📰 Bangla News Text Analysis & Topic Modeling

📌 Objective

The goal of this project is to perform complete text analysis on Bangla news articles scraped from the NTV BD news portal.
It demonstrates practical skills in:

• Web scraping Bangla news articles
• Text preprocessing for Bangla language
• Visual analytics (word frequency, word clouds, bar plots)
• Topic modeling with Latent Dirichlet Allocation (LDA)

---

🛠 Features

• Web Scraping: Extracted Bangla news article texts from multiple category pages on NTV BD.
• Text Preprocessing: Removed punctuation, digits, extra spaces, and common Bangla stopwords.
• Word Frequency Analysis: Visualized top words with WordCloud and frequency bar plots.
• Topic Modeling (LDA): Discovered hidden topics and visualized top words per topic.
• Document–Topic Distribution: Interpreted results with stacked bar plots for each article.

---

📊 Workflow

1. Scraping → Collect articles from NTV BD using rvest.
2. Preprocessing → Clean Bangla text for analysis.
3. Word Frequency Analysis → Generate word cloud & bar plots.
4. Topic Modeling → Apply LDA to discover topics.
5. Visualization & Interpretation → Topic distributions and insights.

---

🧰 Tech Stack

• Language: R
• Libraries:
  • rvest – Web scraping
  • readr – Reading data files
  • dplyr – Data manipulation
  • purrr – Functional programming tools
  • tidyr – Data tidying
  • stringr – String processing
  • tm – Text mining
  • tokenizers – Tokenization
  • wordcloud – Word cloud generation
  • RColorBrewer – Color palettes for visualizations
  • topicmodels – LDA topic modeling
  • tidytext – Text mining using tidy data principles
  • ggplot2 – Data visualization

---

🚀 How to Run

1. Clone the repo:

   git clone https://github.com/WasifAsad/Bangla-News-Text-Analysis.git
   cd bangla-news-text-analysis

2. Open the .R scripts or R Markdown files.

3. Install required libraries:

   install.packages(c("rvest", "readr", "dplyr", "purrr", "tidyr", "stringr","tm", "tokenizers", "wordcloud", "RColorBrewer","topicmodels", "tidytext", "ggplot2"))
   

4. Run script : Bangla-Text-Analysis.R

N.B. :

• You can change your news portal by replacing the website in base_url.
• Also change the output directories (Word Cloud, .csv file, etc.) according to your own folder paths.
• Also you can add other Bangla stop words.

---

📸 Sample Outputs (For the specific day that we ran the code on)

Word Cloud

Top Words Visualization per Topic

---

📜 License

This project is licensed under the MIT License. Feel free to use, modify, and distribute this project.

---

📧 Contact

For inquiries or support, feel free to reach out:
Authors:
Wasif Asad Alvi
Md. Tamjid Hossain
Rifat Talukdar
Md. Tanziul Haque
Email: [wasifasad35@gmail.com]
GitHub: WasifAsad