基于时态软集的诺贝尔科学奖获奖数据分析

Data analysis of the Nobel Prizes in science based on temporal soft sets

针对传统关联规则挖掘方法无法揭示特定时段内项集之间潜在联系的问题,提出一种基于时态软集和Apriori算法的关联规则挖掘算法。考虑经典软集的时态扩展结构,通过时态粒化映射诱导出数据集的粒化结构,构建时态软集及其Q-片段软集。基于时态软集建立描述和挖掘时态关联规则的基本框架,利用Q-片段软集提取特定时段内的数据,并借助所提算法挖掘促进型强时态关联规则,扩展软集理论在时态关联规则挖掘中的应用。对诺贝尔科学奖获奖数据的分析表明,所提方法可以提取出被传统方法忽略的某些强规则,挖掘出的时态关联规则有利于更客观地描述数据中隐藏的事实。

Potential connections between item sets during some specific periods can hardly be revealed with traditional association rule mining methods.To address this issue,a method based on temporal soft sets and the Apriori algorithm is developed to extract association rules.Considering the temporal extension structure of soft sets,the granular structure of a given data set is induced by virtue of temporal granulation mappings,and temporal soft sets with their Q-clip soft sets are constructed.Based on temporal soft sets,a fundamental framework for describing and mining temporal association rules is established.The data in specific periods can be captured by means of Q-clip soft sets.Strong temporal association rules are extracted with the proposed approach,which extends the application of soft sets in temporal association rule mining.The analysis of the award-winning data for the Nobel Prizes in science shows that the proposed method can extract some strong rules ignored by traditional methods,and the extracted temporal association rules are helpful for describing the facts hidden in the data more objectively.