1) Java files :

MovieDataConvert.java
Sorting.java
Sorting2.java	
CF_movies.java
CF_densed.java
Item.java
Predict.java
Merge.java
UserAccuracy.java

2) Datasets :

u.data -- The full u data set, 100000 ratings by 943 users on 1682 items.
        Each user has rated at least 20 movies.The data is randomly
        ordered. This is a tab separated list of 
               user id | item id | rating

movies.csv -- This CSV file is generated manually using MovieDataConvert.java file
	            Space between 3 elements are trimmed and replaced by comma.

sorted_movies.csv -- This CSV file is generated using Sorting.java file .
		               movies.csv file is sorted by the user id in ascending order.

item_sim.csv --  This CSV file is generated as the result of the program Item.java
		         This file shows the result of item-item similarity 

prediction.csv -- This CSV file is generated as the result of the program Predict.java 
	                This file will be used to add the predictions to original dataset

merged.csv -- This CSV file is generated using the Merge.java. This is the result of 
		       movies.csv and prediction.csv

densed.csv -- This CSV file will be the densed dataset resulted from the sorting of 
	            the merged.csv . Sorting2.java is used for this sorting .


CF_result_movies.csv -- This CSV file is result of CF_movies.java program .
			          This is the result of Pure Colloborating Filtering algo using 
			          sorted_movies.csv as Input file .

CF_result_densed.csv -- This CSV file is result of CF_densed.java program .
			        This is the result of User-User CF augmenting by Output of Item-Item CF  using 
			        densed.csv as Input file . 

3) Libraries and dependencies :

• commons-math3-3.4.1.jar
• guava-18.0.jar
• mahout-core-0.9.jar
• mahout-integration-0.9.jar
• mahout-math-0.9.jar
• slf4j-api-1.7.10.jar
• slf4j-nop-1.7.10.jar

• To See the recommendation of the Pure Colloborative Filtering : Java File - CF_movies.java

• To Compile - javac CF_movies.java

• To Run - java CF_movies

• To see the recommendation for improved User-user CF augmenting by Output of Item-item CF using densed dataset Java File - CF_densed.java

• To Compile - javac CF_densed.java

• To Run - java CF_densed

• To see the User Accuracy using Mean Absolute Error : Java File - UserAccuracy.java

• To Compile - javac UserAccuracy.java

• To Run - java UserAccuracy

1 - MovieDataConvert.java -- converts u.data to movies.csv

2 - Sorting.java -- Sorts the user id in ascending order in movies.csv file and generates sorted_movies.

3 - CF_movies.java -- Performs pure collaborative filtering on original dataset sorted_movies.csv

4 - Item.java -- Performs item-item similarity that can be used to add more prediction and to make dataset dense. Input file - sorted_movies.csv Output file - item_sim.csv

5 - Predict.java -- Gives prediction raings using input file - item_sim.csv and generates prediction.csv

6 - Merge.java -- Merge the sorted with movies file sorted_movies.csv and Additional predictions , prediction.csv

7 - Sorting2.java --Sort the final sorted densed file - merged.csv . Output file - densed.csv

8 - CF_densed.java - Performs User-User CF augmenting by Output of Item-Item CF using densed.csv as Input file . Output file - CF_result_densed.csv

9 - UserAccuracy.java - Comparison of results using Mean Absolute Error. Gives the MAE for both - pure collaborative filtering and Improved CF (User-User CF augmenting by Output of Item-Item CF)