kevalmorabia97 / fparm-frequent-patterns-and-association-rule-miner

This is an implementation of Apriori algorithm for frequent itemset generation and association rule generation. The GUI is made using JAVA FX or Cmd_Line version can be used

FPARM - Frequent Patterns and Association Rule Miner

This is an implementation of Apriori algorithm for frequent itemset generation using HashTree data structure and association rules are generated from these frequent itemsets
There is a GUI version and a command line version.

Running The Program from FPARM-Frequent-Patterns-and-Association-Rule-Miner folder
Note: For Command_Line_Version, the dataset path should be specified in src\Command_Line_Version\Main.java file

Command_Line_Version:
$ javac src\Command_Line_Version\Main.java
$ java src.Command_Line_Version.Main

GUI_Version:
$ javac src\GUI_Version\Main.java src\GUI_Version\MainController.java
$ java src.GUI_Version.Main

Note: If GUI_Version is not running properly then you need to install Oracle JDK as below:

$ sudo add-apt-repository ppa:webupd8team/java
$ sudo apt-get update
$ sudo apt-get install oracle-java8-installer

You can directly run the application by extracting the jar inside FPARM-Frequent.Patterns.and.Association.Rule.Miner.zip file and running it. You will need ORACLE JDK to run the JAVA Fx .jar file.
The program can handle any kind of data as long as it is in the required format
Note: Dont directly run from zip file as the files generated will be stored in cache. So first extract then only run the .jar.

---

• Support is an indication of how frequently the itemset appears in the dataset. The support of X with respect to T is defined as the proportion of transactions t in the dataset which contains the itemset X.
  

• Confidence is an indication of how often the rule has been found to be true. The confidence value of a rule, X => Y with respect to a set of transactions T, is the proportion of the transactions that contains X which also contains Y.
  

• Apriori Algorithm uses a breadth-first search strategy to count the support of itemsets and uses a candidate generation function which exploits the downward closure property of support.
  Read More: https://en.wikipedia.org/wiki/Apriori_algorithm

---

1. First, the application generates frequent 1 itemsets.
2. After generating Frequent K itemsets, it generate Frequent K+1 itemsets by using F_k-1 x F_k-1 approach.
3. The frequent itemsets are stored in frequentItemsets.data file
4. When all frequent itemsets are generated, associating rules that have confidence greater than minimum confidence are generated and stored in the file named associationRules.data file.

Read about Association Rule Mining: https://en.wikipedia.org/wiki/Association_rule_learning

---

To generate frequent itemsets and association rules you have to select the transaction file which should be of the following form:

apple,ball,egg
apple,ball,dandiya,egg,fish,apple
ball,caramel,dandiya,fish,milk
apple,caramel,fish,fish
apple,ball,dandiya,fish,milk
ball,caramel,fish

Attribute Names Required? - YES or NO

If YES then first line of dataset mush contain names of attributes seperated by comma(,)
Example: buying,maint,doors,person,lug_boo,LABEL

Required if the data is of the following form:

 | buying | maint | doors | persons | lug_boot | LABEL |
---------------------------------------------------------
 | vhigh  | vhigh |  2    |    2    |  small   | unacc
 | vhigh  |  high |  2    |   more  |   big    | unacc
 | vhigh  |  med  |  2    |    4    |  small   |  acc 
 |  high  | vhigh |  4    |    2    |  small   | unacc 

So, the dataset implies that the data is of the form: buying=vhigh,maint=vhigh,doors=2,persons=2,lug_boot=small,LABEL=unacc
Doing this, ambiguity in data values can be eliminated
That is, buying=vhigh is considered different than maint=vhigh

For generaly market basket datasets, this is not required as the the values in each dataset row
does not mean that it belongs that particular column

Note:

• If the transaction has more than one appearance of an item then only one appearance is considered
• If you have transaction in the form of like this then also the program works

bus=yes,plane=no,rich=yes

• If you have continuous data then you can descretize the data then run this application
  For Ex: divide age in groups like
  Age=[0,10)
  Age=[10,20)
  Age=[20,30)
  and so on...

Age=[21,30),Salary=[70k,120k),Browser=Mozilla

---

After that, the transaction is processed and converted to numbers which makes it easier to handle in the hashtree The processed transaction is saved inside processedTransaction.data file which looks like this:

0,1,2,
0,1,2,3,4,
1,3,4,5,6,
0,4,5,
0,1,3,4,6,
1,4,5,

---

After that the frequent itemsets are generated which are stored in frequentItemsets.data file which looks like this:

Frequent 1 Itemsets:
ball, (8)
egg, (3)
dandiya, (5)
fish, (6)

Frequent 2 Itemsets:
apple, ball, (5)
fish, caramel, (3)

Frequent 3 Itemsets:
apple, ball, egg, (3)

Here the number inside parenthesis indicates the support count of the frequent itemset

---

Finally, the association rules are generated and stored in associationRules.data file which looks like this:

apple, (6) ----> ball, (8) conf(0.8333333)
egg, (3) ----> ball, (8) conf(1.0)
milk, (5) ----> caramel, (6) conf(0.8)
dandiya, (5) ----> fish, (6) conf(0.8)
egg, (3) ----> apple, (6) conf(1.0)

Here the number inside parenthesis besides the itemset is its support count and conf(*) denotes the confidence of the association rule

To-Do: Handling continuous data