论文部分内容阅读
【目的】构建网络信息内容可信度的定量测度模型,以提高虚假信息的筛除效率。【方法】基于贝叶斯推理理论,构建网络信息内容可信度的测度模型;基于贝叶斯决策理论,构建可信度测度有效性的最小错误率评估模型。【结果】基于实际数据集的实验结果表明,随着社会化媒体参与者规模增加,可信度测度的最小错误率呈下降趋势,且贝叶斯可信度测度模型总体优于传统的模糊可信度测度模型。【局限】可信度测度错误率的影响因素只关注参与者规模因素,而其他影响因素,如条件属性或可参照对象等,将需要进一步研究。【结论】基于集体智慧理论,揭示网络信息内容可信度测度的最小错误率会随着参与者规模增加而降低。
【Objective】 To construct a quantitative measure model of the credibility of network information content to improve the screening efficiency of false information. 【Method】 Based on Bayesian inference theory, a model to measure the credibility of network information content is constructed. Based on Bayesian decision theory, a model for evaluating the minimum error rate of credibility measure is constructed. 【Result】 The experimental results based on the actual data set show that with the increase of social media participants, the minimum error rate of confidence measure shows a downward trend, and Bayesian credibility measure model is better than the traditional fuzzy Reliability measurement model. Limitations The factors that influence the error rate of the reliability measure only focus on the scale factor of the participants, while other influencing factors, such as condition attributes or reference objects, will need to be further studied. 【Conclusion】 Based on the theory of collective intelligence, the minimum error rate of revealing the credibility of network information content will decrease with the increase of participants’ size.