An incremental layout method for data asset graph

Abstract

The present invention belongs to the field of data visualization technology, and specifically relates to an incremental layout method for data asset maps. It aims at the pain point problems caused by map changes during the progressive exploration and analysis process, and based on the characteristics of force-directed layout iterative optimization, designs a dynamic map incremental layout algorithm with a force-directed model adaptive optimization strategy for local coordinate prediction from the perspective of the map layout algorithm. The algorithm includes the following steps: current map point edge preprocessing, Tweak algorithm node coordinate prediction, and Tweak algorithm node coordinate prediction. While ensuring the reliability of the incremental purpose, it can make the transition effect smoother, and maintain the user's psychological map before and after exploration without affecting the overall layout of the map. Moreover, it can significantly improve the visual effect of the map after incremental layout, highlight the incremental structure, and help identify and explore important data assets, which is suitable for promotion and application.

Description

Technical Field

[0001] The present invention belongs to the technical field of data visualization, and in particular relates to an incremental layout method of a data asset map. Background Art

[0002] The data asset map is a point-edge dual heterogeneous network with data assets as nodes and the relationships between data assets as edges. It is used to describe the data assets held by one or more companies and the relationships between assets, so as to efficiently mine and integrate the value of enterprise data assets, help enterprises realize data asset operations, accelerate data monetization, and facilitate enterprises to intelligently build industry knowledge bases and quickly build data operation capabilities, thereby realizing data value-added.

[0003] In the current engineering field, data engineers can use data asset graphs to implement one-click retrieval and interactive exploration services for data assets. By matching keywords, relevant data asset sets can be retrieved, and details such as the name, creator, business semantics, and processing of data assets can be viewed. Based on the relationship between data assets, graphs can be flexibly explored to help understand and use related data assets for project development. In particular, the most commonly used and most effective means of exploring complex connections between graphs is the progressive exploration method, which is to capture graphs within the target range by artificially controlling the focus area, thereby effectively visually analyzing the target graph structure.

[0004] However, interactive exploration and analysis tasks for graphs are mostly limited to exploration in rendered complete graphs, that is, graph analysis is usually performed by scaling, dragging, filtering, etc. of graph elements. In some analysis and exploration tasks, the exploration targets that analysts are interested in are often located in some local areas with special structures rather than the entire graph, and the overall graph structure will affect the perception of the target structure. Therefore, the exploration process usually starts from a part of the local structure and refines the graph structure through a progressive exploration method to assist progressive analysis. In this process, nodes and edges in the graph will be added or deleted. At the same time, the layout of the graph is a dynamic process. If there is a lack of context-aware information, simply rendering the nodes and edges of the graph changes directly in the current graph will cause the canvas elements to change abruptly, making the entire analysis process difficult to understand and causing analysts to lose their psychological cognition of the previous graph. Summary of the invention

[0005] To address the defects and shortcomings of the prior art, the present invention aims at the pain points caused by graph changes in the progressive exploration and analysis process. Starting from the perspective of graph layout algorithm, based on the characteristics of force-directed layout iterative optimization, a dynamic graph incremental layout algorithm with a force-directed model adaptive optimization strategy for local coordinate prediction is designed, that is, a method for incremental layout of data asset graphs.

[0006] To achieve the above-mentioned purpose, a technical solution adopted by a data asset map incremental layout method of the present invention is: it comprises the following steps:

[0007] Step 1: Preprocessing of current graph points and edges

[0008] Since the incremental node coordinates in the force-directed layout depend on the origin node in the progressive exploration process, and the user's mental map needs to be kept stable during the incremental layout process, the coordinates of all nodes presented in the current graph G need to be memorized and cached;

[0009] Step 2: Tweak algorithm node coordinate prediction

[0010] The incremental origin node P is used to predict the initial coordinates of the newly added nodes. The initial coordinate processing of the incremental nodes is related to the relevant origin node P and the progressive exploration scheme adopted:

[0011] 01. Point expansion: Since there is only one origin node in the exploration method based on point expansion, the initial coordinates of the newly added nodes can be set within the domain of the origin node P. The specific domain coordinates can be determined by position offset using random numbers. After point expansion of node B, the initial coordinates of the incremental nodes will be determined to be within the domain of node B based on the point-edge connection relationship in the incremental graph.

