'The 100 Most Influential Persons in History': A Data Mining Perspective

摘要

Data mining has been widely applied in various domains, however, there have been limited studies into discovering hidden knowledge from factual data about selected groups of people with special characteristics. It is important to mine data about such group of individuals to extract insightful knowledge that could lead to a better understanding of their personalities, in addition to further sociological conclusions. This paper presents the application and outcome of data mining techniques, namely data clustering and association rules extraction, to find common features and relations among social, environmental and socioeconomic factors from the lives of known influential individuals in history. The mining process was initiated by constructing a dataset through defining, extracting, and retrieving important known facts about these individuals from selected and reliable sources. Second, association rules discovery algorithms were applied in order to show interesting patterns and highlight relations between attributes. Finally, the data were clustered into different groups and each cluster was further analyzed to identify its most strongly defining attributes. The extracted association rules showed how some factors are related, such as the effect of environment type and order of birth in the family on the age at which the individual first engaged with their domain of influence. The clustering exercise demonstrated that influential people who grew up in families of a similar size and financial status share many similar characteristics.