Financial Transactions
Detailed Task Description
This
dataset contains data about the financial transactions between various
accounts. Its sanitized data from BOA. This data has embedded threats in it.
Try to find out everything that is “suspicious” or
“questionable.”
Dataset contains
following files:
a
folder of images of all the disjointed networks
a
file called Exportation.png that describes all the exports from all the
countries we chose (source is Wikipedia)
a
file called keywords.png that links up the keyword ID used in the graphs to a
keyword name (note that not all keywords are used in the ntr data)
and
a file called accountnames.txt that links up the account ID used in the graphs
to the name in the ntr file.
Parsing the ntr file:
number of
keywords (k)
0,
1,
2,
3,
:
:
k-1
number of
accounts (n)
0,
1,
2,
3,
:
:
n-1
number of
transactions (m)
0,
1,
2,
:
:
m-1
Each line in
the transaction is interperted as follows:
sender_ID,
receiver_ID, amount, date, number_of_keywords, keyword_ID(s)
For example,
if sender_ID is 0, the account's name is the first line in the list of
accounts.
The same is
true for receiver_ID and keyword_ID.
The
number_of_keywords tells how many keyword ids there are to follow. For example,
if
number_of_keywords
is 2, the there will be two numbers after that, delimited by a semicolon
(e.g., 6;7).
Take the
first transaction in the file:
7,0,360000.0,2005-10-11,2,1;2
This can be
translated to:
Sender: Account-57
Receiver: Account-76
Amount: 360,000
Date: October 11, 2005
Keywords: Hardware and Machinery, South Africa
Dataset Download