Detecting Hateful and Offensive Language using NLP

Local Chapter Aswan, Egypt Chapter

Coordinated byEgypt ,

Status: Completed

Project Duration: 27 Jan 2023 - 27 Mar 2023

Open Source resources available from this project

Project background.

The spread of hate speech and fake news is a serious side effect of expanding social networks. Most of the time covered by anonymity, social network users feel more comfortable speaking hate or disseminating fabricated information as opposed to real life, where they have to confront the consequences of what they say. All major social networks make an enormous effort to deal with such a problem, and machine learning is a powerful tool in this arena.

The problem.

We want to avoid the impacts of hate speech and fake news in our homes that would negatively affect our friends and families especially: children as they are the least expert with technology. As a result, we would use our technology experience to tackle this problem and provide a safe environment to them.

Project goals.

1. Train datasets. 2. Carryout data pre-processing 3. Develop a Machine learning model 4. Use NLP and advanced Machine Learning 5. Develop an API to be used by individuals.

Project plan.

  • Week 1

    Data Collection

  • Week 2

    Data Pre-Processing

  • Week 3

    Exploratory Data Analysis

  • Week 4

    Create language model (1)

  • Week 5

    Create language model (2)

  • Week 6

    Train data on various models

  • Week 7

    Test data on various models

  • Week 8

    Deploy the system as API

Learning outcomes.

1. The participants will deploy their data analysis skills: (such as data wrangling and preprocessing, problem solving using python,…etc.)
2. Practice Machine learning as well as neural nets background in a real-world project, in order to enrich their experience and skilling new fields (e.g. NLP and more advanced machine learning algorithms).
3. Offer a solution that may serve in solving the challenge they face.

Share project on: