An Efficient Prediction of Missing Itemset in Shopping Cart

Abstract

Many researches has focused mainly on how to expedite the search for frequently co-occurring groups of items in “shopping cart” and less attention has been paid to the methods that exploit these “frequent itemsets” for prediction purposes. This study contributes to this task by proposing a technique that uses the partial information about the contents of a shopping cart for the prediction of what else the customer is likely to buy. Several algorithms have been introduced to detect the frequently co occurring group of items in the transactional databases for prediction purposes. This study presents a new technique whose principal diagonal elements represent the association among items and looking to the principal diagonal elements, the customer can select what else the other items can be purchased with the current contents of the shopping cart and also reduces the rule mining cost. The association among items is shown through Graph. The frequent itemsets are generated from the Association Matrix. Then association rules are to be generated from the already generated frequent itemsets. We conducted extensive experiments and showed that the accuracy of our algorithm is higher than the previous algorithm. Our experiments show that the time needed for predicting the items is highly reduced than other algorithms. Moreover the memory requirement is also less since our work does not generate candidate itemsets. In this study, we have successfully implemented the Rule generation technique and predicted the set of other items that the customer is likely to buy. The performance of our algorithm outperforms the existing algorithm that needs multiple passes over the database in such a way that it efficiently mines the association among the items in the shopping cart and the prediction time of the items is greatly reduced

Redes de regras de associação filtradas e multialvo

“…A maioria dos artigos que utilizaram árvores, o fizeram na fase de extração (ANAND; VINODCHANDRA, 2018;SHEU, 2018;GAO et al, 2017b;SHA, 2017;OLIINYK et al, 2017;TURGEMAN;SCIULLI, 2017;MONDAL;BHATTA-CHARYA;MONDAL, 2016;WEN;ZHONG;WANG, 2015;NIRMALA;PALANISAMY, 2013). As árvores foram utilizadas como estrutura base, nas quais os nós eram os itemsets frequentes e as regras eram obtidas percorrendo a árvore.…”

“…As principais bases de dados utilizadas nos artigos são as disponibilizadas pela UCI Machine Learning Repository 2 . Os trabalhos utilizam essas bases para seus experimentosGUO;USMAN;USMAN, 2016;CARVALHO;REZENDE, 2016;IQBAL et al, 2014;CARVALHO, 2014;NIRMALA;PALANISAMY, 2013; LIU; ZHAI; PEDRYCZ, 2012; ZHAO; ZHANG; ZHANG, 2011) a fim de validar as abordagens propostas e comparar com resultados já existentes na literatura.Nadi et al (2014) utilizaram dois datasets, "Connect" e "Chess". Trabalhos relacionados a dados de Redes Sociais utilizaram datasets em comum como o "Pokec", "YAGO2", "GraMi" e "Web graph" (WANG; XU, 2018; FAN; WU; XU, 2016).…”