| With the rapid development of globalization, social competition is becoming larger and larger. To adapt ever-changing world, people expand their knowledge constantly and improve their abilities. In such a learning social background, the development of distance learning platform undoubtedly has brought people hope and convenience, as it can make people review old knowledge and study new knowledge by network anytime, anywhere.At present, however, the learning processes of many distance learning systems have been arranged by designers in advance, so that every user acquires the same learning contents and styles, which apparently can't meet the demands of learning. Along with the rapid development of theories and technologies of pedagogy,psychology, artificial intelligence and computer network, the concept of "personalize learning" has been cognized by many experts, and become a researched hotspot. Through analyzing the characteristics of learners to construct a personal learning platform, in which each learner may acquire his interested content according to his demands in study or the like ways ,and hence achieves the best learning effects.In this thesis, on the basis of related theories and the development of technology of researching and analyzing personalized learning carefully, a learning system is constructed based on students' interests. This system majorly includes three modules: user management module, resource management module and students' learning modules. User management module retains the registered information of users, and endows their corresponding accesses according to the different identities of registered users. Resource management module, in which,each added learning text must be passed by some processes,such as Chinese word segmentation etc, mainly realizes the teachers'maintenance managements about learning resource and modeling of text vector.Then a corresponding text vector is constructed. Students' learning module mainly includes the first and continuing learning modules of students.When students study firstly, the system makes two ways to collect the learning interested characteristics of students, one is students select the interested keywords directly, another is students describe their interested contents.Then the collected information is expressed by a vector space model. And finally, the learning texts are provided by matching the similarity of interested vectors of students and text vectors of learning resources. When the student studies it next time, he (or she) can choose to continue to study the recommend texts last time, or may demand to recommend new learning texts.At last, this thesis compares the recommended results of traditional model matching keywords directly and the constructed model in it. Returned results show that the recommended effect of the constructed model is better than the traditional model. |
