面向互联网的课程信息集成技术研究

发布时间：2018-09-05 06:12

【摘要】：目前,互联网的普及使得网上的信息量剧增,但由于互联网信息的分散性、复杂性、异构性等特点,想在如此之大的信息世界中查找所需要的信息有时并不是一件容易的事。比如对于在校的学生和教师来说,他们都需要一些优质的习题资源来提高自己的学习和教学水平,由于这些学习资源分散在互联网不同的地方且形式各异,如何充分有效的利用这些资源成为一个亟需解决的问题。本文针对此问题进行了研究,提出了面向互联网的课程信息集成与融合框架及解决策略,主要贡献如下:(1)提出了针对大学课程习题的Web信息集成概念模型。对课程、习题等概念给出了定义,对于不同的习题类型,提出了基于正则表达式的形式化描述方法。(2)提出了面向大学课程的信息集成与融合策略。通过对不同数据源的综合评价确定了恰当的数据源,提出了基于网页快照的习题信息识别与抽取方法,包括习题类型的识别、不同数据源习题信息的融合、基于Web的不完整习题的信息补充等。(3)提出了基于AHP层次分析法和BP神经网络的课程习题评价策略。在综合分析了大量习题样本的基础上,确定了影响习题质量评价的四个因子,即长度、类型、知识点、难度,其中难度因子通过神经网络算法得到,最后通过AHP层次分析法得到了上述四个因子对习题质量的影响权重。(4)基于本文提出的模型和方法,设计实现了一个基于Web的大学课程习题信息集成与融合系统。基于该系统,通过给定一个课程名称,可以自动发现并集成互联网上与该课程相关的习题信息。该系统具有一定创新性,目前尚未看到类似的系统。
[Abstract]:At present, the popularity of the Internet makes the amount of information on the Internet increase dramatically. However, due to the characteristics of the dispersity, complexity and heterogeneity of Internet information, it is sometimes not easy to find the information needed in such a large information world. For example, for students and teachers in school, they all need some high-quality exercise resources to improve their learning and teaching standards, because these learning resources are scattered in different places and in different forms on the Internet. How to make full and effective use of these resources has become an urgent problem. This paper studies this problem and puts forward the Internet-oriented framework and solution of curriculum information integration and integration. The main contributions are as follows: (1) the conceptual model of Web information integration for college curriculum exercises is proposed. The concepts of courses and exercises are defined, and a formal description method based on regular expressions is proposed for different exercise types. (2) A strategy of information integration and fusion for college courses is proposed. Through the comprehensive evaluation of different data sources, the appropriate data sources are determined, and a method of problem information recognition and extraction based on web page snapshot is proposed, including the identification of exercise types, the integration of exercise information from different data sources, Information supplement of incomplete exercises based on Web. (3) the evaluation strategy of course exercises based on AHP analytic hierarchy process and BP neural network is proposed. Based on the comprehensive analysis of a large number of exercise samples, four factors affecting the evaluation of exercise quality are determined, that is, length, type, knowledge point, difficulty, among which the difficulty factor is obtained by neural network algorithm. Finally, the influence weight of the above four factors on exercise quality is obtained by AHP analytic hierarchy process. (4) based on the model and method proposed in this paper, an information integration and fusion system of college course exercises based on Web is designed and implemented. Based on the system, the problem information related to the course can be automatically found and integrated on the Internet by giving a course name. This system has certain innovation, at present has not seen the similar system.
【学位授予单位】：天津理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：G434;TP393.09

【相似文献】