Analysis of visual analysis technology

【 Research status of technology at home and abroad 】

Visualization of spatial information network based on multi-layer features

Luca Rossi Et al. First discussed the visualization strategy of multi-layer network , They believe that although some scholars gradually focus on the research and analysis of multilayer networks , However, there is no professional visualization method for this content , The solution to the current problem can only rely on traditional methods .

De Domennico M Based on self-developed visualization tools MuxViz Finished the project MULTIPLEX, However , At present, the platform can only realize concept demonstration , Interactive visual analysis is not yet supported , Its essence is an experimental platform for multi-layer network visualization . In the field of social networks , Applying the idea of traditional network diagram to multi-layer network layout is an effective way to show the characteristics of network hierarchy ,M. Magnani Et al. Used the social network graph to realize Social network hierarchy topology Visualization of structure , As shown in the figure .

Social network diagram of multi-layer network

As shown in the figure above , Include for 5 Multi layer network social relationship network diagram of layer connection relationship . In the left figure, when a single connection relationship is shown , The social network diagram can clearly show the nodes Clustering , But when considering multiple types of connections , On the right , Because of the diversity and complexity of connection relations, clustering is not obvious 、 There are some problems, such as edge overlap, relationship disorder, and loss of density location information . The above problems are a drawback of this idea , That is, the structural information between relevant layers will be completely lost .

In order to solve the problem of information loss in multi-layer network , A method called expanding node attributes is proposed . In this method, edges and nodes belonging to different layers are colored , Hierarchical information is distinguished by different colors , Generate a social network map of expanded information , Although the visualization results obtained by this method are more beautiful , But in fact, the amount of information added is relatively small , Although some dense areas can be observed , However, it is almost impossible to understand their relationship with multi-layer network structure .

W.Longabaugh And so on Based on a specific matrix, determine the position order of nodes and edges, and establish a sequential social relationship network diagram , This method can not only obtain the overview information of network node degree distribution through sequential social relationship network diagram , It can also investigate neighbor nodes 、 distance 、 Edge type and isomorphism （ Or mismatch ） Etc , However , This method can not directly display the network topology .

Visualization of spatial information network based on dynamic features

From the perspective of spatiotemporal behavior , Spatial information network has remarkable dynamic characteristics . The dynamics of spatial information network includes the dynamics of time dimension , It also includes the dynamics of spatial dimensions , And its time dimension and space dimension are dynamic at the same time , In terms of time , It refers to the change of spatial information network topology caused by time change ; From the space dimension , It refers to the change of topological structure of spatial information network caused by spatial displacement .

Fengkunchen et al. In order to achieve spatiotemporal consistency and multi focus in a single-layer framework + Visualization of context , A time-varying graph drawing method is proposed , This method uses the existing algorithm to realize the preliminary layout , Then based on F+C Construct and form continuous time-varying graphs , So as to achieve the purpose of time-varying graph visualization , The method is shown in the figure .

The focus of + Text time-varying visual rendering method

Rutgers University James Abello In view of the problem that tracking the change of time in large-scale dynamic networks will cause the loss of some network node information , Put forward a kind of Visual analysis method based on modular interest description , The method is based on adjacent structure information 、 A large number of nodes / Edge attributes and evolution information , So as to realize the corresponding Visualization .

University of skatt, Germany F. Beck People think The main challenge of dynamic network visualization is how easy to read 、 It is easy to measure and design effective algorithms to present the evolution relationship of different entities , They systematically classify and mark published methods , This paper systematically expounds the development status of dynamic network visualization , As shown in the figure .

Hierarchical classification dynamic network visualization method

Eindhoven University of Technology S. van den Elzen For the imperfection of the existing methods of visual analysis of dynamic networks , A method of reducing time snapshot to point is proposed , Thus, the network is reduced from high-dimensional to two-dimensional for visualization and interaction , The method mainly includes discretization 、 To quantify 、 Standardization 、 Dimensionality reduction interaction 5 part , This method is effective for some artificial and real networks , The process steps are shown in the figure below .

“ snapshot - spot ” Dynamic network visualization technology

Chenhui Li Et al. Studied large-scale network data visualization based on modules , Through design and development ModuleGraph System , Based on this system The dynamic structure of social networks and spatial networks .

At home , Liu Zhen of Hangzhou University of Electronic Science and technology and others from dynamic network data model and visualization standard 、 This paper systematically summarizes the current development status of dynamic network visualization and visual analysis from the aspects of basic dynamic graph visualization methods and multi task oriented visual analysis , It is one of the few research literatures on dynamic network visualization in China . In my previous research , Focus on the network layout 、 This paper summarizes the typical technologies and methods of dynamic network visualization from the perspectives of attribute visualization and user interaction .

【 summary 】

The research and development of dynamic network visualization technology is still ongoing , However, task oriented visualization and visual analysis should be the mainstream of future development , In particular, it is worth advocating and encouraging to carry out research on relevant aspects in combination with research objects with distinctive characteristics . Because the concept of spatial information network was put forward relatively late , At present, the core issues mainly focused on research are still in the relatively elementary stage 、 It is also the core of dynamic network topology modeling , There is no published literature on its dynamic network visualization and visual analysis .

The development of spatial information network visualization and visual analysis platform integrating multi-layer dynamic features . Spatial information network is the core element of information support in the future battlefield , The visual platform based on multi feature fusion shows and analyzes the macro structure of spatial information network , Query based rendering of overall or local network functions and performance , It is helpful for managers of spatial information network to obtain information enlightenment from it , So as to grasp the overall operation of the spatial information network , So as to ensure that the spatial information network plays its role .

Improve the usability of spatial information network visualization and visual analysis platform . Because ordinary users usually do not have the knowledge of relevant professional fields , Make it difficult to be familiar with 、 Understand and apply visualization and visual analysis platforms , For planners engaged in spatial information networks 、 Designer 、 A certain basic knowledge is necessary for the builders , This limits the popularization and wide use of the visualization platform , So we should improve the usability of the visualization platform .

Network visualization can make full use of human visual perception system , take network data Show it graphically , On the one hand, it is used to understand the internal structure of the network , On the other hand, it helps to tap the useful value inside the network , Its essence is the subclass of information visualization . At present , There is a general view that data visualization is scientific computing visualization 、 Information visualization and visual analysis . Data visualization technology has covered people's life 、 Work 、 All aspects of study and life , Especially the wide spread of the Internet 、 The widespread use of mobile terminals , While people enjoy the dividends of scientific and Technological Development , Also based on visualization technology, it really realizes “ A picture is worth a thousand words ”.

Network visualization can display network data in a formal way , And further visual analysis , Facing the multidimensional characteristics of spatial information network , Visualization technology provides a new idea and method for analyzing its characteristics . Like other visualizations , There are also some basic functional and performance requirements for multi feature visualization for spatial information network . If you want to meet the basic aesthetic criteria throughout the visualization process （Aesthetics criterion）, Always be able to maintain the user's intention map （Mental map） etc. , Its basic aesthetic principles are mainly used to guide the elimination of visual confusion in the network , Including the minimum number of edge intersections 、 The adjacent points should be close to each other 、 Straight edge 、 The side layout should keep the overall balance 、 The node layout should form a hierarchy 、 Layout criteria consistency and network edge traceability ; Intention map mainly refers to whether it is multi-layer visualization , Or dynamic visualization , Users do not need to repeatedly adjust their new understanding and perception of the visualized network , Only by meeting the requirements of these functions and performance can we achieve the ideal visualization effect .

The visualization process of cyberspace information can be summarized as : First, get the key relevant data information of cyberspace , So that the computer can read and recognize ; Then, the computer will map the digital form according to a certain structure algorithm ( The reason is ) Information is processed by a series of data to form a certain form of digital map ; Finally, the digital map is transformed into a displayable map graph or chart 、 The report form is called number in cartography / Mode conversion , In essence, it is the process of spatial information visualization .

（ Part of the data comes from the Internet , If any, please contact to delete ）