Improved Market Basket Analysis with Utility Mining

Srivastava, Nishant; Stuti,; Gupta, K. C.; Baliyan, Niyati

doi:10.2139/ssrn.3170300

Cited by 8 publications

(17 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Association rule mining is the most extensively used (and promising) technique in the data mining field [14]. Since its introduction, it has created several opportunities to mine high-utility data [2]. It helps connect people with common interests [3] and provides new opportunities to improve public health and medicine; the prediction of post-liver transplantation survival is one example [4,5].…”

Section: Related Workmentioning

confidence: 99%

“…If the rule {l} ⇒ {m, n} was chosen as a sensitive rule, then the victim item would be (n) because it has lower support compared to (m), where α(n) = 5 and α(m) = 6. Therefore, (n) should be removed from a certain number of transactions; this number is δ, which can be computed using Equation (2). So, altering one transaction is enough to hide the rule {l} ⇒ {m, n} and this is due to the small size of the dataset in our example.…”

Section: Tracing Examplementioning

confidence: 99%

“…So, for a child with six entries (3, 6, 9, 1, 12, 5), the randomly selected fraction is 50%. Then entries (1,2,3) are checked with the mutation rate; if entry (3) is a hit, it will go to the second step and number (9) should be replaced by another random number selected uniformly by considering the upper and lower limits for this entry.…”

Section: Tracing Examplementioning

confidence: 99%

“…In a set of transactions, ARM focuses on finding associations between frequent items, enabling researchers to predict the frequency of one itemset depending on the frequency of another item in a transaction. The ARM technique has been extensively used in several applications, including privacy preservation [1], market-basket transaction data analysis [2], recommendation [3], health care [4,5], prediction [6], pattern finding in web browsing [7], ranking of text documents [8], and hazard identification [9], among others. The extraction of association rules from transaction datasets was initiated in 1993 [10].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Efficient Association Rules Hiding Using Genetic Algorithms

Bux

Wang

et al. 2018

Symmetry

View full text Add to dashboard Cite

In today’s world, millions of transactions are connected to online businesses, and the main challenging task is ensuring the privacy of sensitive information. Sensitive association rules hiding (SARH) is an important goal of privacy protection algorithms. Various approaches and algorithms have been developed for sensitive association rules hiding, differentiated according to their hiding performance through utility preservation, prevention of ghost rules, and computational complexity. A meta-heuristic algorithm is a good candidate to solve the problem of SARH due to its selective and parallel search behavior, avoiding local minima capability. This paper proposes simple genetic encoding for SARH. The proposed algorithm formulates an objective function that estimates the effect on nonsensitive rules and offers recursive computation to reduce them. Three benchmark datasets were used for evaluation. The results show an improvement of 81% in execution time, 23% in utility, and 5% in accuracy.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Tracing Examplementioning

confidence: 99%

Section: Tracing Examplementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Efficient Association Rules Hiding Using Genetic Algorithms

Bux

Wang

et al. 2018

Symmetry

View full text Add to dashboard Cite

show abstract

“…Security is also considered to be an important aspect in transactions done individually and frequent itemsets for databases that are horizontally partitioned. With the aim of providing security for the frequently utilized itemsets that is purchased finally for the transaction, this research work introduces a novel key security algorithm that employs RSA cryptographic technique in which the classifier optimizes the key values [8,9].…”

Section: Introductionmentioning

confidence: 99%

Market Basket Analysis with Enhanced Support Vector Machine (ESVM) Classifier for Key Security in Organization

Yoganandhini,

Prabakaran

2019

IJEAT

View full text Add to dashboard Cite

Market Basket Analysis is considered to be one among the highly popular and efficient sort of data analysis exploited in the marketing and retailing field. The objective of market basket analysis lies in deciding the products purchased together by the customers. Its name has originated from the concept of customers filling into a shopping cart everything of all they had purchased (a "market basket") while doing shopping in the grocery. Having a knowledge of the products that customers buy in group can be quiteusefulfor a retailer or to any other organization. A store could make the best use of this information to keep the products that are often sold together in the same place, whereas a catalog or World Wide Web (WWW) merchant could utilize it for deciding the structure of their catalog and order form. Since several applications such as market basket analysis, fraud detection in web, medical diagnosis, census data, Customer Relationship Management of business that makes use of association rules exists, the process involving Decision making can be improved. Security is also regarded to bean important facet for transactions done individually and frequent itemsets for database that are horizontally partitioned. In order to render security for lastly bough often used itemsets for transaction purposes, this research work introduces a novel key security algorithm that uses RSA cryptographic technique which is classifier based. The classifier makes use of information about several often utilized itemsets and it provides a key value to the actual company. For instance, in case if there are any reliance users, only the valid users can obtain that market info. The rest of the users belonging to the reliance organization are not allowed to select the data’s key value. First, the frequent itemsets are mined with the help of association rule mining employing Probabilistic Graphical Model techniques. Then the Enhanced Support Vector Machine (ESVM) classifier checks the key values of the mined frequent itemsets.

show abstract

Efficient Apriori Algorithm using Enhanced Transaction Reduction Approach

Arcos

Hernandez

2019

2019 IEEE 13th International Conference on Telecommunication Systems, Services, and Applications (TSSA)

View full text Add to dashboard Cite

Improved Market Basket Analysis with Utility Mining

Cited by 8 publications

References 13 publications

Efficient Association Rules Hiding Using Genetic Algorithms

Efficient Association Rules Hiding Using Genetic Algorithms

Market Basket Analysis with Enhanced Support Vector Machine (ESVM) Classifier for Key Security in Organization

Efficient Apriori Algorithm using Enhanced Transaction Reduction Approach

Contact Info

Product

Resources

About