Datasets:
FredZhang7
commited on
Commit
•
499003f
1
Parent(s):
324be8b
add examples of data preprocessing
Browse files- enron_spam.py +24 -0
- spam_assassin.js +30 -0
- spam_assassin.py +35 -0
enron_spam.py
ADDED
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import pandas as pd
|
2 |
+
|
3 |
+
|
4 |
+
df = pd.read_csv('enron_spam_data.csv')
|
5 |
+
|
6 |
+
data = []
|
7 |
+
for index, row in df.iterrows():
|
8 |
+
message = row['Message']
|
9 |
+
subject = row['Subject']
|
10 |
+
if not message or type(message) == float:
|
11 |
+
message = subject
|
12 |
+
subject = ''
|
13 |
+
elif not subject or type(subject) == float:
|
14 |
+
subject = message
|
15 |
+
message = ''
|
16 |
+
spam_ham = row['Spam/Ham']
|
17 |
+
is_spam = 1 if spam_ham.lower() == 'spam' else 0
|
18 |
+
text = (str(subject) + """
|
19 |
+
|
20 |
+
""" + str(message)).strip()
|
21 |
+
data.append({'text': text, 'is_spam': is_spam})
|
22 |
+
|
23 |
+
df = pd.DataFrame(data)
|
24 |
+
df.to_csv('enron_spam_clean.csv', index=False)
|
spam_assassin.js
ADDED
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
const fs = require('fs');
|
2 |
+
const corpus = require('@stdlib/datasets-spam-assassin');
|
3 |
+
|
4 |
+
const data = corpus();
|
5 |
+
let csvData = "text,is_spam\n";
|
6 |
+
|
7 |
+
|
8 |
+
function cleanText(str) {
|
9 |
+
if (str.startsWith('""') && str.endsWith('""')) str = str.substring(2, str.length - 2)
|
10 |
+
else if (str.startsWith('"') && str.endsWith('"')) str = str.substring(1, str.length - 1)
|
11 |
+
str = str.replace(/["]/g, '""')
|
12 |
+
if (str.includes('\n') || str.includes(',') || str.includes(`""`)) str = '"' + str + '"'
|
13 |
+
return str
|
14 |
+
}
|
15 |
+
|
16 |
+
|
17 |
+
for (let i = 0; i < data.length; i++) {
|
18 |
+
const text = data[i].text;
|
19 |
+
const group = data[i].group;
|
20 |
+
let isSpam = 0;
|
21 |
+
if (group.includes('spam')) {
|
22 |
+
isSpam = 1;
|
23 |
+
}
|
24 |
+
csvData += cleanText(text) + ',' + isSpam + '\n';
|
25 |
+
}
|
26 |
+
|
27 |
+
fs.writeFile('spam_data.csv', csvData, (err) => {
|
28 |
+
if (err) throw err;
|
29 |
+
console.log('CSV file saved!');
|
30 |
+
});
|
spam_assassin.py
ADDED
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import pandas as pd
|
2 |
+
|
3 |
+
special = []
|
4 |
+
df = pd.read_csv('spam_data.csv')
|
5 |
+
|
6 |
+
for index, row in df.iterrows():
|
7 |
+
text = row['text']
|
8 |
+
|
9 |
+
if text is not None and text != '' and 'Subject: ' in text:
|
10 |
+
subject = text.split("""
|
11 |
+
Subject: """)[1]
|
12 |
+
|
13 |
+
if '\n' in subject:
|
14 |
+
subject = subject.split("\n")[0].strip()
|
15 |
+
elif '\r' in subject:
|
16 |
+
subject = subject.split("\r")[0].strip()
|
17 |
+
else:
|
18 |
+
raise Exception('No newline found in subject: ' + str(text))
|
19 |
+
content = text.split("""
|
20 |
+
|
21 |
+
""")[1:]
|
22 |
+
content = """
|
23 |
+
|
24 |
+
""".join(content)
|
25 |
+
|
26 |
+
text = subject + """
|
27 |
+
|
28 |
+
""" + content
|
29 |
+
|
30 |
+
df.at[index, 'text'] = text
|
31 |
+
else:
|
32 |
+
special.append({'text': text, 'is_spam': row['is_spam']})
|
33 |
+
|
34 |
+
df.to_csv('spam_data_clean.csv', index=False)
|
35 |
+
pd.DataFrame(special).to_csv('special.csv', index=False)
|