Evaluating the utility and usability of an adaptive hypermedia system

Adaptive web‐based learning: accommodating individual differences through system's adaptation

“…Compared with other approaches, the proposed guidelines support designing several levels of adaptation, ranging from full system-control to full student-control, and Changing their learning style on the student the educational material (Hartley et al, 1995) model, students may plan the presentation of the educational material accordingly Students may take full instructional Students should be provided with a view of the Students have the option to deactivate the control over the system internal workings of the system and of the dynamic lesson generation process and select influence of their actions on system's the lesson contents, e.g. the outcome functions (Höök et al, 1998) concepts, in case that they just want to revise specific concepts of the domain combine instructional design theories with the learning styles theory to develop an adaptation framework that is educationally effective and technologically feasible. This framework unifies several processes that mainly affect system's adaptation, such as structuring the domain model and developing the educational material; assessing learner's knowledge level, and exploiting individual traits (eg, students' dominant learning style); planning the lessons content, delivery and presentation, and providing the appropriate navigation support to learners.…”

Reuse and challenges in evaluating language generation systems

“…The design of the tasks follows the design used in the evaluation of two other adaptive hypermedia applications -PUSH (Höök, 1998) and (Wills et al, 1999). Each of the participants was first given a set of three tasks -each set contained one browsing, one problem-solving, and one information location task.…”

“…Due to the fact that HYLITE+ generates hypertext which content and links are adapted to the user, it can be evaluated following strategies from two fields: NLG and adaptive hypertext. After reviewing the approaches, used for evaluation of the NLG and adaptive hypertext systems most similar to ours,e.g., (Cox et al, 1999), (Reiter et al, 1995), (Höök, 1998), we discovered that they were all evaluated extrinsically by measuring human performance on a set of tasks, given different versions of the system. The experiments were typically followed by an informal interview and/or questionnaire, used to gather some qualitative data, e.g., on the quality of the generated text.…”

Methods and techniques for the evaluation of user-adaptive systems

“…For instance, users' suggestions can be taken into consideration in order to discover the effects generated by the exploitation of adaptive techniques such as reduction of the complexity of the interaction, deeper knowledge of the system functionality, increased satisfaction, reduction of the interaction anxiety, and so on. Höök (1997) argued that the evaluation of an interactive system is always a hard task, and when the system is adaptive the task becomes even harder because the evaluation must distinguish the adaptive features from the general usability problems. Since most adaptive system evaluations are comparisons between the system with and without adaptations, the problem is clear: most of the time the non-adaptive version is not well designed for the tasks.…”