Динамический сетевой анализ
Динамический сетевой анализ (DNA) - научная область на стадии становления, которая объединяет традиционный социальный сетевой анализ (SNA), анализ связи (LA), социальное моделирование и системы мультиагента (MAS) в пределах сетевой науки и сетевой теории.
NetLogo как средство создания динамической социограммы
Пакет работает в разных операционных системах. Соберем множество картинок в единый гиф:
convert -delay 150 ?.png ??.png ???.png dm.gif
Можно было использовать
convert *.png dm.gif
но нам нужно не перепутать последовательность картинок
Переведем gif в виде.mpg
convert dm.gif dm.mpg
Audacity - программа для записи и редактирования аудиофайлов. Мы видим продолжительность нашего видео, где человечки бегают по экрану. Добавим к этим человечкам поясняющий рассказ и сохраняем этот рассказ как mp3 звуковой файл.
подробнее - https://ru.wikipedia.org/wiki/FFmpeg нам понадобится только строчка объединяющая видео и звук
ffmpeg -i dm.mpg -i dmv.mp3 -vcodec copy -acodec copy dm.avi
И все. Получило видео со звуком. Положили на YouTube
Еще интересный вариант - пропустить использование ImageMagic и собрать фильм из картинок напрямую только при помощи ffmpeg
История с сочинением
Дополнения (только FFmpeg)
- FFMPEG An Intermediate Guide/image sequence - https://en.wikibooks.org/wiki/FFMPEG_An_Intermediate_Guide/image_sequence
ABM для динамического сетевого анализа
1.De Caux R. et al. Dynamic, small-world social network generation through local agent interactions // Complexity. 2014. Vol. 19, № 6. P. 44–53.
To model agent relationships in agent-based models, it is often necessary to incorporate a social network whose topology is commonly assumed to be “small-world.” This is potentially problematic, as the classification is broad and covers a wide-range of network statistics. Furthermore, real networks are often dynamic, in that edges and nodes can appear or disappear, and spatial, in that connections are influenced by an agent's position within a particular social space. These properties are difficult to achieve in current network formation tools. We have, therefore, developed a novel social network formation model, that creates and dynamically adjusts small-world networks using local spatial interactions, while maintaining tunable global network statistics from across the broad space of possible small-world networks. It is, therefore, a useful tool for multiagent simulations and diffusion processes, particularly those in which agents and edges die or are constrained in their movement within some social space. We also show, using a simple epidemiological diffusion model, that a range of networks can all satisfy the small-world criterion, but behave quite differently. This demonstrates that it is problematic to generalize results across the whole space of small-world networks.
The second issue concerns the behavior of nodes. NA has to reconcile two different and sometimes apparently irreconcilable aspects: the need of generating a network through appropriate form/severe rules and the need of embedding in such rules a stylized version of meaningful social and economic behaviors. It seems rules that govern the formation of links that the literature in traditional network theory to date employs are usually very straightforward and often lack empirical foundations (see Roth 2007). It follows that, through these models, we can only generate theoretical networks are essentially abstract in nature.
Примеры использования в образовании:
- McFarland, Daniel A. 2004. “Resistance as a Social Drama – A Study of Change-Oriented Encounters.” American Journal of Sociology 109 (6): 1249–1318.
- McFarland, Daniel A. 2001. “Student Resistance: How the Formal and Informal Organization of Classrooms Facilitate Everyday Forms of Student Defiance.” American Journal of Sociology 107 (3): 612-78.
- McFarland, Daniel A. and Skye Bender-deMoll. 2003. “Classroom Structuration: A Study of Network Stabilization.” Working paper, Stanford University.
- The Art and Science of Dynamic Network Visualization - http://www.cmu.edu/joss/content/articles/volume7/deMollMcFarland/
- Moody J., Mcfarl D., Bender-demoll S. Dynamic Network Visualization 1.
