Research And Improvement Based On Apriori Algorithm And Its Application In Wisdom Endowment

The advent of the era of big data brings convenience to all industries,but at the same time it also causes the problem of too much information,from which people cannot get effective information in a timely manner.Information overload,in which the internal relationships between different factors and different transactions cannot be presented.The emergence of data mining technology helps people to explore the special relationship within transactions,so as to analyze and explore more intuitively.The research focus of this thesis is mainly on association rules.Apriori algorithm,a classical algorithm,is selected to introduce the conceptual properties of the algorithm,and the advantages and disadvantages of the algorithm in the process of use are analyzed.The classical Apriori algorithm can effectively discover the hidden internal relationship between data,but the algorithm also has the problem that the larger the number of candidate sets,the greater the cost.To solve this problem,a new improved algorithm,RTI?Apriori(Repeat Transaction and Item Sum),is proposed.The algorithm for the shortage of the traditional Apriori algorithm is improved,the idea is to introduce Boolean matrix to store the transaction information,the rows of the matrix corresponding to the transactions in the transaction database set,the columns of the matrix corresponding to the transaction on the project,item exists in the transaction focus is marked as "1" in the corresponding position,otherwise set to "0".By scanning the Boolean matrix constructed by the transaction database and the "logical and" operation of the matrix,we can find the corresponding association rules.At the end of the matrix,1 row and 2 columns were added to assist to calculate the support degree of item set,to help the matrix compress,so as to simplify the operation,and then scan and delete the item set that does not meet the conditions according to the result.At the same time,the parallel method is introduced to optimize the algorithm in parallel,so as to improve the algorithm efficiency.The experimental results show that the improved algorithm is more efficient than the original algorithm without multiple scanning of the database.The RTI?Apriori algorithm is applied to the intelligent pension system to dig out the internal correlation between the elderly's pension mode and many factors,such as children,spouse,income,occupation,education background,etc.The core of the system is the dynamic management of the elderly information files,while giving the elderly daily care,understand the elderly's health needs,for the society and family members to provide more reference and help.