Analysis of massive data sets
CF.java is a program which solves the problem of Collaborative Filtering Recommender System It can give a prediction of user-user or item-item filtering. It expects the input in the following format: N M N rows and M columns Q q queries; query looks like this (I, J, T, K)
example:
5 5
1 2 X 2 4
2 X 3 X 5
3 1 X 4 X
X 2 4 X 4
1 X 3 4 X
3
1 3 0 1
4 1 0 2
5 5 1 3
PARAMETERS:
X is the value in matrix where the user or item doesn't have a rating
N <=100 is number of rows of the following matrix, also items
M <=100 is number of columns of following matrix, also users
Q <=100 is number of queries
I <= N is i coordinate of the rating we want to predict in the matrix
J <= M is j coordinate of the rating we want to predict in the matrix
T is type of filtering, if 0 it is user-user, if 1 it is item-item
K <= N, M is cardinality, i.e. biggest number of values taken into account
NodeRank.java is a programm that calculates rank of each node in a directed graph. This is done by PageRank algorithm implementation. It doesn't consider dead end nodes. It expects the input in the following format: n B // n is number of nodes, B is beta probability A01 A02 A03 ... A0k0 // in the i-th line there are nodes to which i-th node is looking, 0-based indexing A11 A12 A13 ... A1k3 A21 A22 A23 ... A2k2 ... An−11 An−12 An−13 ... An−1kn q // number of queries n1 t1 // n-node index, t-iteration number [1, 100] n2 t2 ... nq tq