|
信息传播模型中的对等和权威压力
Peer and Authority Pressure in Information-Propagation Models |
|
| |
| 课程网址: |
http://videolectures.net/ecmlpkdd2011_terzi_peer/
|
| 主讲教师: |
Evimaria Terzi
|
| 开课单位: |
波士顿大学
|
| 开课时间: |
2011-10-03
|
| 课程语种: |
英语
|
| 中文简介: |
现有的信息扩散模型假设同伴影响是观察到的传播模式的主要原因。在本文中，我们研究了权威压力对观察到的信息级联的作用。我们通过将网络中的特征节点称为“权限”节点来模拟这种直觉。这些节点可以影响大量的对等体，而它们自身不会受到对等体的影响。我们提出了一个模型，它将每个项目与两个参数相关联，这两个参数量化了同行和权力压力对项目传播的影响。给定网络和项目的观察扩散模式，我们从数据中学习这些参数，并表征迭代对等或权限传播。我们还开发了一项随机化测试，评估我们研究结果的统计学意义，并使我们的项目表征对噪声具有鲁棒性。我们利用在线媒体和科学合作网络的实际数据进行的实验表明，这些网络存在强大的权威压力信号。
|
| 课程简介: |
Existing models of information diffusion assume that peer influence is the main reason for the observed propagation patterns. In this paper, we examine the role of authority pressure on the observed information cascades. We model this intuition by characterizing some nodes in the network as "authority" nodes. These are nodes that can influence large number of peers, while themselves cannot be influenced by peers. We propose a model that associates with every item two parameters that quantify the impact of the peer and the authority pressure on the item’s propagation. Given a network and the observed diffusion patterns of the item, we learn these parameters from the data and characterize the item as peer- or authority-propagated. We also develop a randomization test that evaluates the statistical significance of our findings and makes our item characterization robust to noise. Our experiments with real data from online media and scientific-collaboration networks indicate that there is a strong signal of authority pressure in these networks.
|
| 关 键 词: |
信息扩散; 权威压力; 信息级联
|
| 课程来源: |
视频讲座网
|
| 最后编审: |
2020-06-04：dingaq
|
| 阅读次数: |
56
|
|
图 书 馆