The nature of mobile devices with their small screens and limited input capabilities can not replace yet the standard desktop computers or laptops. However, the same properties can make them efficient in learning domain, if certain constraints are kept like Short modules and Context-Based content. One of the primary assessment goals is to provide feedback on whether the course and its learning objectives have been satisfied; this feedback is used to improve aspects of the educational process. So, it is important that the assessment data are accurate. In line with many researchers we distinguish three types of assessment: a) Diagnostic Assessment b) Formative Assessment, and c) Summative Assessment. We are developing an Intelligent Mobile Learning System which uses web technologies and specifications and which learns by experience, changes its knowledge base on the basis of students’ feedback and follows students’ progress supporting diagnostic, formative and summative assessment; Questions are exchanged by means of an XSL engine and they are directly transformed in html tests and java objects. System’s knowledge base contains: • Course structures expressed as related topics. • History data concerning students’ assessment (topic-based or not). • Mechanism for locating distributed Learning Objects (LO). The architecture allows • Insertion or deletion of new Learning Objects without changing a course structure. • Creation of new courses by reusing Learning Objects without any effect on history data or other learning objects. A history data module is updated by the data received from the assessment results. Useful conclusions can be made based on the history data concerning a student’s progress, effort, motivation, instructor’s efficiency, and course modification. The history data are used to formulate a customised assessment session. First year students with empty history base participate in a diagnostic assessment in order to enrich the history base and define an approximate student level. The system records all significant actions of a student (lessons chosen, number of self tests taken, test difficulties, results taken, test time, progress level). When a student finishes a lesson, besides the indirect assessment data, the system conducts a direct assessment session in order to confirm that the material has been actually understood. The test corrector module takes answers as input and produces a score and feedback. The corrector knows the correct solutions and records a student’s performance. The total score evaluated is normalized in order to obtain a human-understandable mark. The test corrector module can be either on the server (all the answers are posted to the server and the correction is performed by the server) or on the client (answers posted to the server but correction is performed by the client). An Intelligent Mobile Learning System, like the one we briefly described, among other changes, adds new roles to an educational environment. Such a system introduces new ways of exercising teaching and assessment and changes fundamentally the role of teachers and students make them both producers and consumers of Learning Objects. An educational and assessment material becomes a set of Learning Objects automatically adapted to meet the new requirements. In such an environment a consumer either student or teacher expresses a particular behaviour called Development With Reuse while a producer develops Learning Objects suitable for reuse expressing a behaviour called Development For Reuse. In such a world with no contractual and legal issues between producers and consumers the equilibrium state is a level where demand is satisfied.

An Intelligent Mobile Learning System, like the one we briefly described, among other changes, adds new roles to an educational environment. Such a system introduces new ways of exercising teaching and assessment and changes fundamentally the role of teachers and students make them both producers and consumers of Learning Objects. An educational and assessment material becomes a set of Learning Objects automatically adapted to meet the new requirements. In such an environment a consumer either student or teacher expresses a particular behaviour called Development With Reuse while a producer develops Learning Objects suitable for reuse expressing a behaviour called Development For Reuse. In such a world with no contractual and legal issues between producers and consumers the equilibrium state is a level where demand is satisfied.