search

Lopezm0908 / SMS-filtering

0 stars 0 forks source link

readme

Inbox Guardian

Description

Inbox Guardian is an SMS spam filter meant to be deployed in addition to any existing SMS application allowing users 
the ability to enjoy all the features of their favorite SMS application along with the added protection against spam 
and fishing messages. Inbox Guardian employs the use of a logistic regression model, the BERT language model, and 
other holistic methods such as keyword recognition to flag spam messages.

Why?

Spam is a pervasive threat to our productivity and safety when using mobile devices, we at Inbox Guardian find the 
current integrated spam filters in most messaging applications to be inadequate to block the constant barrage of 
ever evolving spam messages. This is why we designed the Inbox Guardian, a robust and comprehensive protective measure 
using the latest artificial intelligence to block spam with a tested 90% accuracy rate.

Quick Start

Setup the inbox Guardian API

1. Clone repository [Link](https://github.com/CameronJ2/InboxGuardian-API)
2. Install requirements listed in the requirements text folder contained in the main directory using the command 
    pip install -r requirements.txt
3. Run the server.py file using the command 
    python server.py

Run the project in Android Studio

1. Open the application under smsfilteringapplication 
2. Run the mainactivity.kt file

Ensure app is default SMS app

1. Open settings 
2. Navigate to apps
3. Navigate to Default Apps 
4. Navigate to SMS app 
5. Select smsfilteringapplication from the list of apps given

Usage

Available Features:

  1. Whitelist a Number - Exempt a number from analysis
  2. Blacklist a Number - Block all incoming texts and calls from a specific number
  3. Keyword Blocking - Block messages containing specific keywords
  4. Message Reporting - Report a message as spam, this blacklists the number and deletes the message.
  5. Evaluation Mailbox - If a message is blocked by keyword analysis, BERT analysis, or logistic regression analysis, the message is sent here. The user can then decide whether to flag the message as spam and block the number or send the message through to the inbox.
  6. BERT Analysis - Using the BERT language model we can accurately detect whether or not a message is spam or not spam.
  7. Logistic Regression Analysis - Using a machine learning model trained on a large dataset of spam messages we are able to accurately detect whether or not a message is spam or not spam.
  8. Settings Menu - Opt in and out of different filters and features to make the application better suit an individual users needs

    Examples

Adding a number to the whitelist

1. Use the gui to navigate to the whitelist menu
2. Press the add number button
3. Enter the number of your choice
4. Hit the confirm button
5. Your number is now in the whitelist

Adding a number to the blacklist

1. Use the gui to navigate to the blacklist menu
2. Press the add number button
3. Enter the number of your choice
4. Hit the confirm button
5. Your number is now in the blacklist

Adding a keyword to filter out

1. Use the gui to navigate to the keyword menu
2. Press the add keyword button
3. Enter the keyword of your choice
4. Hit the confirm button
5. Your keyword is now a flag for spam messages

Reporting a message as spam

1. Use the gui to navigate to the message reporting screen
2. Select the message you would like to report by pressing on the message itself 
3. In the pop up menu just press confirm 
4. The message is now deleted and the number is blocked.

Evaluating a message

1. Use the gui to navigate to the evaluation mailbox screen
2. Select the message you would like to evaluate by pressing on the message itself
3. In the pop up menu select either flag or approve 
4. If you approve the message it will be sent to the inbox
5. If you flag the message the message is deleted and the number that sent it is added to the blacklist

Edit Settings

1. Use the gui to navigate to the settings menu 
2. Use the checkboxes to opt in and out of specific features or filters

Demo

https://github.com/Lopezm0908/SMS-filtering/assets/158241209/9370c61f-fa78-4955-aba4-418992dd5b7c