CyberGuards

The popularity of online social networks has created massive social communication among their users. This communication leads to a huge amount of user-generated data. Cyberbullying has grown into a major problem with the growth of online communication and social media which has been recognized as a

Project Title

CyberGuards

Project Area of Specialization

Cloud Infrastructure,

Project Summary

The popularity of online social networks has created massive social communication among their users. This communication leads to a huge amount of user-generated data. Cyberbullying has grown into a major problem with the growth of online communication and social media which has been recognized as a serious mental health issue among social network users and developing an efficient detection model holds tremendous practical significance. Twitter have become the house of cyberbullying. The problem is that bullied children rarely share this with their parents or guardians. To solve this problem, we will develop a classification system that can detect bullying messages and its severity from online social network platforms such as, Twitter. Furthermore, a web-based or android application will be designed and developed for end users to demonstrate the effectiveness of developed classification model. The developed web or android system will have capability to send the detected bullying messages to victims’ parents in order to monitor bullying activities happening with their children and to provide valuable assistance to their children.

Project Objectives

The software system “CyberGuards” will be developed for the parents of teenagers who are conscious about their child safety in online world. The system will let them know if the child is being bullied online, this will help parents to counsel their children.

Project objectives are stated as under:

• Development of Classification Model
• Evaluation of Classification Model
• Development of Android Application
• Evaluation of application results and User Experience

Project Implementation Method

This system has three main modules to develop namely, classification model development for detecting the bullying messages and its severity, development of web-based application for end-users, and the evaluation of developed web-based application and constructed classification model.In the first module we will develop classification model for bullying message detection using supervised machine learning approaches. In this module, initially the tweets will be collected from the Twitter. Afterwards, the collected tweets will be labelled into either ‘low bullying’, ‘medium bullying’, ‘high bullying’, or ‘no bullying’ classes. After labelling, the dataset will be divided into training and testing set using random sub-sampling method where 70% of the tweets will be used for training and remaining for testing. Furthermore, several preprocessing techniques will be employed to remove the irrelevant features or text from the collected tweets. The preprocessed tweets then will be transformed into numeric vectors through state-of-the-art feature engineering to form a numeric master feature vector. This master feature vector will be given as an input to machine learning algorithm to develop classification model. Finally, the performance of the developed classification model will be evaluated using the test set. Several performance metrics including accuracy, F-score, and area under the curve will be used to measure the classification model performance.

In the second module, the aforementioned classification model will be embedded in a web-based application that will be used by end-user. The end-user will link the twitter accounts of their children to web application by giving username. Periodically, the accounts will be checked if any user will send or receive the bullying message then our developed system will detect that text and inform the sender and receiver parents through message about the bullying activity.

In a third module, we will conduct a user experience survey and to evaluate the performance of our developed system.

Benefits of the Project

There are no such commercialized platforms providing this kind of parental control feature although there are many researches going-on on detecting cyberbullying and According to Cyberbullying Research Center Articles, 35% of Teenagers are the victims of online cyberbullying and due to not sharing this with others worst consequences occurs. Our Project will definitely provide benefits to potential market and it will provide valuable assistance to parents in counseling their victimized children.

Technical Details of Final Deliverable

Technical details of final deliverables are as under:

• Platform: Android App
• Operating System version: above android jellybean 4.2 Database: Google Firebase
• Servers: AWS Services:
• Twitter integration Maintainance: monthly/weekly system
• updates Future Support: Yes.
• API/Services: Tweepy to access twitter data and python web service to connect backend front end.
• System Requirement Specification
• Developed Classifier
• System Specifiaction Document
• Login Screen
• Signup Screeen

Final Deliverable of the Project

Software System

Type of Industry

Health , Security , Telecommunication

Technologies

Artificial Intelligence(AI), Cloud Infrastructure, Others

Sustainable Development Goals

Good Health and Well-Being for People, Sustainable Cities and Communities

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com