A re-ranking method for locality mining in graph data

Abstract

The present invention relates to a reordering method for graph data locality mining, comprising: step S1, converting an original image dataset into its corresponding tree structure; step S2, performing a hierarchical traversal on the first m layers of nodes in the tree structure based on a deep branching reordering algorithm to obtain a hierarchical traversal result; wherein m is a specified number of layers, and m is a positive integer less than the total number of layers of the tree structure; step S3, performing a depth-first traversal on the remaining nodes of the tree structure except the first m layers of nodes based on a deep branching reordering algorithm to obtain a depth-first traversal result; step S4, storing the hierarchical traversal result and the depth-first traversal result in the same result set array to obtain a reordering result of the original image dataset. With the aid of the above technical solution, the embodiments of the present application can effectively reduce the running time of the graph application and the computational overhead of the algorithm operation.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a reordering method for locality mining of graph data. Background Art

[0002] The unstructured and irregular nature of large-scale graph data, as well as the highly random and poor locality of memory access, limit the efficiency of graph computing systems. Multi-type input graph data leads to significant performance gaps due to input dependencies. Furthermore, current distributed graph computing systems increase locality by studying data organization formats, reordering graph data, or partitioning the graph.

[0003] Therefore, there is an urgent need for a re-ranking method for locality mining in graph data, so as to effectively reduce the running time of graph applications and the computational overhead of algorithm operation. Summary of the Invention

[0004] (1) Technical issues to be solved

[0005] In view of the above-mentioned shortcomings and deficiencies of the prior art, the present invention provides a reordering method for graph data locality mining, which can effectively reduce the running time of graph applications and the computational overhead of algorithm operation.

[0006] (2) Technical solution

[0007] In order to achieve the above objectives, the main technical solutions adopted by the present invention include:

[0008] In a first aspect, an embodiment of the present invention provides a reordering method for locality mining of graph data, comprising: step S1, converting an original image dataset into its corresponding tree structure; step S2, performing a hierarchical traversal on the first m layers of nodes of the tree structure based on a deep branching reordering algorithm to obtain a hierarchical traversal result; wherein m is a specified number of layers, and m is a positive integer less than the total number of layers of the tree structure; step S3, performing a depth-first traversal on the remaining nodes of the tree structure except the first m layers of nodes based on the deep branching reordering algorithm to obtain a depth-first traversal result; step S4, storing the hierarchical traversal result and the depth-first traversal result in the same result set array to obtain a reordering result of the original image dataset.

[0009] Optionally, step S2 includes the following steps: step S21, determining whether the current first visited node has been visited; wherein the current first visited node is a node in the first m layers of nodes; step S22, if it is determined that the current first visited node has not been visited, storing the current first visited node in a pre-created access record array using a breadth-first search method, and accessing the current first visited node; wherein the access record array is used to record nodes that have been visited; step S23, if it is determined that the current first visited node has been visited, updating the current first visited node to the next node of the current first visited node, and returning to step S21.

[0010] Optionally, after visiting the current first visited node, step S22 further includes the following steps: step S221, determining whether the specified layer number m has been reached; step S222, if the specified layer number m has not been reached, updating the current first visited node to the next node of the current first visited node, and returning to step S21; step S223, if the specified layer number m has been reached, executing step S3.

[0011] Optionally, step S3 includes the following steps: step S31, visiting the current second visited node; wherein the current second visited node is one of the remaining nodes; step S32, determining whether the current second visited node has an adjacent point; wherein the adjacent point is a child node adjacent to the current second visited node; step S33, if it is determined that the current second visited node has an adjacent point, updating the current second visited node according to the adjacent point; step S34, if it is determined that the current second visited node does not have an adjacent point, updating the current second visited node to the next node of the current second visited node.

[0012] Optionally, there are multiple adjacent points; updating the current second visited node according to the adjacent points includes the following steps: step S331, determining whether the current adjacent point has been visited; step S332, if it is determined that the current adjacent point has been visited, updating the current second visited node to the next adjacent point, and returning to step S32; step S333, if it is determined that the current adjacent point has not been visited, updating the current second visited node to the current adjacent point, and returning to step S31.

[0013] Optionally, after step S34, the reordering method further includes: step S35, determining whether the updated current second visited node is the last node; step S36, if it is determined that the updated current second visited node is the last node, ending the depth-first traversal.

[0014] Optionally, the reordering method further includes: step S37, if it is determined that the updated current second visited node is not the last node, then determining whether the updated current second visited node has been visited; step S38, if it is determined that the updated current second visited node has been visited, then updating the updated current second visited node to the next node of the updated current second visited node, and returning to step S35; step S39, if it is determined that the updated current second visited node has not been visited, then returning to step S31.

[0015] Optionally, the starting node set of the depth-first traversal in step S3 is the child node set of the m-th layer node.

[0016] In a second aspect, an embodiment of the present application provides a storage medium having a computer program stored thereon, which, when executed by a processor, executes the method of the first aspect or any optional implementation of the first aspect.