Т.е. они снимают данные с поведения в классе и переводят эти данные в форму динамического графа - интересно, что связи не только образуются, но и обрываются
Below, we present the network movie generated by the chaining of Kamada-Kawai generated graphs (see figure 14). In the figure, there are various colors and shapes that have ascribed meaning: yellow squares = teachers; grey nodes = students; circles = females; squares = males; darker nodes = older students. The arcs are depicted in three colors: black = task; blue = sociable; and red = conflict (any negative evaluation). Arrows show arc directionality. Other specifics: We use a circle as a starting configuration upon which to reach an optimum in the first slice; we then chain subsequent graphs by using the prior layout as a starting configuration; and we ignore isolates. Other settings use default parameters.
The routine of group work is basically characterized by dyadic task and social interactions that persist in multiple clusters. Not all persons engage in these groups, and a couple students sit quietly by themselves. The group work routine breaks down as social activity increases within the groups and the teacher emits broadcast sanctions in an effort to redirect student attention back on task (~16 minutes). The task breaks down again at the end of class, but this time because the adults make closing announcements. In such a fashion, the movie shows teachers involved in the task engaging their students as they monitor interaction. When students become too social, a teacher usually arrives, disperses the group, and then reforms it via task interactions. Hence, there is a “dance” here, and it entails relatively bounded groups of individuals that free-associate over tasks and drift into social affairs, while teachers refocus affairs by indirect means of broadcasts and by direct means of directed speech.
Innovations in data collection (Choudhury and Pentland 2004; Choudhury 2004; Motoyoshi et al. 2002), the application of network methodologies to new fields (Barabasi 2002; Dunne et al. 2002; Powell et al. 2005), and the use of simulations as data sources (Ashish et al. 1995; Banks and Carley 1996; Carley 1999; Jin et al. 2001) are yielding data with high sampling rates, and occasionally time data that identifies when changes arise in individual relations
However, some of these problems are less prevalent for observation-based data, computer data, and other automated data collection techniques. As advances in technology and methodology increase our ability to collect larger and larger volumes of raw data with accurate information about the timing of events, it is still crucial that we explicitly consider how the raw event data relates to the abstract network we want to analyze or visualize. We cannot expect visualization techniques to give stable, consistent, useful results unless the definition of the social space we are trying to visualize is itself stable and consistent. This means that when working to draw a network, we should have a clear sense of the functional relationship between the data and the underlying "network" that we want to picture.
We are mostly concerned here with visualizing social network data sets that change over time. These networks tend to be small to medium sized networks (<1,000 nodes), with directional ties that are fairly sparse, and which often entail cycles (unlike trees), bridging connections, isolates, and disconnected components.
Thin slices give the response to the question "show me the network at time t.”
The field of network visualization is not particularly unified or comprehensively theorized. It might be described as a set of useful but ad hoc methodologies specific to various problem domains. Although recent increases in computing power and media technology have led to considerable advances over the original sketches made on paper, the size and complexity of the datasets under consideration has also increased, thereby creating dramatic challenges for data presentation (Pajek; see Batagelj 1998, 2004). While we do not have a polished theory or general classification scheme to present, it does seem worthwhile to consider what our aims are in creating pictures of networks, and how they might relate to some of the issues we have described above.
Social Cartography - Социальная картография
As with any visual representation, visualization of network data serves multiple purposes. Often, the framework used to conceptualize a problem is closely related to the shared metaphor that is used to describe and communicate about it. Enhancing the power and flexibility of visualization techniques can increase our intuitive understanding and ability to communicate abstractly about networks in general. At the same time, visualization can provide a means for understanding specific networks by presenting their data in media that are sufficiently visually accessible to give purchase for intuition, and sufficiently accurate to allow substantive comparison and argument (Tufte 1983).
Ideally, network visualization might serve a role similar to geographical cartography—whose strengths and limitations are well understood and techniques so widely used that they are transparent, allowing us to concentrate on the relationships revealed rather than the tools used to present them. However, like geographical visualization or any charting procedure, network visualization has the power to distort as well as to inform.
