Electronic Paper Automatic Generation System Based On Semantic Online Knowledge Resource

As a huge information space, web has become an important way for people to acquire learning information. Nowadays, there are many online computer programming problem archives for beginner to training. They usually provide a html format problem, and need problem solver to use a command-line program to solve it. But when beginner can hardly get any help information when he can't solve problems. This will make the beginner feel very hard to get start.There are mainly three reasons for this phenomenon: 1) Most of problem archives are lack of difficulty analysis and knowledge statistic 2) When a beginner can't solve a problem, he can't get the any hint on that 3) Current online problem archives don't help user to choose problem. So for some beginner, they don't which problem is the most suitable for them.For these reasons, this thesis describes a way to convert the online problem to be semantic, so that computer can divide the hard problem to be smaller and easier ones. Also, this paper describes a new way to analyze the difficulty of computer programming problems based on many experimental data.The main focus of this thesis is on how to make the online computer programming problem semantic, so that they can be understand by computer. And then computer can divide the problem to some smaller and simpler problem based on some rules. Moreover, after the problem is divided, the new problem must have the description that can be read by human beings. So another work of this thesis is to convert the problem in semantic format to nature language description. The main technology that this thesis used to generate nature language is the Nature Language Generation (NLG). Besides problem dividing, this thesis also describes a method to analyze the difficulty of single problem. So that computer can combine suitable problem for beginner. This system has been used on our university programming training many times. The result proves that the feasibility and correctness of this thesis.The result of this thesis, for example nature language generation and difficulty of problem, not only can be used in computerprogramming domain but also can be used in other domain after somechanging. For example, the nature language generation can be used inon line shopping system. So the result of this thesis is general in certaindegree.