Research On Web Multimedia Resources Analysis And Text Information Extraction

With the development and popularization of information technology, educational informatization has become an important part and the main symbol of modernization of education. It not only helps to train their innovative spirit and comprehensive ability and to improve teaching effectiveness, but also provides the conditions and safeguards for the overall development of students. The development and utilization of information resources of education is the core of education informatization. It is the key to success of educational informatization. Information resources of education play a very important role in the creation of constructivism learning environment. To design constructivism learning environments, it is necessary to provide rich and related with problem solving information resources for the learners.The rapid development of Internet and the rising level of information contribute to that the number of Web pages is presentting the explosive growth in geometric progression. The growing number of the pages contain large amounts of multimedia resources, including images, audio, video and flash, etc, and the multimedia resources become richer and richer. As an important part of information resources, Multimedia resource plays a very important role for its vivid and intuitive advantages in the constructivist learning environment. It has become more difficult to find their desired multimedia information in the vast Web. To build Web multimedia resource library, and make it apply in the area of education, to help teachers and learners can quickly and accurately locate multimedia resources which they need, is an important problems to be solved by educational technology workers.Multimedia resources in the Web are generally embedded in Web pages, to accurately find and locate these resources, we need a description of these multimedia resources, to form the indexed database of multimedia resources. But if to search for and describe its information in artificial way, the efficiency is very low and operation is quite complicated. If the the text infomation of multimedia resources can be automatically extracted from the Web page, then it is helpful to describe and retrieve Web multimedia resources and construct Web multimedia resources indexed database of multimedia resources.By analysing a large number of multimedia resources on the Web page and summarizing various types of multimedia resources in the form of Web presence of various types of multimedia resources, using Web Multimedia Web page searcher to collect Web pages that contains multimedia resources,on the basis of the above work,in the thesis, a system is designed to analyse Web multimedia resources and extract the text information. The system use a set of heuristic rules to locate the region of multimedia resources in Web pages, then extract the related text of multimedia resources, and translate the text, segment chinese words, filter the results of sementation and extract keywords, and then mark multimedia resources in the Web.The experimental results show that the system has higher accuracy of extracting the text information of Web multimedia resources, which has positive significance to improve the recall and precision of multimedia information retrieval system. If the above method is applied in the field of education, this can be very helpful to create constructivist learning environment and to help students find the multimedia resources they need more effective and accurately, and can greatly improve teaching effectiveness.