The past two decades of network research have been characterized by increasing quantitative sophistication. As methods become more specialized and technical, there is greater and greater need for inductive, exploratory tools that enable researchers to more readily intuit and understand social network form, content, and process. Network visualization is frequently heralded as such a tool.
Dynamic Network Visualization
Moody J., Mcfarl D., Bender-demoll S. Dynamic Network Visualization // American Journal of Sociology. 2005. Т. 110. № 4. С. 1206 – 1241.
Social network research has made extensive use of visualization since Moreno first introduced the sociogram (Brandes, Raab, andWagner 2001; Freeman 2000a, 2000b; Freeman, Webster, and Kirke 1998). Actors are usually represented as points, and relations among actors are represented by lines, with relational direction indicated by arrows. Early sociograms were drawn by hand (Whyte 1943; Coleman 1961), and the layout was determined by the artistic and analytic eye of the author. Such early graphs were usually simple, having few relations per person or a clear hierarchical structure. The purpose of dynamic network visualizations is to help augment theoretical intuition provided by summary statistics and standard static visualizations. Until now, visualizations of network change have tended to take two forms. The first common visualization approach plots network summary statistics as line graphs over time. For example, Doreian et al. (1996) present change in reciprocity and transitivity for the Newcomb data (see also Gould 2002). However, such summary statistics provide information on a single dimension of a network’s structure. For example, one might find that a network reaches a given equilibrium transitivity level, but since transitivity is a single average for the graph as a whole, we cannot know if this—in itself—means the graph is now relationally stable. The second common visualization approach is to examine separate images of the network at each point in time. Unfortunately, such images are often difficult to interpret, since it is impossible to identify the sequence linking node position in one frame to position in the next.
THEORETICAL IMPLICATIONS OF NETWORK DYNAMICS A standing critique of social network research has focused on a “structural bias” that implicitly denies much of the dynamic nature of social relations (Emirbayer 1997; Emirbayer and Goodwin 1994). For some types of relations (such as conversations that occur in real time), one could argue that the networks are largely artificial constructions built by aggregating dead past events. The network “structure” as such only emerges from this aggregation. While we do not think this argument should be pushed too far, it raises important questions about how the temporal embeddedness of relations defines a dynamic social space. While discussions of meaning and temporal abstraction in themselves are
Using Visualizations to Explore Network Dynamics
Chu K.-H., Wipfli H., Valente T.W. Using Visualizations to Explore Network Dynamics // Journal of Social Structure. 2013. Vol. 14.
Network graphs, in particular, are a useful tool that can help model relations, summarize data, and represent abstract concepts in a clear and intuitive way. The value of using network graphs to visualize data has been applied in different fields, and has helped improve our knowledge of disease spread (Christakis and Fowler 2010), international telecommunications (Barnett 2001), ecological systems (Stefano, Alonso and Pascual 2008), social networks (Moody and White 2003), health studies (Valente 2010), among many others. Social network analysis (SNA) often uses a sociogram to clarify different concepts. Sociograms are network graphs in which nodes represent actors and ties represent relationships between them.
The sociogram is a powerful analysis tool, helping researchers identify points of interest such as clusters (Newman and Girvan 2004), boundary spanners - это такие медиаторы и объединители сообществ (Levina and Vaast 2005), central and peripheral layers (Borgatti and Everett 2000), and other structural properties that otherwise would not be obvious in numeric data (e.g. an adjacency matrix). Today, there are online communities that form around every conceivable topic, so it is no surprise that SNA has become popular for online social network research.
Growing in parallel with SNA is the availability of different software tools. Since Moreno’s (1932) small hand drawn examples, modern computer technology can now create networks with 10’s of millions of users (Mislove, Massimiliano, Gummadi, Drushel and Bhattacharjee 2007). The development of SNA software has aided SNA research, as increased computing power has enabled fast complex calculations and supported large-scale network analyses (e.g. visualizing million node networks). Researchers can conduct studies based on network structures, and many of the calculations and measurements are made immediately available.
Given the power of SNA, there are still gaps that have only recently started to be addressed. For example, sociograms are, by nature, static representations. They are snapshots of a network in a single moment in time, giving no hints as to how or why the network developed into a particular structure, or what it could potentially become. More studies into the evolution of social networks would be beneficial for research, especially in online communities, which can grow at tremendous speeds. Within social network analysis, researchers have recognized the value in emphasizing important features of social structures, the similarities and differences in positions occupied by the actors, searching for groups and positions, and understanding the patterns that link sets of actors (Freeman 2000). Freeman noted the strength of the sociogram as a method of exploration, and also predicted that as computing processing power and storage continued increasing, there would be a growth in graph-generating software. While browser-based Java applets and VRML tools did not become as popular as he predicted, there are many standalone network analysis software packages that have been developed.
Over a decade later, we are still learning how to visualize social networks. Recently, Correa and Ma (2011) identified 4 types of social network visualization: structural, semantic, temporal, and statistical.
Packages such as SocialAction have been developed that better integrate classical methods in exploratory data analysis and statistics with SNA visualizations (Perer and Shneiderman 2008). SocialAction was used to find different levels of partisanship in US senators by interactively filtering the data on various statistical measures.
Along with other visualization-focused technologies (e.g. Gephi, ORA, NetLogo), new tools are being developed that enable network graphs to be integrated into different types of research. The study of dynamic networks greatly benefits from visualizations that can illustrate ideas and concepts not immediately visible in a static sociogram. In fact, “The ability to see data clearly creates a capacity for building intuition that is unsurpassed by summary statistics” (Moody, McFarland and Bender-deMoll 2005). Moody and others’ research emphasizes how the ability to see data can be superior to summary statistics, and illustrates the need to visualize how networks develop and change over time. Additionally, they lay the foundation of how dynamic network visualizations should be presented (e.g. differentiating between discrete and continuous time), and recommend visualization and analysis be interactive. These theoretical ideas were developed in parallel with SoNIA, a software package for visualizing dynamic network data (Bender-deMoll and McFarland 2006).
SNA metrics and concepts can also be useful in helping to understand actions within online communities. Mislove and others (2007) studied the communities in Flickr, YouTube, LiveJournal, and Orkut, representing some of the largest online communities at the time (over 11.3 million users and 328 million links). They found the communities exhibited a strongly connected core of high-degree nodes, surrounded by many small clusters of low-degree nodes. There was also a high degree of reciprocity in directed user links, leading to a strong correlation between user in-degree and out-degree, with a power law degree distribution. A more recent study by Kairam and colleagues (2012) investigated the Ning network to compare diffusion and nondiffusion membership growth, and found that clustering promotes diffusion growth, although it is more likely to lead to smaller eventual groups.
Online communities can provide large amounts of accurate data, usually based on web server log files. Popular web servers such as Apache3 provide logs with rich information, including user-identifiers (e.g. IP address), precise timestamps, mouse clicks, text typed into any field, etc. Browser cookies allow servers to remember users across multiple web surfing sessions. These resources offer researchers rich and accurate information in how people communicate and interact in any online medium. It is an ideal source of data for dynamic network studies.
Studies of online communities have expanded as the number of online communities continues to increase. In particular, social networking sites have become a valuable source of data for different types of studies. User actions can be easily and accurately collected over extended periods of time, for large numbers of websites. Each site can have distinct characteristics, distinguished by the purpose for which people joined, the type of interactions that occur, or the media available for use by members (Chu and Suthers 2013). These variations afford many opportunities to study the similarities and differences between online communities (Hether, Murphy and Valente in-press).