[0012] 02. Edge expansion: Since there are two origin nodes in the exploration method based on edge expansion, it is necessary to classify the newly added nodes according to the edge relationship in the incremental graph structure M, find the close origin node P to which they belong, and then adopt the processing ideas of random numbers and domain range offset in point expansion to predict the coordinates. After expanding the edge between node A and node B, the specific origin nodes of the incremental nodes will be determined according to the edge relationship in the incremental graph and the corresponding incremental coordinates will be initialized respectively;

[0013] Step 3: Adaptive iterative optimization of force-directed layout

[0014] The repulsive force in the charge model under force-guided layout and the spring force in the spring model are used to perform dynamic adaptive iterative optimization and visual rendering of the layout. At the same time, the energy system attenuation coefficient alphaDecay and the friction factor velocityDecay are introduced to balance the optimization process of incremental layout transition effect.

[0015] For the repulsive force and spring force in the classic charge model, the algorithm's force model calculation is tuned according to the characteristics of the graph data to optimize the iterative effect of the layout, including three aspects:

[0016] a. Adaptive node edge length optimization based on node degree;

[0017] b. Adaptive charge force optimization based on node degree;

[0018] c. Introduce collision detection rules.

[0019] Preferably, the memory cache in step 1 is based on the point-edge data on the current graph, and a pair of node coordinates are mapped and cached through a hash dictionary according to the formula:

[0020] hashMap={node.guid:{node.x,node.y}for node in G.nodes} (Formula 1)

[0021] Preferably, in step 2, the incremental origin node P is used to initialize the coordinates of the newly added nodes. According to the progressive exploration method adopted, the point expansion and edge expansion are respectively initialized by the following coordinate processing:

[0022] Among them, point expansion, according to equations 2 and 3, the initial coordinates of the newly added nodes are set within the domain range of the origin node P, and the domain range is determined by random offset.

[0023] node.x=P.x+RandomPosition (Formula 2)

[0024] node.y=P.y+RandomPosition (Formula 3)

[0025] In addition, for edge expansion, according to Equations 4 and 5, the newly added nodes are classified according to their connection with the edge nodes, and the close origin node P to which they belong is found, and then the initial coordinates are determined by random offset in the domain range.

[0026] node.x=node.relatedEdgeNode.x+RandomPosition (Formula 4)

[0027] node.y=node.relatedEdgeNode.y+RandomPosition (Formula 5)

[0028] Preferably, for the repulsive force and spring force in the classical charge model described in step 3, the force model of the algorithm is optimized from the following three aspects according to the business data characteristics:

[0029] 3.1. Adaptive node edge length optimization based on node degree: When the degrees of the nodes at both ends of the edge are large, the two nodes can be regarded as two different cluster structures, and the edge length between the two nodes should be appropriately increased;

[0030] 3.2. Adaptive charge force optimization based on node degree: A node with a larger degree can be regarded as a cluster center node, and it should have a larger attraction to its neighboring nodes; at the same time, for cluster center nodes with two high degrees, the repulsion between them should be appropriately increased;

[0031] 3.3. Introduce collision detection rules: treat each node as an entity with quality characteristics, calculate and detect the entity gap, so as to avoid node overlap.

[0032] Preferably, in step 3, a dynamic adjustment strategy is designed for the energy system attenuation coefficient alphaDecay and the friction factor velocityDecay:

[0033] I. A larger decay rate is set before the incremental layout iteration starts to render the initial position of the newly added nodes faster;

[0034] II. When the rendering is completed, the attenuation coefficient is scaled, and the system is iteratively optimized using the responsive attenuation coefficient and friction factor based on the incremental map scale, thereby highlighting the transition effect of incremental structural rendering. Among them, the attenuation coefficient and friction factor are updated according to Formula 6 and Formula 7 respectively.

[0035] alphaDecay = min{0.0001, 0.00025-0.00001*M.nodes.length} (Formula 6)

[0036] velocityDecay=min{0.95,0.45+0.015*M.nodes.length} (Formula 7)

[0037] The data asset map incremental layout method described in the present invention can make the transition effect smoother while ensuring the reliability of achieving the incremental purpose, and maintain the user's psychological map before and after exploration without affecting the overall layout of the map; moreover, it can significantly improve the visual effect of the map after incremental layout, highlight the incremental structure, and facilitate the identification and exploration of important data assets, making it suitable for promotion and application. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the present invention is described in detail through the following specific implementations and drawings.

