针对计算字符串相似度的RKR-GST算法,分析了与该算法相关的技术并给出算法的流程图,然后在Visual Studio 2008中对该算法进行了实现,详细描述了实现过程中涉及的类与数据结构图,最后对算法的复杂度及算法运行过程中一些参数的选取进行...针对计算字符串相似度的RKR-GST算法,分析了与该算法相关的技术并给出算法的流程图,然后在Visual Studio 2008中对该算法进行了实现,详细描述了实现过程中涉及的类与数据结构图,最后对算法的复杂度及算法运行过程中一些参数的选取进行了讨论。RKR-GST算法在剽窃检测、DNA序列匹配等领域具有广阔的应用前景,该算法在.NET中的实现具有良好的可移植性与可扩展性,可以在多个应用领域中推广使用。展开更多
Evaluation measures play an important role in the design of new approaches, and often quality is measured by assessing the relevance of the obtained result set. While many evaluation measures based on precision/recall...Evaluation measures play an important role in the design of new approaches, and often quality is measured by assessing the relevance of the obtained result set. While many evaluation measures based on precision/recall are based on a binary relevance model, ranking correlation coefficients are better suited for multi-class problems. State-of-the-art rank- ing correlation coefficients like Kendall＇s T and Spearman＇s p do not allow the user to specify similarities between differ- ing object classes and thus treat the transposition of objects from similar classes the same way as that of objects from dissimilar classes. We propose ClasSi, a new ranking corre- lation coefficient which deals with class label rankings and employs a class distance function to model the similarities between the classes. We also introduce a graphical representation of ClasSi which describes how the correlation evolves throughout the ranking.展开更多
Cohesion is a design quality that has a great im- pact on the posterior development and maintenance. As software evolves, the cohesion of the system becomes weaker due to the changes introduced during evolution. Over ...Cohesion is a design quality that has a great im- pact on the posterior development and maintenance. As software evolves, the cohesion of the system becomes weaker due to the changes introduced during evolution. Over evolution, a single responsibility class may be unintentionally assigned other responsibilities, which makes the class less cohesive and more complex and consequently increases the complexity of the entire system. There has been much work on decomposing class responsibilities based on internal class relationships such as method-attribute referencing and internal method calls. However, object-oriented systems involve significant external class relationships carrying important behavioral semantics, which should be taken into account in identifying class responsibilities. In this paper, we present a novel approach for identifying and decomposing classes responsibilities based on method similarity using both internal and external class relationships. We extend the existing work for measuring similarity of internal class relationships and present a distance-based method for measuring external class relationships. We evaluate the approach using three open source applications -- JMeter, JHotDraw, and ArgoUML. The evaluation shows that the presented approach improves precision over the existing work. We validate the results using independent samples T-test and ANOVA applied to a set of hypotheses. The validation confirms that the results are statistically significant.展开更多
文摘针对计算字符串相似度的RKR-GST算法,分析了与该算法相关的技术并给出算法的流程图,然后在Visual Studio 2008中对该算法进行了实现,详细描述了实现过程中涉及的类与数据结构图,最后对算法的复杂度及算法运行过程中一些参数的选取进行了讨论。RKR-GST算法在剽窃检测、DNA序列匹配等领域具有广阔的应用前景,该算法在.NET中的实现具有良好的可移植性与可扩展性,可以在多个应用领域中推广使用。
文摘Evaluation measures play an important role in the design of new approaches, and often quality is measured by assessing the relevance of the obtained result set. While many evaluation measures based on precision/recall are based on a binary relevance model, ranking correlation coefficients are better suited for multi-class problems. State-of-the-art rank- ing correlation coefficients like Kendall＇s T and Spearman＇s p do not allow the user to specify similarities between differ- ing object classes and thus treat the transposition of objects from similar classes the same way as that of objects from dissimilar classes. We propose ClasSi, a new ranking corre- lation coefficient which deals with class label rankings and employs a class distance function to model the similarities between the classes. We also introduce a graphical representation of ClasSi which describes how the correlation evolves throughout the ranking.
文摘Cohesion is a design quality that has a great im- pact on the posterior development and maintenance. As software evolves, the cohesion of the system becomes weaker due to the changes introduced during evolution. Over evolution, a single responsibility class may be unintentionally assigned other responsibilities, which makes the class less cohesive and more complex and consequently increases the complexity of the entire system. There has been much work on decomposing class responsibilities based on internal class relationships such as method-attribute referencing and internal method calls. However, object-oriented systems involve significant external class relationships carrying important behavioral semantics, which should be taken into account in identifying class responsibilities. In this paper, we present a novel approach for identifying and decomposing classes responsibilities based on method similarity using both internal and external class relationships. We extend the existing work for measuring similarity of internal class relationships and present a distance-based method for measuring external class relationships. We evaluate the approach using three open source applications -- JMeter, JHotDraw, and ArgoUML. The evaluation shows that the presented approach improves precision over the existing work. We validate the results using independent samples T-test and ANOVA applied to a set of hypotheses. The validation confirms that the results are statistically significant.
