Ti-ABSA：藏文方面级情感分析数据集

基于藏语的情感分析数据集 Ti ABSA 由微博数据和微博评论数据两部分组成。这两个部分都包含三个情绪类别：积极、消极和中立。Ti ABSA 数据集共包含 10577 个样本，其中微博数据集 5872 个样本，微博评论数据集 4705 个样本。数据集以 JSON 格式存储，每个文件包含三个字段：text、aspectTerm 和 aspectPolarity。text 字段记录藏文文本的内容;aspectTerm 字段捕获文本中的藏文 aspect 项（表示正在评估的特定对象或 face），而 aspectPolarity 字段记录与每个 aspect 项对应的情感极性。情绪极性分为积极、中性或消极，分别由值 2、1 和 0 表示。微博数据共包含 36705 个方面词，而微博评论包含 9043 个方面词。

The Tibetan-based sentiment analysis dataset Ti ABSA consists of two parts: Weibo data and Weibo comment data. Both parts include three sentiment categories: positive, negative, and neutral. The Ti ABSA dataset contains a total of 10,577 samples, among which 5,872 samples belong to the Weibo dataset and 4,705 samples belong to the Weibo comment dataset. The dataset is stored in JSON format, and each file contains three fields: text, aspectTerm, and aspectPolarity. The text field records the content of the Tibetan text; the aspectTerm field captures the Tibetan aspect terms in the text, which refer to the specific object or target being evaluated, while the aspectPolarity field records the sentiment polarity corresponding to each aspect term. The sentiment polarities are categorized into positive, neutral, and negative, represented by values 2, 1, and 0 respectively. The Weibo dataset contains a total of 36,705 aspect terms, while the Weibo comment dataset contains 9,043 aspect terms.

搜集汇总

背景与挑战

背景概述

Ti-ABSA是一个专门用于藏文方面级情感分析的数据集，包含10577个样本，涵盖积极、消极和中性三种情感类别。数据集以JSON格式存储，包含文本内容、方面词及其情感极性，适用于藏文情感分析研究。