by Dick and Carey (Dick & Carey 1996) and by Ragan and Smith (Smith & Ragan 1993). On the other side, several organizations in the technical world such as IEEE, ADL and IMS, are investing much in the development of metadata standards for Learning Objects (see e.g. LTSC LOM) and for the description of whole learning environments (noticeably IMS Learning Design, see IMS 2003a and IMS 2003b). Although these developments do not belong to the field of ID, their advancement will influence it in a probably near future. What is E2ML? EML – Educational Environment Modeling Language is proposed as a blueprint for educational environment modeling. As conceptual framework, EML is composed by an explicit definition of the learning process and of the educational activities; as a formal modeling language, it defines UML-like representation syntax. Differently from the greatest part of ID models, EML was developed for representing the educational activity being designed, and not the process of design itself or the support materials developed. This makes it profitably integrable with other models. When is E ML useful? Instructional design models are structured design methods, and as such require an overhead cost. A structured design is cost-effective only when the complexity of the designed object is high. EML was developed for supporting educational design in technology-rich contexts, and it is useful when an interdisciplinary team is at work on a course that exploits different media and different kinds of educational activities and settings. In particular, it addresses the following issues: 1. The subject of e-learning design is often an interdisciplinary team, which may include subject experts, media designers, technicians, tutors, psychologists, etc. External partners too may have a role in delivery or supply (see e.g. Pigni 2002, Ardizzone & Oliveto 2002). EML can ease and enhance communication in the design team and with external partners. Moreover, an E2ML representation of the educational activity could be the basis for a project management approach to course development, structured according to other classroomoriented or system-oriented ID models. 2. The definition of requirements for the tools to be used in a specific e-learning environment is a delicate issue, as it often concerns great investments, both in the case of off-the-shelf solutions and in that of custom application development. EML is a tool for systematically defining and expressing the educational requirements of software applications, thus providing a first input to product-oriented ID models. 3. Setting up an e-learning environment is great economic effort (developing content, digitizing, buying and customizing applications, training tutors, etc.), and should be balanced by an adequate return on investment. EML can support static quality assessment through the expression of formal features. This can help controlling the quality of the learning experience at design time and distinguishing quality-critical applications or content from optional nice-to-have features and optimizing investments. 4. Given the complexity of e-learning environments and the uniqueness of each class and of each learner, unexpected learning outcomes may rise. EML modeling may be used as a diagnostic tool for identifying relevant issues and for figuring out viable redesign solutions. The EML modeling of a learning environme nt requires four main activities: the expression of goals (goal statement and goal mapping); the compilation of resource lists (roles & actors, locations, tools); the definition of action diagrams and finally the creation of overview diagrams (course breakdown statement – CBS, dependencies, timeline). These activities are described in the following paragraphs in their standard form. Nevertheless, they can be adapted (simplified or detailed) to the needs of any specific educational context or design team. Moreover, their presentation in this paper does not coincide with their actual performance in a real design process – experience proved that most of the time the activities run in parallel and interact with each other.