Identifying Fake Accounts on Social Networks Based on Graph Analysis and Classification Algorithms

Table 1

Generation of synthetic examples (SMOTE).

---

Consider a sample arr1 and let arr2 be its nearest neighbor.

Arr1 is the sample for which K-nearest neighbors are being identified.

Arr2 is one of its K-nearest neighbors.

Arr1= (0.0045, 0.0014, 0.0145, 0.0046)

Arr2= (0.003, 0.0004, -0.0135, 0.0057)

Let: f1_1=0.0045 f2_1=0.003 f2_1-f1_1= -0.0015

f1_2=0.0014 f2_2=0.0004 f2_2-f1_2= -0.001

f1_3=0.0145 f2_3= -0.0135 f2_3-f1_3= -0.028

f1_4=0.0046 f2_4=0.0057 f2_4-f1_4=0.0011

The new samples will be generated as

(f1^', f2^', f3^', f4^') = Arr1 + rand (0-1) (-0.0015, -0.001, -0.028, 0.0011)

Rand (0-1) generates a random number between 0 and 1.

---

Arr1 and Arr2 denote similarity between v1 and v2 four nodes.