# Binary Biplob: Can You Decode Emotions? — Data

来源：https://www.kaggle.com/competitions/binary-biplob-can-you-decode-emotions/data

本地 Meta Kaggle 快照；不是本次实时页面。

# Dataset Description

# Dataset Description

## Dataset Files

This competition provides three main files:

- **`example.csv`** — A sample dataset containing labeled examples to help participants understand the data format and features.
- **`test.csv`** — The test set for which participants must predict sentiment labels. It contains all feature columns except the target label.
- **`submission.csv`** — The submission file template where participants upload their predictions.

---

The dataset consists of multilingual social media text samples combining **Bangla** and **English** languages, exhibiting common code-switching patterns commonly found in South Asian digital communication. Each sample is labeled with one of three sentiment classes: **positive**, **negative**, or **neutral**.

## Dataset Features

| Column Name | Description                                          | Type    |
|-------------|------------------------------------------------------|---------|
| `id`        | Unique identifier for each text sample               | String  |
| `text`      | Text content containing Bangla, English, or mixed code-switching | String  |
| `label`     | Sentiment annotation for the text                     | String  |

## Label Classes

| Label    | Description                       |
|----------|---------------------------------|
| positive | Expresses positive sentiment     |
| negative | Expresses negative sentiment     |
| neutral  | Expresses neutral or mixed sentiment |

## Note on Emoticons

The text data contains **emoticons and emojis** that often **change or emphasize the tone** of the message. Participants should be aware of their impact when building sentiment models.