Update README.md

README.md

@@ -31,33 +31,71 @@ Changes:
 - Added Polarized column
 ```
 
-IndoToxic2024 is an Indonesian dataset collected before and during the 2024 Indonesia presidential election.
-The data are obtained from social media and are annotated by 29 annotators of diverse backgrounds.
-The tasks supported by this dataset are text classification tasks around hate speech, toxic, and polarizing content.
+# IndoToxic2024: A Multi-Labeled Indonesian Discourse Dataset
 
-`annotators_id` is a list of strings, containing `annotator_id` of the text's annotator. 
-12700 out of 28448 (44.64%) entries are annotated by more than 1 annotator.
+## Dataset Overview
 
-You can load this dataset by doing the following:
-```py
-from datasets import load_dataset
+IndoToxic2024 is a multi-labeled dataset designed to analyze online discourse in Indonesia, focusing on **toxicity, polarization, and annotator demographic information**. This dataset provides insights into the growing political and social divisions in Indonesia, particularly in the context of the **2024 presidential election**. Unlike previous datasets, IndoToxic2024 offers a **multi-label annotation** framework, enabling nuanced research on the interplay between toxicity and polarization.
 
-ds = load_dataset("Exqrch/IndoToxic2024", "main")
-```
+## Dataset Statistics
 
-If you want to load the annotator information, do the following instead:
-```py
-from datasets import load_dataset
+- **Total annotated texts:** **28,477**  
+- **Platforms:** X (formerly Twitter), Facebook, Instagram, and news articles  
+- **Timeframe:** September 2023 – January 2024  
+- **Annotators:** 29 individuals from diverse demographic backgrounds  
 
-ds = load_dataset("Exqrch/IndoToxic2024", "annotator")
-```
-# PENDING Baseline Performance
-<!-- The table below is an excerpt from the paper, listing the baseline performance of each task:
-![image/png](https://cdn-uploads.huggingface.co/production/uploads/61e997f66735d3a73e291055/Z-88FC_nxAYgLUUWhwfmQ.png) -->
+### Label Distribution
 
-If you use this dataset, please cite:
-```
-Susanto, L., Wijanarko, M. I., Pratama, P. A., Hong, T., Idris, I., Aji, A. F., & Wijaya, D.
-(2024, June 27). IndoToxic2024: A Demographically-Enriched Dataset of Hate Speech and
-Toxicity Types for Indonesian Language. arXiv.org. https://arxiv.org/abs/2406.19349
-```
+| Label        | Count |
+|-------------|-------|
+| **Toxic**   | 2,156 (balanced) |
+| **Non-Toxic** | 6,468 (balanced) |
+| **Polarized** | 3,811 (balanced) |
+| **Non-Polarized** | 11,433 (balanced) |
+
+## Dataset Structure
+
+The dataset consists of texts labeled for **toxicity and polarization**, along with **annotator demographics**. Each text is annotated by at least one coder, with **44.6% of texts receiving multiple annotations**. Annotations were aggregated using majority voting, excluding texts with perfect disagreement.
+
+### Features:
+- `text`: The Indonesian social media or news text
+- `toxicity`: List of toxicity annotations (1 = Toxic, 0 = Non-Toxic)
+- `polarization`: List of polarization annotations (1 = Polarized, 0 = Non-Polarized)
+- `annotators_id`: List of annotator_id that annotate the text (anonymized) -- Refer to `annotator` subset for each annotator_id's demographic informatino
+
+## Baseline Model Performance
+
+![image/png](https://cdn-uploads.huggingface.co/production/uploads/61e997f66735d3a73e291055/mWNkGL_RdqdKQE7rzK59a.png)
+
+### Key Results:
+
+We benchmarked IndoToxic2024 using **BERT-based models** and **large language models (LLMs)**. The results indicate that:
+
+- **BERT-based models outperform LLMs**, with **IndoBERTweet** achieving the highest accuracy.
+- **Polarization detection is harder than toxicity detection**, as evidenced by lower recall scores.
+- **Demographic information improves classification**, especially for polarization detection.
+
+### Additional Findings:
+- **Polarization and toxicity are correlated**: Using polarization as a feature improves toxicity detection, and vice versa.
+- **Demographic-aware models perform better for polarization detection**: Including coder demographics boosts classification performance.
+- **Wisdom of the crowd**: Texts labeled by multiple annotators lead to higher recall in toxicity detection.
+
+## Ethical Considerations
+
+- **Data Privacy**: All annotator demographic data is anonymized.
+- **Use Case**: This dataset is released **for research purposes only** and should not be used for surveillance or profiling.
+
+## Citation
+
+If you use IndoToxic2024, please cite:
+
+```bibtex
+@misc{susanto2025multilabeleddatasetindonesiandiscourse,
+      title={A Multi-Labeled Dataset for Indonesian Discourse: Examining Toxicity, Polarization, and Demographics Information}, 
+      author={Lucky Susanto and Musa Wijanarko and Prasetia Pratama and Zilu Tang and Fariz Akyas and Traci Hong and Ika Idris and Alham Aji and Derry Wijaya},
+      year={2025},
+      eprint={2503.00417},
+      archivePrefix={arXiv},
+      primaryClass={cs.CL},
+      url={https://arxiv.org/abs/2503.00417}, 
+}```