[0039] Figure 1 A schematic diagram of a flow chart of an incremental layout method of a data asset map according to the present invention;

[0040] Figure 2 Schematic diagram of initial coordinate prediction of point extension nodes based on Tweak algorithm of the present invention

[0041] Figure 3 Schematic diagram of initial coordinate prediction of edge extension nodes based on Tweak algorithm of the present invention

[0042] Figure 4 This is a diagram showing the effect of the first embodiment of the present invention;

[0043] Figure 5 This is a diagram showing the effect of the second embodiment of the present invention; DETAILED DESCRIPTION

[0044] In order to make the purpose, technical scheme and advantages of the present invention clearer, the present invention is described below by the specific embodiments shown in the accompanying drawings. However, it should be understood that these descriptions are only exemplary and are not intended to limit the scope of the present invention. In addition, in the following description, the description of well-known structures and technologies is omitted to avoid unnecessary confusion of the concept of the present invention.

[0045] It should also be noted that, in order to avoid obscuring the present invention due to unnecessary details, only structures and / or processing steps closely related to the solutions according to the present invention are shown in the accompanying drawings, while other details that are not closely related to the present invention are omitted.

[0046] The input and output of the algorithm of the present invention and the symbolic terms used are as follows: the input of the algorithm is the current graph structure G, such as Figure 2 -a; the context increment origin node P that triggers the incremental behavior of the graph during the progressive exploration process, such as Figure 2 -B node in b; the graph structure M to be added, such as Figure 2 -b is a group of b nodes around node B, and the edges connected to node B are new edges. The output is a dynamic, responsive, and transition-friendly incremental graph structure G', such as Figure 2 -c.

[0047] Among them, for the progressive exploration scheme that expands from a point, the incremental origin node is a single node, and for the progressive exploration scheme that expands from an edge, the incremental origin nodes are the two endpoints of the edge to be expanded.

[0048] like Figure 1 As shown, this specific implementation provides an incremental layout method for a data asset map, comprising the following steps:

[0049] Step 1: Preprocessing of current graph points and edges

[0050] Since the incremental node coordinates under force-directed layout depend on the origin node in the progressive exploration process, and it is necessary to maintain the stability of the user's mental map during the incremental layout process, it is necessary to memorize and cache the coordinates of all nodes presented in the current graph G.

[0051] Step 2: Tweak algorithm node coordinate prediction

[0052] The incremental origin node P is used to predict the initial coordinates of the newly added nodes. The initial coordinate processing of the incremental node is related to the relevant origin node P and the progressive exploration scheme adopted:

[0053] (1) Point expansion. Since there is only one origin node in the exploration method based on point expansion, the initial coordinates of the newly added nodes can be set within the domain of the origin node P. The specific domain coordinates can be determined by using random numbers for position offset. Figure 2 As shown, after point expansion of node B, the initial coordinates of the incremental node will be determined to be within the domain of node B according to the point-edge connection relationship in the incremental graph.

[0054] (2) Edge expansion. Since there are two origin nodes in the exploration method based on edge expansion, it is necessary to classify the newly added nodes according to the edge relationship in the incremental graph structure M, find the close origin node P to which they belong, and then adopt the processing idea of ​​random number and domain range offset in point expansion to predict the coordinates. Figure 3 As shown, after the edge between node A and node B is expanded, the specific origin node of the incremental node will be determined according to the edge relationship in the incremental graph and the corresponding incremental coordinates will be initialized respectively.

[0055] Step 3: Adaptive iterative optimization of force-directed layout

[0056] The repulsive force in the charge model under force-guided layout and the spring force in the spring model are used to perform dynamic adaptive iterative optimization and visual rendering of the layout. At the same time, the energy system attenuation coefficient alphaDecay and the friction factor velocityDecay are introduced to balance the incremental layout transition effect optimization process.

[0057] For the repulsive force and spring force in the classic charge model, the algorithm's force model calculation is tuned according to the characteristics of the graph data to optimize the iterative effect of the layout, including three directions:

[0058] a. Adaptive node edge length optimization based on node degree;

[0059] b. Adaptive charge force optimization based on node degree;

[0060] c. Introduce collision detection rules.

[0061] A dynamic adjustment strategy is designed for the energy system attenuation coefficient alphaDecay and the friction factor velocityDecay to speed up the layout and highlight the transition effect of incremental structure rendering.

[0062] Furthermore, in step 1, the current point-edge data is preprocessed, and according to the point-edge data on the current graph, a pair of node coordinates is mapped and cached through a hash dictionary according to the formula:

[0063] hashMap={node.guid:{node.x,node.y}for node in G.nodes} (Formula 1)

[0064] In formula 1, hashMap represents the node hash dictionary key-value pair object, the key is the node identifier, and the value is the node coordinate information;

[0065] node represents a node element, node.guid represents a node id identifier, node.x and node.y represent the horizontal and vertical coordinates of the node on the two-dimensional plane respectively;

[0066] G represents the current vertex-edge graph data object, and G.nodes represents all the node data in the current graph.

[0067] The specific example process of this step is as follows Figure 2 -a shows that there are 5 nodes and 4 edges in the current graph, so hash cache is performed based on the coordinate information of the nodes on the plane in the current graph for subsequent reading applications.

[0068] Furthermore, in step 2,

[0069] The node coordinate prediction of the Tweak algorithm uses the incremental origin node P to initialize the coordinates of the newly added nodes. According to the progressive exploration method adopted, the point expansion and edge expansion are respectively initialized as follows:

[0070] (1) Point expansion. According to Equations 2 and 3, the initial coordinates of the newly added nodes are set within the domain of the origin node P, and the domain is determined by random offset.

[0071] node.x=P.x+RandomPosition (Formula 2)

[0072] node.y=P.y+RandomPosition (Formula 3)

[0073] In Formula 2 and Formula 3, node.x and node.y represent the horizontal and vertical coordinates of the newly added nodes under point expansion, respectively;

[0074] Px and Py represent the horizontal and vertical coordinates of the incremental origin node respectively; RandomPosition is a random function used to generate random values ​​within the domain range, and the default domain range is [-20, 20].

[0075] (2) Edge expansion: According to Equations 4 and 5, the newly added nodes are classified according to their connection with the edge nodes, and the close origin node p to which they belong is found. Then, the initial coordinates are determined by random offset in the domain range.

[0076] node.x=node.relatedEdgeNode.x+RandomPosition (Formula 4)

[0077] node.y=node.relatedEdgeNode.y+RandomPosition (Formula 5)

[0078] In formula 4 and formula 5, node.x and node.y represent the horizontal and vertical coordinates of the newly added nodes under edge expansion respectively;

[0079] node.relatedEdgeNode indicates the close origin node to which the newly added node belongs.

[0080] node.relatedEdgeNode.x and node.relatedEdgeNode.y represent the horizontal and vertical coordinates of the origin node of the close increment respectively;

[0081] RandomPosition has the same meaning as above. It is a function used to generate random values ​​within a range. The default range is [-20, 20].

[0082] The specific example process of this step takes point expansion as an example. Figure 2 -b. The graph performs a point expansion operation on the B node in the upper left corner to obtain a newly added B node group and related edges. At this time, for the newly added B node group, according to Formula 2 and Formula 3, it can be preliminarily determined that its coordinates are within the domain of the B node; and for other non-newly added nodes, their original coordinates can be directly read from the hash cache in step 1. Thus, the preliminary graph node coordinates are set. For edge expansion, such as Figure 3 As shown, after expanding the edge between node A and node B, node group a and node group b will be obtained. According to formula 4 and formula 5, it can be preliminarily determined that node group a is within the scope of node A and node group b is within the scope of node B.

[0083] Furthermore, in step 3,

[0084] The force model calculation of the algorithm is tuned according to the characteristics of the graph data to optimize the iterative effect of the layout, including the following steps:

[0085] Step 3.1, adaptive node edge length optimization based on node degree: When the degrees of the nodes at both ends of the edge are large, the two nodes can be regarded as two different cluster structures, and the edge length between the two nodes should be appropriately increased.

[0086] like Figure 2 -c, in the incremental graph, both node A and node B are high-degree nodes and can be regarded as two cluster structures. Therefore, the length of their connecting edges can be increased to optimize the visual effect.

[0087] Step 3.2, adaptive charge force optimization based on node degree: A node with a larger degree can be regarded as a cluster center node, which should have a larger attraction on its neighboring nodes; at the same time, for cluster center nodes with two high degrees, the repulsive force between them should be appropriately increased.

