---
language:
- en
license: mit
task_categories:
- text-generation
- text-classification
- token-classification
- summarization
- translation
size_categories:
- 10K<n<100K
tags:
- humor
- social-media
- comments
- memes
- occupations
- instagram
configs:
- config_name: default
  data_files:
  - split: train
    path: data/comments.csv
---

# My Job Sounds Illegal Dataset

A dataset of ~42,000 humorous social media comments collected from a viral reel prompt:

> "Comment your job but make it sound illegal."

The dataset contains creative, exaggerated, and comedic descriptions of professions and daily work activities written to sound suspicious, criminal, or absurd while remaining harmless and humorous.

Examples include:
- "I manipulate vulnerable people into buying things they don't need."
- "I convince tiny humans to obey me for 8 hours a day."
- "I drug people legally and charge them for it."

This dataset is intended for:
- NLP experimentation
- Humor generation
- Meme and internet culture research
- Text classification
- Social media language analysis
- Fine-tuning conversational or creative language models

## Dataset Structure

The dataset is provided as a CSV file with the following format:

| column | type | description |
|---|---|---|
| comment | string | A user comment describing a job in a humorous "illegal sounding" way |

Example:

```csv
comment
"I sell addictive substances to tired adults every morning."
"I interrogate children about triangles and fractions."
"I break into people's houses and leave with their trash."
````

## Usage

### Load with 🤗 Datasets

```python
from datasets import load_dataset

dataset = load_dataset("YOUR_USERNAME/my-job-sounds-illegal")
print(dataset["train"][0])
```

### Load CSV directly

```python
import pandas as pd

df = pd.read_csv("comments.csv")
print(df.head())
```

## Intended Use

This dataset is designed for research and entertainment purposes only.

Possible use cases:

* Humor detection
* Style transfer
* Creative paraphrasing
* Meme culture analysis
* Social media NLP
* Instruction tuning for comedic outputs

## Limitations

* Comments are user-generated and may contain:

  * slang
  * sarcasm
  * spelling mistakes
  * offensive language
  * stereotypes or sensitive jokes
* The dataset reflects internet humor and should not be treated as factual occupational descriptions.
* No demographic or personal metadata is included.

## Ethical Considerations

The dataset contains publicly posted social media comments collected for research and educational purposes. Users should ensure compliance with platform terms of service and privacy considerations before redistribution.

## Citation

If you use this dataset in research or projects, please cite:

```bibtex
@dataset{my_job_sounds_illegal,
  title={My Job Sounds Illegal},
  year={2026},
  author={YOUR_NAME},
  url={https://huggingface.co/datasets/YOUR_USERNAME/my-job-sounds-illegal}
}
```

## License

MIT

```
```