A collection of course materials, tutorials, and projects covering the full pipeline of social media analytics — from data collection to advanced NLP modelling — with a focus on Indonesian-language content and the Twitter/X platform.

social-media-analytics/
├── Tutorial1_TextMining/          # Text mining fundamentals
├── Tutorial2_Topic Modelling/     # Topic modelling with LDA & variants
├── Tutorial3_Data Collection/     # Twitter data collection (API & Twint)
├── tugas_1/                       # Assignment 1: basic data analysis
├── proyek tengah semester/        # Mid-term project: user profiling
├── proyek akhir semester/         # Final project: stance detection
├── graph1.ipynb                   # Graph/network analysis basics
├── graph2.ipynb                   # Extended graph analysis
├── pagerank.ipynb                 # PageRank algorithm on social networks
└── script.py                      # Twitter API utility script (Tweepy)

Covers the core NLP pipeline applied to social media text:

• Text preprocessing (tokenization, stopword removal, normalization)
• Word Embeddings (Word2Vec, GloVe)
• Transformer-based Language Models (BERT, IndoBERT)

Unsupervised discovery of latent topics from Twitter corpora:

• Latent Dirichlet Allocation (LDA)
• Indonesian-language datasets (e.g., trending topics on Twitter)
• Preprocessing with colloquial lexicon & abbreviation dictionaries

Methods for collecting social media data:

• Twitter API v2 via tweepy
• Twint for scraping without API rate limits
• Structured storage of collected tweets

Predicting user attributes from tweet content and profile metadata:

• Gender classification using TF-IDF, LSTM, and Transformer models
• Occupation classification using large Transformer models
• Exploratory Data Analysis (EDA) and error analysis included

End-to-end analysis of opinion and influence on Twitter:

• Tweet collection via Twint
• Stance detection (e.g., pro/against a topic) using fine-tuned Transformers
• Network analysis: retweet/like graphs, PageRank-based influence scoring

1. Clone the repository

   git clone https://github.com/nichsedge/social-media-analytics.git
   cd social-media-analytics

2. Set up a virtual environment

   python -m venv .env
   source .env/bin/activate        # Linux/macOS
   .env\Scripts\activate.bat       # Windows

3. Install dependencies (per tutorial/project folder as needed)

   pip install tweepy pandas numpy nltk scikit-learn gensim transformers networkx matplotlib seaborn

4. Open notebooks

   jupyter notebook

• Some notebooks use Indonesian-language datasets and lexicons (e.g., colloquial-indonesian-lexicon.csv, stopwordsID.csv).
• Twitter API credentials in script.py are for reference only — replace with your own keys before running.
• Large dataset files (.csv) may not be included in the repository due to size constraints.

This repository is intended for educational purposes as part of a Social Media Analytics course.