DiPD: Disruptive event Prediction Dataset from Twitter

DiPD数据集由印度马勒瓦亚国立技术学院创建，专注于从Twitter收集的破坏性事件预测数据。该数据集包含263,561条记录，分为事件和非事件两类，分别标记为1和0。数据集通过提取用户特征如关注者数量和地理位置，以及推文特征如转发次数，来分析推文的传播和影响。创建过程中，使用Python的Tweepy API从Twitter获取数据，并进行了多次提取以确保数据的多样性和独特性。该数据集主要用于机器学习领域，特别是事件分类和识别，帮助政府和安全机构监控和预防可能的破坏性事件。

The DiPD dataset was developed by Malaviya National Institute of Technology, India, and focuses on destructive event prediction data collected from Twitter. This dataset contains 263,561 records, categorized into two classes: event and non-event, labeled as 1 and 0 respectively. It analyzes the propagation and influence of tweets by extracting user features including follower count and geographic location, as well as tweet features such as retweet counts. During the dataset construction process, data was retrieved from Twitter using Python's Tweepy API, with multiple extraction rounds carried out to ensure data diversity and uniqueness. This dataset is primarily applied in the machine learning domain, especially for event classification and recognition, to help governments and security institutions monitor and prevent potential destructive events.