In recent years, the industrial field and the academic circles attach importance to E-learning unceasingly. Proposing many standards E-learning teaching material specifications likes SCORM. Although many E-learning all have evaluative works that start the learning as the new student be improvement basis. However these evaluation all is only the pure feedback, is unable to achieve lets the new student have the adaptive learning environment to obtain the best study result. This research is taking the learners use Dreamweaver teaching study manufacture homepage as an example. Discussing and analyzing different characteristic learners which choice as obtained as result in study process. Diagnoses the different learnerpsilas characteristic which teaches materials using the Bayesian networks and calculates its weight. This teaching material conforms to the SCORM standard E-learning material system. So the teaching material system might defer to recently the different characteristic learners diagnosis result, accumulate and analyze a set of adaptive learning teaching material, could enable the new learner fast to find the suitable teaching material, solved the problem and enhanced the learning efficiency.