Design and Implementation of a Machine Learning-Based Authorship Identification Model

Table 2

PAN12 datasets used in the experiments.

Dataset	Description	Training documents	Testing documents
A1	Instance-based (original text)	6	6
A2	Instance-based (n-grams)	6	6
A3	Profile-based (original text)	3	6
A4	Profile-based (n-grams)	3	6
C1	Instance-based (original text)	16	8
C2	Instance-based (n-grams)	16	8
C3	Profile-based (original text)	8	8
C4	Profile-based (n-grams)	8	8
I1	Instance-based (original text)	28	14
I2	Instance-based (n-grams)	28	14
I3	Profile-based (original text)	14	14
I4	Profile-based (n-grams)	14	14