[0088] like Figure 2 -c, similarly to step 3.1, increasing the repulsion of cluster center nodes A and B on neighboring nodes can highlight the cluster structure characteristics.

[0089] Step 3.3, introduce collision detection rules: treat each node as an entity with quality characteristics, calculate and detect the entity gap to avoid node overlap.

[0090] like Figure 2 -c, when there are too many cluster edge nodes in the cluster structure, collision detection based on the node radius can eliminate node overlap, thus avoiding Figure 2 - The occlusion phenomenon of the b node group during the initial layout in b.

[0091] The dynamic adjustment strategy is designed for the energy system attenuation coefficient alphaDecay and the friction factor velocityDecay, including the following steps:

[0092] Step 3.4: A larger decay rate is set before the incremental layout iteration starts to render the initial positions of the newly added nodes faster.

[0093] Step 3.5: After the rendering is completed, the attenuation coefficient is scaled, and the system is iteratively optimized using the responsive attenuation coefficient and friction factor based on the incremental map scale, thereby highlighting the transition effect of the incremental structural rendering. The attenuation coefficient and friction factor are updated according to Formula 6 and Formula 7, respectively.

[0094] alphaDecay = min{0.0001, 0.00025-0.00001*M.nodes.length} (Formula 6)

[0095] velocityDecay=min{0.95,0.45+0.015*M.nodes.length} (Formula 7)

[0096] In formula six and seven:

[0097] alphaDecay and velocityDecay represent the decay coefficient and friction factor to be updated respectively;

[0098] M represents the incremental graph data related to point expansion or edge expansion, and M.nodes.length represents the number of nodes in the incremental graph data.

[0099] The effect diagram of the embodiment related to this algorithm is as follows Figure 4 and Figure 5 As shown, it can be seen that the data asset map incremental layout method provided by the present invention can make the transition effect smoother while ensuring the reliability of achieving the incremental purpose, and maintain the user's psychological map before and after exploration without affecting the overall layout of the map; moreover, it can significantly improve the visual effect of the map after incremental layout, highlight the incremental structure, facilitate the identification and exploration of important data assets, and is suitable for promotion and application.

[0100] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above and that the present invention can be implemented in other specific forms without departing from the spirit or essential features of the present invention. Therefore, the embodiments should be considered exemplary and non-restrictive in all respects, and the scope of the present invention is defined by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims be included in the present invention.

[0101] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A method for incremental layout of a data asset map, characterized in that: It contains the following steps: Step 1: Preprocessing of current graph points and edges Since the incremental node coordinates in the force-directed layout depend on the origin node in the progressive exploration process, and the user's mental map needs to be kept stable during the incremental layout process, the coordinates of all nodes presented in the current graph G need to be memorized and cached; Step 2: Tweak algorithm node coordinate prediction Use the incremental origin node P to predict the initial coordinates of the newly added nodes; The initial coordinate processing of the incremental node is related to the relevant origin node P and the progressive exploration scheme adopted:

01. Point expansion: Since there is only one origin node in the exploration method based on point expansion, the initial coordinates of the newly added nodes can be set within the domain of the origin node P. The specific domain coordinates can be determined by position offset using random numbers. After point expansion of node B, the initial coordinates of the incremental nodes will be determined to be within the domain of node B based on the point-edge connection relationship in the incremental graph.

02. Edge expansion: Since there are two origin nodes in the exploration method based on edge expansion, it is necessary to classify the newly added nodes according to the edge relationship in the incremental graph structure M, find the close origin node P to which they belong, and then adopt the processing ideas of random numbers and domain range offset in point expansion to predict the coordinates. After expanding the edge between node A and node B, the specific origin nodes of the incremental nodes will be determined according to the edge relationship in the incremental graph and the corresponding incremental coordinates will be initialized respectively; Step 3: Adaptive iterative optimization of force-directed layout The repulsive force in the charge model under force-guided layout and the spring force in the spring model are used to perform dynamic adaptive iterative optimization and visual rendering of the layout. At the same time, the energy system attenuation coefficient alphaDecay and the friction factor velocityDecay are introduced to balance the optimization process of incremental layout transition effect. For the repulsive force and spring force in the classic charge model, the algorithm's force model calculation is tuned according to the characteristics of the graph data to optimize the iterative effect of the layout, including three aspects: a. Adaptive node edge length optimization based on node degree; b. Adaptive charge force optimization based on node degree; c. Introduce collision detection rules; For the repulsive force and spring force in the classical charge model described in step 3, the force model of the algorithm is optimized from the following three aspects according to the business data characteristics: 3.

