Please enable JavaScript.
Coggle requires JavaScript to display documents.
AR-MBA, Sekar Arum Sari
19522362 - Coggle Diagram
AR-MBA
Association Rule & MBA
Association Rule is a data mining technique to find associative rules between a combination of items or a job to determine which attributes will be obtained together.
The Association Rule is well known for its application to analyze the contents of shopping carts in supermarkets, so the Association Rule is also often referred to as Market Basket
Analysis.
An example of the Association Rule from a purchase analysis in a supermarket is to find out how likely a buyer is to buy bread along with milk.
Association Rule is also known as one of the data mining techniques that is the basis of various other data mining techniques
Algorithm of AR-MBA :
Apriori Algorithm
A priori algorithm is an algorithm to reduce the search space for combination items so that analysis can be done more quickly.
Two main processes carried out in the Apriori algorithm (Han & Kamber, 2006) :
Join (merging). In this process, each item is combined with other items until no more combinations are formed.
Prune (pruning). In this process, the results of the combined items are then trimmed using the minimum support specified by the user.
FP-Growth Algorithm
Frequent Pattern Growth (FP-Growth) is one alternative algorithm that can be used to determine the set of data that most often appears (frequent itemset) in a data set (Samuel, 2008).
The characteristic of the FP-Growth algorithm is the data structure used is a tree called FP-Tree. By using FP-Tree, the FP-growth algorithm can directly extract frequent Itemset from FP-Tree.
-
Parameters of AR-MBA
Support. Support is a combination of items in the database, if there is item A and item B, then support is the percentage of transactions in the database containing A and B.
Confidence. Confidence is a correlation or relationship in the database containing Item A and Item B, indicating that how often Item B is purchased when buying Item A.
Lift Ratio. Lift Ratio provides information on whether product A is actually purchased together with product B (Nofriansyah, 2016).
AR Implementation
For marketing, increasing sales can be achieved by reorganizing items so that items sold together are always found together. This
E-commerce sites can also use association rules to provide recommendations (recommender system/suggestive sell) for buyers based on what items are already in their shopping carts.
For direct marketing, promotion to existing customers is preferred, because their purchase data is known.
Google Reader site which is a web application where users can subscribe to feeds from news sites or blogs (users will be able to find out whether a site has been updated or not without having to visit each site) also uses Market Basket Analysis.
-