Auditor/INFO at master · HarkonenBade/Auditor · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
#MIME TYPE plugin

k-neighbour distance normalisation
	Too small a range -> not enough differentiation
	Problematic - needs some measure on reasonable maximums for all values

k-nearest common prefix combination
	Restricted in some way to prevent perfect matching at root

Text plugin
	Most common nouns
	pairwise comparison for distance

THING RIP MENTIONED

Evaluation
	Scripted tests
	Accuracy+speed
	Send prelim graphs to Rip
	profiling

K-nearest neighbour
	division by type


====================

test plan addition

graph performance/number of files


====================

Mention naive baysian clssifier: performance issues when new files introduced, recompute whole database slices.