加权关联规则的个性化科研素养挖掘研究

Application Study in Personalized Science Literacy Using Weight Association Rules Algorithm

针对高校科研统计数据集,通过探索性的数据分析,提取出主要影响科研素养的影响因子,在对各影响因子的特征选择的基础上,进行特征值的离散化、数据规约和特征构造等方面的分析。考虑到科研素养的各影响因子中,不同影响因子有着不同的重要性的特点,在Apriori算法思想的基础上,对各影响因子设定权重值,并将加权关联规则算法,引入到个性化的科研素养的发现中,从而挖掘个性化的、有意义的关联信息。实验结果证明:该算法能够更真实地挖掘并发现高校科研素养中个性化的有趣信息和关联规则,挖掘结果可为科研素养的评价提供决策参考。

This paper, based on the statistical data sets for scientific research with exploratory data analysis, extracted the main factors of scientific literacy in various factors on the basis of feature selection, diseretization of the feature values, data structures, and characteristics such as statute perspectives. Taking into account the impact factor scientific literacy, the author found that different factors have different importance of features in the Apriori algorithm based on the i- deas of various factors setting the weights, and introducing weighted association rules to the personalized scientific literacy discoveries so as to tap the personalized, meaningful information and the association. The experimental results showed that the algorithm can dig and find more real, scientific research accomplishment in the personalized information and interesting association rules, and the mining results can be evaluated for scientific literacy to provide decision - making.