Social networks have become a major platform for people to disseminate information, which can include negative rumors. In recent years, rumors on social networks has caused grave problems and considerable damages. We attempted to create a method to verify information from numerous social media messages. We propose a general architecture that integrates machine learning and open data with a Chatbot and is based cloud computing (MLODCCC), which can assist users in evaluating information authenticity on social platforms. The proposed MLODCCC architecture consists of six integrated modules: cloud computing, machine learning, data preparation, open data, chatbot, and intelligent social application modules. Food safety has garnered worldwide attention. Consequently, we used the proposed MLODCCC architecture to develop a Food Safety Information Platform (FSIP) that provides a friendly hyperlink and chatbot interface on Facebook to identify credible food safety information. The performance and accuracy of three binary classification algorithms, namely the decision tree, logistic regression, and support vector machine algorithms, operating in different cloud computing environments were compared. The binary classification accuracy was 0.769, which indicates that the proposed approach accurately classifies using the developed FSIP.

社交网络已成为人们传播信息的主要平台，其中可能包括负面谣言。近年来，社交网络上的谣言造成了严重的问题和相当大的损失。我们试图创建一种方法来验证来自众多社交媒体消息的信息。我们提出了一种通用架构，将机器学习和开放数据与聊天机器人相结合，并基于云计算（MLODCCC），可以帮助用户在社交平台上评估信息的真实性。提出的 MLODCCC 架构由六个集成模块组成：云计算、机器学习、数据准备、开放数据、聊天机器人和智能社交应用模块。食品安全已引起全球关注。最后，我们使用提议的 MLODCCC 架构来开发食品安全信息平台 (FSIP)，该平台在 Facebook 上提供友好的超链接和聊天机器人界面，以识别可靠的食品安全信息。比较了决策树、逻辑回归和支持向量机三种二元分类算法在不同云计算环境中运行的性能和准确性。二进制分类精度为 0.769，这表明所提出的方法使用开发的 FSIP 进行了准确分类。比较了在不同云计算环境中的操作。二进制分类精度为 0.769，这表明所提出的方法使用开发的 FSIP 进行了准确分类。比较了在不同云计算环境中的操作。二进制分类精度为 0.769，这表明所提出的方法使用开发的 FSIP 进行了准确分类。