Survey Of The Status Quo Of Network Teaching Resources Construction In Primary And Secondary Schools Based On Web Data

The big data technology creates new chance for informational teaching, providing the new data sources and methods to study many problems in education for researchers and managers. In this teaching data which is to acquire, teaching resource is token as a teaching data that bear the weight of information of learning behavior, learning activity and learning progress, which imply abundant analyzing value. The emergence of network not only changes people's lifestyle but also traditional teaching methods. And the instructional mode with the help of the network has sprung up. The construction of teaching resource has been valued as an especially pivotal segment, which has been paid more attention. In this context, investigating the present situation in the construction of teaching resources is one of the important means to promote the orderly development of education informatization. In order to investigate the construction situation, the thesis refers to three aspects, discipline, territory and homogeneity. The content is as followed.Firstly, unstructured data on the primary and secondary resources website design the collection and storage stragies, which collects and deals with Hadoop distributed cluster. It can collect the data information on the primary and secondary resources website through the secondary development web crawler Nutch and reserve in the distributed database Hbase, providing data platform to subsequent pages analysis and resources data.Secondly, According to the characteristics of network teaching resources and the structure of web page, a method of establishing teaching resource base is proposed. Teaching resources data consists of the title, keywords, dates, content and so on, which uses developmental tools jsoup to extract the title and keyword, patter matching algorithm to extract the date and media type, line block of distribution function to extract the contents to establish network teaching resources base according to the relationship among the website, webpage and interlinkage.Thirdly, website resources of primary and secondary mostly classify on the subject discipline and region according to the link structure and title to establish the model of VSM, using pattern matching technology and vector similarity technology to analysis the mode of discipline and region. In the analysis of resources homogeneity, we define the homogeneity of the network teaching resource from three levels:the first one is that the resource with the same name is primary homogeneity; followed by the same name and the same type and the third level is the same content. It utilizes the model LDA to establish a subject pattern for the teaching resources content in the third level and analyzes topic model similarity on the basis of hierarchical clustering algorithm.Fourthly, in order to solve a large amount of data calculation problem, we extract website teaching resources and redesign the model VSM and LDA and clustering algorithm to make it run on the programming model of MapReduce, which lays a solid foundation for the entire survey.Experimental results found that the Chinese, Maths and English occupies the proportion of large resources and take an examination of course more than don't need. The resources distribute significantly higher on the east coast and the Pearl River Delta region.website resources reconstruction phenomenon still exists and have a considerable number. Finally, the design of experiments show that the survey is reliable and effective.