基于编译优化和反汇编的程序相似性检测方法.PDFVIP

2008 6 June 2008 34 6 Journal of Beijing Univers ity of Aeronautics and Astronautics Vo.l 34 No6 赵长海 晏海华 金茂忠 ( , 100191) : 提出了基于编译优化和反汇编的程序相似性检测方法, 能 检测出标识符 重命名增加冗余语句等价的控制结构替换等 12种学生常用的抄袭手段. 基于该方法, 设计 和实现了一个程序相似性检测系统 BuaaSmi , 采用编译优化和反汇编技术将源程序转化为汇 编指令集合, 删除和替换汇编指令中对程序本质特征影响不大的易变元素, 使用 一个与指令顺 序无关的决策函数计算程序相似度; 还给出一个简单有效的聚类算法, 从程序集合中聚类出相 似的程序子集. 通过与著名的 JPlag系统针对两份典型的抄袭样本集进行评测对比, 表明本文 方法的检测效果具有明显的优势. : 抄袭; 程序相似性; 相似性检测; 编译优化 : TP 311. 56 : A : 1001-5965( 2008) 06-0 11-05 A pp roach based on com p iling op tmi ization and d isassem b ling to detec t p rog ram smi ila rity Zhao Changhai Yan H aihua Jin M aozhong ( School of Computer Science and T echnology, Beijing University ofA eronautics andA stronautics, Beijing 100191, China) Abstract: An approach based on compiling optmi ization and disassembling w as proposed to detect smi -i larity in computer programs. It can detect 12 modification strategies that are often used by students, such as renam ing identifiers, add ing redundant statements and replacing control structuresw ith equ ivalent structures. The mi plemented softw are, called BuaaSmi , translates source code into assembly instructionsw ith the help of comp iler and d isassembler, removes and rep laces those easily changed elem ents in the assembly instructions, and applies a decision function to calculate the smi ilarity, w h ich doesnt depend on the order of assembly in- structions. A smi p le clustering algorithm w as also introduced to find all groups of smi ilar programs. By using wt o sets of plagiarized transcripts as testing program s, the comparative evaluation show s thatBuaaSmi hasmore advantages than JPlag,