Merge retrde's branch

Code/Malicious URL detection tool/python/main.py

@@ -0,0 +1,159 @@
+import pandas as pd
+import numpy as np
+import sklearn
+
+import random
+from tkinter import *
+from tkinter import ttk 
+import time
+
+from sklearn.feature_extraction.text import CountVectorizer
+from sklearn.cross_validation import train_test_split
+from sklearn.linear_model import LogisticRegression
+from sklearn.feature_extraction.text import TfidfVectorizer
+
+boolean= False
+to_ret=[]
+
+
+def getTkns(input):
+	tknsBySlash= str(input.encode('utf-8')).split('/')
+	allTkns= []
+	for i in tknsBySlash:
+		 tkns= str(i).split('-')
+		 tknsByDot= []
+		 for j in range(0, len(tkns)):
+			 tempTkns= tkns[j].split('.')
+			 tknsByDot= tknsByDot + tempTkns
+		 allTkns= allTkns+tkns+tknsByDot
+	 allTkns= list(set(allTkns))
+	 # Since .com is most common domain, we actually don't need it.
+	 if 'com' in allTkns:
+		 allTkns.remove('com')
+	 return allTkns
+
+
+def trainer():
+	global to_ret
+	global txt
+
+	# ONE TIME EXECUTION STARTS HERE
+	csv= pd.read_csv("data.csv", error_bad_lines= False)
+	data= pd.DataFrame(csv)
+	```
+	data= np.array(data)
+	random.shuffle(data)
+
+	y= [d[1] for d in data]
+	corp= [d[0] for d in data]
+
+	count_vectorizer= CountVectorizer(tokenizer= getTkns)
+	tf_vectorizer= TfidfVectorizer(tokenizer= getTkns)
+
+	X1= count_vectorizer.fit_transform(corp)
+	X2= tf_vectorizer.fit_transform(corp)
+
+	# Fit the Logistic Regression Model
+	X1_train, X1_test, y1_train, y1_test= train_test_split(X1, y, test_size= 0.2, random_state= 42)
+	X2_train, X2_test, y2_train, y2_test= train_test_split(X2, y, test_size= 0.2, random_state= 42)
+	lgs_count= LogisticRegression()
+	lgs_tf   = LogisticRegression()
+	lgs_count.fit(X1_train, y1_train)
+	lgs_tf.fit(X2_train,y2_train)
+
+	out1= "The accuracy of Model with Count Vectorizer is "+str(lgs_count.score(X1_test, y1_test))+"\n"
+	out2= "The accuracy of Model with TFIDF Vectorizer is "+str(lgs_tf.score(X2_test, y2_test))+"\n"
+	txt.insert(0.0,out1)
+	txt.insert(1.0,out2)
+
+
+	to_ret.append(count_vectorizer)
+	to_ret.append(tf_vectorizer)
+	to_ret.append(lgs_count)
+	to_ret.append(lgs_tf)
+	return to_ret
+
+
+
+
+def callback():
+
+	global boolean
+	global to_ret
+	global txt
+	global choice
+
+	txtname= entry.get()
+	if txtname=="":
+		txt.insert(0.0,"Enter valid URL\n")
+		return
+	Choice= choice.get()
+	if Choice == "":
+		txt.insert(0.0,"Please choose any one of the two options.\n")
+		return
+
+
+	X_predict= [txtname]
+    	if Choice=="CV":        
+		X_predict= to_ret[0].transform(X_predict)
+		y_predict= to_ret[2].predict(X_predict)
+		out= str(txtname)+" is found "+str(y_predict)+"\n"
+		txt.insert(0.0, out)
+    	if Choice=="TF":        
+		X_predict= to_ret[1].transform(X_predict)
+		y_predict= to_ret[3].predict(X_predict)
+		out= str(txtname)+" is found "+str(y_predict)+"\n"
+		txt.insert(0.0, out)
+
+  	return
+
+
+
+root= Tk()
+choice= StringVar()
+frame= ttk.Frame(root)
+frame.pack()
+frame.config(height= 600, width= 800)
+frame.config(relief= RAISED)
+
+label= ttk.Label(frame, text="Malicious URL Detection Tool")
+label.config(foreground='#1aaedb',background='#dbbf23')
+label.config(justify=CENTER)
+label.config(font=('segoe',18,'bold'))
+label.pack()
+
+label2= ttk.Label(frame, text='\n\n')
+label2.pack()
+
+ttk.Radiobutton(frame,text='Use Count Vectorizer',variable=choice,value='CV').pack(anchor='w')
+label3= ttk.Label(frame, text='\n')
+label3.pack()
+
+ttk.Radiobutton(frame,text='Use Tf-Idf Vectorizer',variable=choice,value='TF').pack(anchor='w')
+label4= ttk.Label(frame, text='\n')
+label4.pack()
+
+label5= ttk.Label(frame, text='Enter any valid URL to Check')
+label5.pack()
+
+to_check= StringVar()
+entry= ttk.Entry(frame,textvariable=to_check,width=60)
+entry.pack()
+
+
+
+label5= ttk.Label(frame, text='\n')
+label5.pack()
+
+progressbar= ttk.Progressbar(frame,orient= HORIZONTAL,length=200)
+progressbar.pack()
+progressbar.config(mode= 'indeterminate')
+
+
+button= ttk.Button(frame,text='Check',command=callback).pack()
+
+
+txt= Text(frame,height=10, wrap= WORD)
+txt.pack()
+trainer()
+root.mainloop()

Code/Malicious URL detection tool/python/run-all.sh

@@ -0,0 +1,4 @@
+# Giving sudo priviledges 
+
+sudo python3 init.py
+sudo python3 main.py

Code/Malicious URL detection tool/readme.md

@@ -0,0 +1,14 @@
+## Malicious URL detection tool
+
+### Execution options:
+
+- Python:
+  - execute main.py from terminal or ide having installed dependencies of requirements.txt
+
+- Web-app:
+  - open index.html at any javascript compatible HTML5 engine with CSS3 support
+
+### To-do:
+
+- Complete web-app files index.html, add some styling and javascript code
+- If possible, convert Python tool to a web-app!

Code/Malicious URL detection tool/web/index.html

@@ -0,0 +1,8 @@
+<!DOCTYPE HTML5>
+<head>
+
+</head>
+<body>
+  Add site compatible page of tool
+</body>
+</HTML>

Code/Malicious URL detection tool/web/js/index.js

@@ -0,0 +1,3 @@
+var train(){
+  // Create JS equivalent function for trainer() from python
+}

Code/Malicious URL detection tool/web/readme.md

@@ -0,0 +1,6 @@
+## Web version of Malicious URL detection tool
+
+### To-do list:
+
+- Add a free subdomain for site
+- Compelete the site :(

Code/Malicious URL detection tool/web/style/main.css

@@ -0,0 +1,9 @@
+.background {
+  color: white,
+  image: none
+}
+
+.h1 {
+  color: black,
+  font-size: 18
+}