This project involves data analysis on the Global Terrorism Dataset, leveraging Natural Language Processing (NLP) techniques to analyze incident descriptions and classify terrorist organizations.
The analysis focuses on exploring and understanding patterns within the Global Terrorism Dataset, aiming to derive insights and trends related to terrorist incidents worldwide.
- Pandas
- NumPy
- Matplotlib
- Seaborn
- Data Loading: Importing the Global Terrorism Dataset.
- Exploratory Data Analysis (EDA): Exploring key statistics, trends, and visualizing the data.
- Geospatial Analysis: Mapping incidents to understand geographical patterns.
- Temporal Analysis: Analyzing trends over time.
- Statistical Insights: Extracting statistical information to better understand the dataset.
NLP techniques are applied to analyze the textual descriptions of each incident in the dataset. This involves extracting meaningful information and patterns from the unstructured text.
- BERT
- WordCloud
- Text Preprocessing: Cleaning and preparing the incident descriptions.
- Tokenization: Breaking down text into tokens.
- Sentiment Analysis: Understanding the sentiment of incident descriptions.
- WordCloud Visualization: Visualizing the most frequent words in the dataset.
Classification algorithms are employed to categorize and label terrorist organizations based on specific attributes or features.
- PyTorch
- Feature Engineering: Identifying relevant features for classification.
- Model Selection: Choosing a suitable classification algorithm.
- Training and Evaluation: Training the model and evaluating its performance.
- Confusion Matrix: Analyzing the classification results.
Summarize key findings and insights from the data analysis, NLP, and classification processes.
Provide instructions on how to replicate the analysis and use the code in this repository.
If you'd like to contribute to this project, please follow the guidelines in [CONTRIBUTING.md].
This project is licensed under the [MIT License] - see the [LICENSE.md] file for details.