Research On Personalized Writing Resource Push Based On The Needs Of Primary School Students

In the long-term study related to composition teaching for elementary school students, the researcher finds that it's difficult for the teacher to consider every student's individual differences in the model of class teaching about writing. After all,time and energy the teacher can pay for students are so limited. With the educational informization developing quickly, a lot of websites and apps targeted the primary school students' writing have sprung up on the market, which provide services, such as writing guides, supports for writing process, writing resources and so on. However,the writing systems which can recommend personalized writing resources to students are very few, so are the related researches. This paper is based on the problem of individual differences in the writing of primary school students. It takes the way of getting the writing requirements of primary school students as the starting point, and studies how to design a complete model to push personalized writing resources to students in the mature writing platform. In this way, not only it can take care of the individual differences in the writing of primary school students to a certain extent, but also it can help to relieve the teachers' pressure in writing teaching. In recent years,the technology used to push personalized information has been widely used in the field of domestic education, such as e-books, digital textbooks, distance education system and so on, which provides good theoretical basis and technical support for this study.Based on the status quo of researches in the mining of user requirements, this paper firstly adopts two methods to get the primary school students' writing requirements: by analyzing a series of compositions of one student and by the data the student has rated the writing resources. The indices of the first way are determined by the related questionnaire survey and the problems' analysis in the compositions. And the second way mainly stems from the collaborative filtering technology. Secondly,based on the existing researches about personalized push model and the ways this study has adopted to gain the primary school students' demands, this thesis designs a model used to push personalized writing resources to elementary school students in the network writing system, which includes the push technology, the push strategy,the push content and the push form. Thirdly, this study achieves the push model on the Eclipse platform with the mode of B/S. Finally, in order to verify the effectiveness of the pushing technologies, this paper tests the push technology based on the content analysis and the push technology of user-based collaborative filtering, with user survey and offline experiment respectively.The results show that the accuracy rate of push technology based on composition analysis is higher than 75%, but the consistency rate is less than 50%, which indicates that this technology's degree of personalization is not high; To the push technology of user-based collaborative filtering, the method of user similarity calculation adopted in this paper has better recommended quality than the traditional method of Pearson similarity calculation and the method of modified Pearson similarity calculation. In general, the model designed to push personalized writing resources to primary school students in this paper has a certain significance for the research of the recommendation system in the network writing platform. However, the problems in this study, such as the indices used to analysis compositions are not comprehensive,the recommended algorithm is not optimal, the prototype has issues in its stability and security and so on, still need to be further improved.