There are many popular software tools, such as Pajek and UCINet, which have been used for various types of SNA research. However, they do not provide support for dynamic visualization and are less beneficial for longitudinal analyses. This paper does not attempt to comprehensively compare different software packages that can visualize dynamic networks. There are several available off-the-shelf tools that support some form of longitudinal analysis, including DyNet5, GUESS6, SoNIA (Bender-deMoll and McFarland 2006), TeCFlow (Gloor and Zhao 2004), and JUNG (O'Madadhain, Fisher, White and Boey 2003). In this paper, we use Gephi, an open source multipurpose platform for network visualization.
Gephi is an interactive visualization and exploration platform for all kinds of networks and complex systems, dynamic and hierarchical graphs. Some of its most attractive features are its support of many different native graph formats, real-time interactive features, and easy-to-use interface. Most importantly, it has many supporting features built for dynamic network analysis that incorporate functions such as live filtering, a combination of static and dynamic metrics, a multitude of layouts, and a timeline component that can generate various longitudinal reports.
Our web crawler provided information for GLOBALink members that included their home country, the countries of their referring members, and the date when their account was approved. This was stored in a comma-separated values (CSV) file. Custom parsing scripts were run against the CSV file to create two Gephi-readable files: a dynamic nodes list, and a dynamic ties list. The dynamic nodes are a list of all nodes in the network (i.e. countries), with a time interval based on when a member from a country first joins. This was achieved by parsing through all members, grouped by their home countries, and selecting the one with the earliest account approval date. The layout of the network is created using Gephi’s ForceAtlas2 (FA2) algorithm, which created an easy-to-interpret graph. The FA2 algorithm is continuous and optimized for speed (suitable for dynamic graphs, as it will efficiently update in real time) and offers various options to help fine-tune the results. More details can be found on Gephi’s website.
Modularity is a measurement of how well a network can be divided into smaller clusters, or modules (Newman and Girvan 2004), and is useful in finding community structure (Newman 2006). High modularity indicates that a network has a higher rate of intra-module edges relative to inter-module ones.
The referral network appears to create a majority of sub-groups that are divided geographically. These regions align closely with regional WHO offices that provided the platform for regional consultations and the formation of regional negotiating positions in between the global negotiating sessions held between 2000 and 2003. Other clusters appear to encompass the sphere of influence or language affinity between particular countries (e.g. Dark blue countries associate with French language/influence).
Data visualizations have been useful in many scientific fields, including social network analysis. The evolution of the sociogram, especially in conjunction with modern computers and software, has helped advance SNA studies, providing researchers with a better understanding of network characteristics such as structural patterns, positions that actors occupy, or where clusters emerge. As we push progress in studying longitudinal networks, the tools and methods used must also continue to evolve. Dynamic network analysis began as a collection of SNA-derived extensions, but its application in longitudinal network studies (e.g. terrorist cells, diffusion models) has demonstrated its necessity and utility. Research in how to conduct dynamic network studies, and consequently, the visualizations used, will help provide us with the necessary tools to better understand the unique nature of how networks are formed, patterns of evolution, and the metrics used to study them.
- Borgatti, S. and M. Everett (2000). "Models of Core/Periphery Structures." Social Networks 21(4): 375-395.
- Brandes, U. and C. Pich (2012). "Explorative Visualization of Citation Patterns in Social Network Research." Journal of Social Structure
- Carley, K. (2003). Dynamic Network Analysis. Dynamic Social Network Modeling and Analysis: Workshop Summary and Papers. K.
- Kairam, S., D. J. Wang and J. Leskovec (2012). The Life and Death ofOnline Groups: Predicting Group Growth and Longevity.
- Levina, N. and E. Vaast (2005). "The Emergence of Boundary Spanning Competence in Practice: Implications for Implementation and Use of Information Systems." MIS Quarterly
29(2): 335-363. Liben-Nowell, D. and J.
- Newman, M. E. (2006). "Modularity and community structure in networks." Proceedings of the National Academy of Sciences 103(23): 8577-8582.
- Newman, M. E. and M. Girvan (2004). "Finding and evaluating community structure in networks." Physical review E 69(2).