Research On Knowledge Link Discovery Of Massive Solutions To Programming Problems

Currently, there are many programming resources on the web, such as massive programming problems on the major programming OJ system. The famous internal OJ systems are known as HDU, POJ, ZOJ, the popular foreign OJ system are known as UVA and so on. Every OJ system has provided thousands of programming problems on the web. At the same time, there are massive solution reports to these programming problems on the web, which are written by programming players or amateurs. Unfortunately, these massive solution reports have not been well organized and used by vast programming learners and amateurs.Based on the National Natural Science Foundation of China (NSFC), this paper focuses on more specific problems; the range of study is located on programming knowledge resources on the web, which is an instance of the web knowledge resources on the web. Massive programming problems scattered on various OJ systems are connected to the corresponding solution reports to these programming problems. And we extract the programming knowledge from their corresponding solution reports. Then we form a learning sequence of programming problems and their solution reports according to the difficulty of each programming problem and the frequency of each knowledge point used by the solution reports on the purpose of better used for programming teaching.This paper expands as follows:Firstly, The process of discovering programming resources automatically involves the following process:a) downloading the title information of programming problems from OJ systems on the web; b) finding keywords from the downloaded title information of programming problems for searching their solution reports; c) getting the URLs of solution reports of the programming problems with a search engine, and removing several invalid URLs; d) downloading the solution reports according to the URLs;Secondly, we collect and collate the name and alias of data structure, math knowledge, and algorithm knowledge which ACM/ICPC requests, including the Chinese name and English name, and write them to an expert file.Thirdly, we extract the programming knowledge from their corresponding solution reports by matching the predefined programming knowledge points or their alias in the expert file. Then we take programming knowledge set from various problems’ solution reports through knowledge identification as the problem-solving character set; combine with all kinds of problem-solving character set to get the commonly used problem-solving character set; and form a learning sequence of programming problems and their solution reports according to the difficulty of each programming problem and the frequency of each knowledge point used by the solution reports.