imranraad/github-emotion-surprise

---
task_categories:
- text-classification

---
# AutoTrain Dataset for project: github-emotion-love

## Dataset Description

Dataset used in the paper: Imran et al., ["Data Augmentation for Improving Emotion Recognition in Software Engineering Communication"](https://arxiv.org/abs/2208.05573), ASE-2022.

This is an annotated dataset of 2000 GitHub comments. Six basic emotions are annotated. They are Anger, Love, Fear, Joy, Sadness and Surprise. This repository contains annotations of all emotions.

## Dataset Structure

Dataset is in CSV format. The columns are:

```id, modified_comment, Anger, Love, Fear, Joy, Sadness, Surprise```

Here, `id` is a unique id for each comment. Each emotion is marked as 1 or 0.

### Dataset Splits

This dataset is split into a train and test split. The split sizes are as follows:

| Split name | Num samples |
| ------------ | ------------------- |
| train | 1600 |
| test | 400 |

原始信息汇总

AutoTrain Dataset for project: github-emotion-love

数据集描述

该数据集用于论文：Imran et al., "Data Augmentation for Improving Emotion Recognition in Software Engineering Communication", ASE-2022。

这是一个包含2000条GitHub评论的标注数据集。标注了六种基本情绪：愤怒、爱、恐惧、喜悦、悲伤和惊喜。该存储库包含所有情绪的标注。

数据集结构

数据集格式为CSV。列包括：

id, modified_comment, Anger, Love, Fear, Joy, Sadness, Surprise

其中，id是每个评论的唯一标识符。每种情绪标记为1或0。

数据集划分

该数据集被划分为训练集和测试集。划分大小如下：

划分名称	样本数量
train	1600
test	400