Automated planning for personalised course composition Antonio Garrido, Eva Onaindia and Oscar Sapena Phone: (+34) 963877007 Fax: (+34) 963877359 Universidad Politécnica de Valencia Camino de Vera s/n, 46022, Valencia (Spain) {agarridot, onaindia, osapena}@dsic.upv.es Abstract Authoring tools for building Intelligent Educational Systems must provide support to ensure flexibility, adaptability of content to the user profile, reusability and sharing of learning objects. These facilities are essential to develop automated decision processes for providing course compositions tailored to the specific characteristics of each individual learner. We present a LOM-compliant learning approach that uses an automated planning process to create personalised learning courses while giving special attention to the development of reusable learning objects. 1. Introduction Authoring tools for building Intelligent Educational & Tutoring Systems must provide support to ensure flexibility, adaptability of content to the user profile, reusability and sharing of e-learning objects (LOs) [5]. These are also key features in the development of automated decision processes for providing course compositions tailored to the specific characteristics, goals and preferences of each individual learner. The selection and composition of LOs require meta- data that denote a level of semantic specification enough to enable consistent runtime automated semantics [6]. This is especially relevant in the case of automated planning processes where the access to specific content of a given course requires that all resources are accurately described by structural relationships which reflect the logical sequence of content [7]. However, since meta-data labelling is usually an arduous task and a not often attained goal, this has led to repositories where the learning content does not meet the necessary requirements to serve as the basis for common automated learning activities. In addition, the roles of relationships are not free of ambiguity, which seriously hampers the possibilities of consistent composition [6]. The efficiency of a planning process relies, among other things, on the accuracy of the LOs relationships. Therefore, a major issue in building a personalised course (plan) is how to provide LOs with appropriate meta-data annotations in order to increase the automation level for the composition of LOs and services (planning). In this paper, we present a LOM-compliant learning approach [4] that uses an automated, adaptive planner to create personalised learning courses that are portable to any IMS-LD run-time environment while also focusing on the modelling of reusable LOs. Our motivation for reusability is to achieve a better identification of LOs when searching for sets of related resources, for which it is crucial to create a shared semantic basis for metadata elements usage. This semantic information can be expressed through structural relationships between LOs to facilitate knowledge sharing and reuse. Our approach follows the IMS specification and facilitates the completion and extension of meta-data records of LOs, specially those ones related to the structural and logical relationships of LOs. Particularly, we present a contribution to: 1. Import LOs from sharable repositories. 2. Improve meta-data labelling by extending the instructional design rationale, thus promoting a personalised access to the LOs. 3. Design our own LOs. 4. Export the improved content into sharable repositories. 5. Create personalised goal-oriented learning routes through an automated planner. 2. General schema The general schema of our approach is depicted in Figure 1. First, a teacher uses our modelling tool to design, from scratch or by reusing LO collections, a pool of LOs. These LOs can be part of one or more courses, defined as instructional designs, based on an explicit, topic-centered structured representation of the learning domain. Note that this information is profile- independent, so it is valid for students with different learning styles. Second, students choose a whole course to follow, or simply the learning outcomes they are interested in from the course, indicating their personal characteristics (profile), background and preferences. The course structure, together with the