GENERATYVINIŲ MOKYMOSI OBJEKTŲ GRANULIACIJOS ANALIZĖ PAKARTOTINIAM PANAUDOJIMUI PAREMTI
ANALYSIS OF GRANULARITY WITHIN GENERATIVE LEARNING OBJECTS TO SUPPORT REUSABILITY

Summary/Abstract: Reusability of the learning objects (LOs) and its usefulness is the main engine for the LOs theory development in the e-learning domain. We discuss the granularity problems which are directly related with LO reusability in the context of using generative learning objects (GLOs). The main problem is to select an appropriate LO granularity degree without loosing the possibilities of LOs reusability. Naturally, the smaller the LO and the more unrelated it is to a specific context the more reusable it is in various contexts. However, decomposing LO to separate elements and eliminating the context we loose the LO pedagogic value, i.e. such LOs become non-efficient with respect to the learner. Therefore, to solve the problem, we suggest creating GLOs combining two technological paradigms: feature diagrams (FDs) for GLO specification in the early development phase and generative techniques for GLO implementation. The chosen paradigms ensure the evaluation of the possibilities of reusability in the early GLO development phase (it means, we have a possibility to evaluate the efficiency of the composed GLO) and the generative reuse. GLO enables the user to select the requested parameters from the meta-interface on which a specific LO item, meeting the requirements of the user, is generated. By specifying with GLO feature diagrams we may foresee possible GLO contexts, filling GLOs with versatile information of a particular field. Thus the GLO becomes pedagogically efficient, adaptive and its granularity degree does not restrict the possibilities of reusability. Therefore, we may draw a conclusion that composing a GLO we automatically partially solve the existing problem as the GLO granularity does not restrict the possibilities of reusability. However, we must know how to evaluate the GLO granularity degree because while describing the GLO by means of meta-data it is necessary to indicate the aggregation level. To solve the problem we suggest using a formal graph-based model which not only enables the creation of learning scenarios from the generated LO items but also allows the evaluation of the granularity degree.

• Details
• Contents