最近老板给推荐了个CMU的ORA，个人觉得真的很合适做sna，dynamic network analysis，link analysis，multi-agent system之类的任务，综合了pajek gephi ucinet citespace等等的功能，且给出了很完整的user guide。（竟然可以topic analysis，我惊了）当然也可以配合CMU CASOS tools，UCINET, KeyPlayer, and Analyst's Notebook等进行额外分析。总之如果你的数据量不大（agents<2000) 且不想买高阶版时非常好用，尤其是当你的ucinet等突然打不开的时候（没错就是我最近），墙裂推荐。当然只要熟悉统计分析思路，是不限于使用软件的。附CASOS: Center for Computational Analysis of Social and Organizational Systems (http://www.casos.cs.cmu.edu/index.html)

软件大概200多M（记不清楚了哈哈），有student版本和付费版本，不同版本能处理的节点及关系数量的能力不同。以下对于一些可能常用到的动作和功能作个记录：

Preface：

ORA uses a Java interface for ease of use, and a C++ computational backend. The most current edition of ORA software, Version 3, is available in two versions. ORA-LITE is available on the CASOS website: http://casos.cs.cmu.edu/projects/ora/. It contains hundreds of social network, dynamic network metrics, trail metrics, procedures for grouping nodes, identifying local patterns, comparing and contrasting networks, groups, and individuals from a dynamic meta-network perspective.

GETTING STARTED

1. File format 能支持的数据导入格式

• DyNetML (.xml) -- stored as default

• DL - Data Language (.dl) 一定要储存为plain text
  大概是这样

  
  DL.png

• UCINET (.##h and .##d)

• CSV - Comma Seperated Values (.csv)

  
  Again note that the cell A1 is blank. This is common in all network files created in this manner. .png

• Text Files (.txt)

• Pajek (.net)

• GraphML (.graphml)

2. Quickstart

User interface 主要两个界面：editor 以及visualizer
Versions：Lite-学生教育版本，max 2000个nodes
下载地址：Go to http://casos.cs.cmu.edu/projects/ora/software.php

2.1 Main Interface

1） Editor GUI image.png

2）Meta-network
包含noteset和network，可点击查看
Nodeset: nodes and attibutes.

Nodeset class: Nodesets (and hence the contained nodes) can be classified as "Agent", "Organization", "Knowledge", "Resource", "Belief", "Event", "Task", "Location", "Role", "Action" or "Unknown".Multiple nodesets in a metanetwork can be assigned to the same nodeset class. Many predefined ORA analyses use the nodeset class to determine inputs.

Node
Attribute

Networks:

• Binary, weighted;
• Self-looped

Dynamic network

image.png

2.2 Import file 基本操作

select File > Data Import Wizard
可以导入矩阵 或node的attributes表 可在选项中选择

node关系表
attribute表
第三个弹出窗口
Match

记得导入data 和 attributes，导入后可查看info和editor

对于dynamic network， 可以merge meganetwork也可以直接导入file

2.3 Edit data and attributes

To edit a nodeset, select the nodeset in the Metanetwork Manager pane. Then click the Editor tab. 注：在进行新操作时，最好是使用copy后的文件，原文件一旦改动则无法undo
对nodes：可进行过滤选择，可合并节点，移除节点，建新节点
对attributes：
对links：可在meta network-->network-->editor-->display options中修改matrixview还是其他view。display options中也同样可以进行排序。同样在Editor的Convert Links中一些名词解释：Binarize ->Change all link values to 1 (non-links remain 0) 表示有无联系，Remove self-loops (diagonal) -》Remove all links that point from a node to itself. 还有各种remove links的操作

2.4 Visualizations

在Visualizer的界面下：

image.png
常用：
hide links/nodes/isolates； Node appearance 改size和color，link appearance改size，View改字体，Node Selector改颜色 可见度 appearance cleaning 等等
Tools > Node Locator : Selecting a node from the dialog box will reveal the location of the node in Visualizer.

左侧的configure bar可以把自己常用的功能放在一起

2.5 Reports

有4个最重要的reports
1）Key Entity : Identifies key entities and groups who, by virtue of their position in the network, are critical to its operation.
2）Standard Social Network Analysis : Calculates the standard network analysis measures.
3）View Role and Sphere of Influence : For each individual, identifies the set of actors, groups, knowledge, resources, etc. that influence, and are influenced, by that actor.

For example

4）Locate Groups : Identifies the groups present in the network using various grouping algorithms. Algorithms work on unimodal or bimodal parts of the meta-network.

这些reports可以分析的对象包括：
1）Actors；
2）Region; Once the key locations are discovered agents connected to that location can be analyzed.

3. Time. 如果有time data

General contents

• 1, 2 or n Mode Data

  
  
  3 mode
• Aggregation (更多指时空的组合，这个数据我还不晓得怎么输入）
• Correlation

• Cube analysis: 可以做mental model分析
  Three key communicative dimensions.
  Intensity/consensus : Weighted degree or number of graphs of individual's mental models or number of reports the tie is presented in.
  Conductivity : in-degree*out-degree or betweenness
  Density : degree

  
  Concept Embedding
• Aggregation 比如把多个放在一个time period里
• 指数随机图模型. 使用ERGM模型时要记住的一点是，一个网络被建模为一个单一的多变量观测，这限制了它们在用于一段时间内的建模网络。此外，只适用于二进制网络。
• Hamming Distance. 两个网络（同一维度）的Hamming距离是一个链路存在于一个网络中但不存在于另一个网络中的次数。然后，可以通过除以可能的链接数来规范化该计数。
• Monte Carlo Methods. The method is useful for obtaining numerical solutions to problems which are too complicated to solve analytically. The most common application of the Monte Carlo method is Monte Carlo integration.
• Moran-I and Geary-C.

References 整理

1）Where to Learn More: Barnes, J. A. (1954). Class and committees in a Norwegian island parish. Human relations, 7(1), 39-58.
2）Scott, J., (2012). Social Networks (3rd edition), Sage.
3）Wasserman, S. & Faust, K. (1994). Social Network Analysis: Methods and Applications (Vol. 8). Cambridge University Press.
4）Carley, K. M., (2003). Dynamic Network Analysis. In Breiger, R., Carley, K. M., & Pattison, P. (Eds.) Dynamic Social Network Modeling and Analysis: Workshop Summary and Papers, (pp. 133-145). Washington, DC: National Academies Press.
5）Carley, K. M., (2002). Smart Agents and Organizations of the Future. In Lievrouw, L. & Livingstone, S. (Eds.) The Handbook of New Media, (pp. 206-220). Thousand Oaks, CA: Sage.

6. Davis, A. et al. (1941). Deep South. Chicago: University of Chicago Press.
7. Breiger R. (1974). The duality of persons and groups. Social Forces, 53, 181-190.
8. G. Robins, P. Pattison, Y. Kalish, D. Lusher, An introduction to exponential random graph (p*) models for social networks, Social Networks, Volume 29, Issue 2, May 2007, Pages 173191.