Association rules analysis calculator
(Market Basket Analysis calculator)
To calculate an association analysis (market basket analysis) online, simply copy your data into the table above and select the data you want.
Make sure that your data for the association analysis is in one of the following formats:
- Variant 1: Each row is a transaction or a purchase. One means bought, zero means not bought. So the first person has bought jeans, shirt and shoes.
- Variant 2: Each row is a transaction, the purchased products are in one column and separated by a comma.
In both cases, please note that the first row must be a name, i.e. it is not part of the data. DATAtab will then give you the results.
Definition market basket analysis
This is just a brief overview, please also visit our detailed tutorial on market basket analysis.
Association rule analysis is a technique used in data mining to discover relationships or associations between variables in large datasets. It's most commonly applied in the context of market basket analysis to find out which products tend to be bought together.
With market basket analysis, an online department store can find links between purchased items. The market basket analysis gives an answer to the question, how likely it is that a customer buys product A (Rhs), if he already has product B (Lhs) in the market basket. So which items or products are frequently bought together? The association rules are therefore in the form:
- Left Side Products (Lhs) ==> Right Side Products (Rhs)
where the rule says that if the left side products are present in a transaction, the right side products are likely also present. A simple example of a retail transaction could be that a customer who buys jeans and a shirt (the Lhs) is likely to buy shoes (the Rhs) too.
Results of market basket analysis
The result of an association rules analysis typically consists of:
Itemsets
These are sets of one or more items. For example, "bread, butter" is an itemset with two items.
Rules
These describe the association between two itemsets. For example, the rule "bread" -> "butter" means that if bread is bought, there's a likelihood that butter is also bought. Each rule has two main parts: an antecedent (left-hand side) and a consequent (right-hand side).
Frequency
The frequency in the table of results tells us how often the products of the Lhs and Rhs appear in a transaction.
Support
The support tells us in what percentage of all transactions the products of the Lhs and Rhs occur.
Confidence
Confidence now tells us if the products under Lhs appear in an order, how likely it is that the products under Rhs are then also in the market basket.
Lift
The lift indicates the factor by which the probability of buying the products under Rhs increases if the products under Lhs have already been bought.
Association Rule Mining Calculator
Association rule mining is a method which aims to find frequently occurring patterns, correlations, or associations from datasets. Association Rules are used to explain patterns in data from relational databases and transactional databases.
Therefore, with the support, confidence and lift calculator you can discover relationships or associations between variables in large datasets.