1. Adaptive node edge length optimization based on node degree: When the degrees of the nodes at both ends of the edge are large, the two nodes can be regarded as two different cluster structures, and the edge length between the two nodes should be appropriately increased; 3.

2. Adaptive charge force optimization based on node degree: A node with a larger degree can be regarded as a cluster center node, and it should have a larger attraction to its neighboring nodes; at the same time, for cluster center nodes with two high degrees, the repulsion between them should be appropriately increased; 3.

3. Introduce collision detection rules: treat each node as an entity with quality characteristics, calculate and detect the entity gap, so as to avoid node overlap.

2. The incremental layout method of a data asset map according to claim 1 is characterized by: The memory cache in step 1 is based on the point-edge data on the current graph, and a pair of node coordinates are mapped and cached through a hash dictionary according to the formula: hashMap={node.guid:{node.x,node.y}for node in G.nodes} Formula 1 In formula 1, hashMap represents the node hash dictionary key-value pair object, the key is the node identifier, and the value is the node coordinate information; node represents the node element, node.guid represents the node id identifier, node.x and node.y represent the horizontal and vertical coordinates of the node on the two-dimensional plane respectively; G represents the current vertex-edge graph data object, and G.nodes represents all the node data in the current graph.

3. The incremental layout method of a data asset map according to claim 1 is characterized in that: In the step 2, the incremental origin node P is used to initialize the coordinates of the newly added nodes. According to the progressive exploration method adopted, the point expansion and edge expansion are respectively initialized as follows: Among them, point expansion, according to equations 2 and 3, the initial coordinates of the newly added nodes are set within the domain range of the origin node P, and the domain range is determined by random offset. no d e .x=P.x+RandomPosition n Formula 2 node.y=P.y+RandomPosition Formula 3 In Formula 2 and Formula 3, node.x and node.y represent the horizontal and vertical coordinates of the newly added nodes under point expansion, respectively; Px and Py represent the horizontal and vertical coordinates of the incremental origin node respectively; RandomPosition is a random function used to generate random values ​​within the domain range, and the default domain range is [-20,20] node.x=node.relatedEdgeNode.x+RandomPosition Formula 4 node.y=node.relatedEdgeNode.y+RandomPosition Formula 5 In formula 4 and formula 5, node.x and node.y represent the horizontal and vertical coordinates of the newly added nodes under edge expansion respectively; node.relatedEdgeNode indicates the close origin node to which the newly added node belongs. node.relatedEdgeNode.x and node.relatedEdgeNode.y represent the horizontal and vertical coordinates of the origin node of the close increment respectively; RandomPosition has the same meaning as above. It is a function used to generate random values ​​within a range. The default range is [-20,20] In addition, for edge expansion, according to Equations 4 and 5, the newly added nodes are classified according to their connection with the edge nodes, and the close origin node P to which they belong is found, and then the initial coordinates are determined by random offset in the domain range.

4. The incremental layout method of a data asset map according to claim 1 is characterized in that: In step 3, a dynamic adjustment strategy is designed for the energy system attenuation coefficient alphaDecay and the friction factor velocityDecay: I. A larger decay rate is set before the incremental layout iteration starts to render the initial position of the newly added nodes faster; II. When the rendering is completed, the attenuation coefficient is scaled, and the system is iteratively optimized using the responsive attenuation coefficient and friction factor based on the incremental map scale, thereby highlighting the transition effect of incremental structural rendering. alphaDecay=min{0.0001,0.00025-0.00001*M.nodes.length} Formula 6 velocityDecay=min{0.95,0.45+0.015*M.nodes.length} Formula 7 In formula six and seven: alphaDecay and velocityDecay represent the decay coefficient and friction factor to be updated respectively; M represents the incremental graph data related to point expansion or edge expansion, and M.nodes.length represents the number of nodes in the incremental graph data; Among them, the attenuation coefficient and friction factor are updated according to Formula 6 and Formula 7 respectively.

Images