This repository contains code for sentiment analysis of Indonesian text data, likely related to internship (magang) experiences or MSIB. The project uses data collected from social media Twitter or X, preprocesses it, and applies a sentiment analysis model to classify the text as positive, negative, or neutral. The process.ipynb notebook provides the main workflow, while function.py contains reusable functions for preprocessing and sentiment prediction.
To set up the project, follow these steps:
-
Clone the repository.
-
Install the required Python packages using pip:
pip install -r requirements.txt
-
Run the
process.ipynbnotebook to execute the data processing and sentiment analysis pipeline.
The data/ directory contains several CSV files with text data from different years, presumably related to "magang" (internship) experiences. Key datasets include:
data/dataset_final.csv: The final processed dataset with sentiment labels.data/magang.csv: Raw tweet data related to "magang" or internship.data/magang2021.csv,data/magang2022.csv,data/magang2023.csv,data/magang.old.csv: Various raw tweet datasets from different years.
The primary analysis is performed within the process.ipynb notebook. This notebook reads the datasets, preprocesses the text data using functions from function.py, performs sentiment analysis using a pre-trained model, and generates visualizations.
- process.ipynb: Jupyter Notebook containing the main data processing and analysis workflow.
- function.py: Python script containing functions for text preprocessing and sentiment analysis.
- requirements.txt: List of Python packages required to run the project.
- data/dataset_final.csv: The final dataset after processing.
The sentiment analysis is performed using the ayameRushia/roberta-base-indonesian-1.5G-sentiment-analysis-smsa model from Hugging Face Transformers.
An interactive dashboard visualizing the sentiment analysis results is available at: