Scribd
Upload
7 views3 pages

Project A1

The document outlines a project focused on detecting cyberbullying on Twitter using a hybrid deep learning approach that combines Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) models. It emphasizes the importance of this system in preventing harm, supporting mental health, and assisting moderators, while detailing the steps to build the system, including data collection, preprocessing, labeling, training, testing, and deployment. Additionally, it highlights the technologies involved and the educational benefits for students participating in the project.

Uploaded by

wickedsohail99
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
7 views3 pages

Project A1

The document outlines a project focused on detecting cyberbullying on Twitter using a hybrid deep learning approach that combines Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) models. It emphasizes the importance of this system in preventing harm, supporting mental health, and assisting moderators, while detailing the steps to build the system, including data collection, preprocessing, labeling, training, testing, and deployment. Additionally, it highlights the technologies involved and the educational benefits for students participating in the project.

Uploaded by

wickedsohail99
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Cyberbullying Detection on Twitter Using Hybrid Deep Learning

1. What is Cyberbullying?

Cyberbullying means hurting someone through online messages, like insults, threats, or harmful
comments. Twitter is a platform where this happens often, because people can post short messages
(tweets) freely and anonymously.

This project is all about building a system that can automatically detect whether a tweet is harmful
or bullying in nature, using Artificial Intelligence (AI) and Deep Learning.

2. What Does “Hybrid Deep Learning” Mean?

In deep learning, we use models that try to mimic how the brain works when processing
information.

A hybrid model means we're combining two types of deep learning models to make a stronger,
smarter system:

• CNN (Convolutional Neural Network): Usually used in image processing, but here it's used
to recognize patterns in text (like spotting bad words or phrases).

• LSTM (Long Short-Term Memory): A special kind of neural network that understands
sequences of words (like grammar or sentence context).

By combining CNN and LSTM, the system can:

• Understand patterns (bad phrases),

• AND also understand the context of a sentence (whether it’s really meant to harm or not).

3. Why This Project is Important to Society?

• Prevents Harm: Helps detect and stop harmful content before it reaches the victim.

• Supports Mental Health: Reduces online harassment, especially among teenagers.

• Assists Moderators: Makes the job easier for platforms like Twitter by flagging harmful
content.

• Educational Value: Students learn real-world AI applications with social impact.

4. How Do You Build This System?

Step-by-Step:

1. Collect Tweets:

o Use Twitter’s API to get real tweets (both normal and bullying ones).

2. Preprocess Tweets:

o Clean the text (remove @mentions, links, emojis).


o Break down sentences into words.

o Remove common filler words (like “the”, “is”).

3. Label the Tweets:

o Manually or semi-automatically tag tweets as "bullying" or "not bullying".

4. Train the Hybrid Deep Learning Model:

o Input the cleaned tweets into the CNN+LSTM model.

o Let it learn to identify patterns of bullying.

5. Test the Model:

o Try it on new tweets to see how well it detects cyberbullying.

6. Deploy the Model:

o Host the model using a web framework so others can use it (like schools, parents,
platforms).

5. Technologies Involved

Technology Purpose

Python Main programming language

TensorFlow / Keras Libraries used to build and train deep learning models

NLP (Natural Language Processing) Used to process and understand human language

Twitter API For collecting real-time tweet data

CNN (Convolutional Neural Network) Detects patterns in text

LSTM (Long Short-Term Memory) Understands word sequences and context

Flask / Django To deploy the model as a website or service

Jupyter Notebook For writing and testing code interactively

6. What Will a Student Learn from This?

• Real-world AI project experience

• Understanding of text data and how to clean it

• How deep learning models are trained and tested

• How to combine models for better results

• How to build a full pipeline: collect data → train → test → deploy

You might also like