[0017] In a third aspect, an embodiment of the present application provides an electronic device comprising: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor and the memory communicate through the bus, and when the machine-readable instructions are executed by the processor, a method of executing the first aspect or any optional implementation method of the first aspect is executed.

[0018] (3) Beneficial effects

[0019] The beneficial effects of the present invention are:

[0020] The embodiment of the present application provides a reranking method for locality mining of graph data. Since the DBR algorithm combines the advantages of hierarchical community mining and deep community mining, the embodiment of the present application can effectively reduce the running time of the graph application and the computational overhead of the algorithm operation, thereby improving the retrieval efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0022] Figure 1 A flowchart of a reordering method for graph data locality mining provided by an embodiment of the present application is shown;

[0023] Figure 2A flowchart of the layer traversal part of a reordering method for graph data locality mining provided by an embodiment of the present application is shown;

[0024] Figure 3 A flowchart of a depth-first traversal portion of a reordering method for graph data locality mining provided by an embodiment of the present application is shown;

[0025] Figure 4a A schematic diagram of a tree structure provided by an embodiment of the present application is shown;

[0026] Figure 4b A schematic diagram of a tree structure after sorting one layer of child nodes provided by an embodiment of the present application is shown;

[0027] Figure 4c A schematic diagram showing a reordering result after sorting one layer of sub-nodes provided by an embodiment of the present application is shown;

[0028] Figure 5a A schematic diagram of a tree structure after sorting two layers of sub-nodes provided in an embodiment of the present application is shown;

[0029] Figure 5b A schematic diagram showing a reordering result after sorting two layers of sub-nodes provided by an embodiment of the present application is shown;

[0030] Figure 6a A schematic diagram showing another tree structure provided by an embodiment of the present application is shown;

[0031] Figure 6b A schematic diagram showing another tree structure after sorting sub-nodes in two layers provided in an embodiment of the present application is shown;

[0032] Figure 6c A schematic diagram showing another reordering result after sorting two layers of sub-nodes provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0033] In order to better explain the present invention and facilitate understanding, the present invention is described in detail below through specific implementation methods in conjunction with the accompanying drawings.

[0034] Currently, sophisticated graph reordering techniques can effectively reduce the running time of graph applications, but the reordering step also brings corresponding computational overhead.

[0035] Based on this, an embodiment of the present application provides a reordering method for locality mining of graph data. Since the depth-branch-reorder (DBR) algorithm combines the advantages of hierarchical community mining (or hierarchical traversal) and deep community mining (or depth-first traversal), it starts from the source node and proceeds downward in a hierarchical community mining manner. When it reaches the specified number of layers m, the child nodes of the specified layer m are traversed in a depth-first manner. After the branches of a child node are traversed in a depth-first manner, other branches are traversed until the entire graph is completely traversed, thereby effectively reducing the running time of the graph application and the computational overhead of the algorithm operation.

[0036] To better understand the above technical solutions, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments described herein. Instead, these embodiments are provided to enable a clearer and more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0037] See Figure 1 , Figure 1 FIG1 shows a flow chart of a reordering method for graph data locality mining provided by an embodiment of the present application. Figure 1 As shown, the reordering method can be performed by an image reordering device, and the specific device of the image reordering device can be set according to actual needs, and the embodiments of the present application are not limited thereto. For example, the image reordering device can be a computer or a server. Specifically, the reordering method includes:

[0038] Step S110, converting the original image data set into its corresponding tree structure;

[0039] Step S120, performing a hierarchical traversal on the first m layers of nodes of the tree structure based on a deep branch reordering algorithm to obtain a hierarchical traversal result; wherein m is a specified number of layers, and m is a positive integer less than the total number of layers of the tree structure;

[0040] Step S130, performing a depth-first traversal on the remaining nodes of the tree structure except the first m layers of nodes based on a depth-branch reordering algorithm to obtain a depth-first traversal result;

[0041] Step S140 : storing the level-first traversal result and the depth-first traversal result into the same result set array to obtain a reordering result of the original image data set.

[0042] Therefore, since the DBR algorithm combines the advantages of hierarchical community mining and deep community mining, the embodiment of the present application can effectively reduce the running time of the graph application and the computational overhead of the algorithm operation, thereby improving the retrieval efficiency.

[0043] For ease of understanding Figure 1 The specific process is as follows Figure 2 and Figure 3 to describe.

[0044] See Figure 2 , Figure 2 The flowchart of the layer traversal part of a reordering method for graph data locality mining provided by an embodiment of the present application is shown. Specifically, Figure 2 As shown, the related methods of the layer traversal part include:

[0045] Step S211: convert the original image data set into its corresponding tree structure.

[0046] Specifically, when the image compression format contained in the original image dataset is from the initial vertex to the end vertex (for example, (1,2), (1,3) and (1,4), etc.), the original image dataset can be converted into a tree structure of a specified form (specifically, parent node: leaf node set, for example, (1:2,3,4), etc.), thereby preparing for subsequent data locality mining.

[0047] Step S212: Create an access record array, a queue of nodes to be accessed, and a result set array. The access record array is used to record nodes that have been accessed; the queue of nodes to be accessed is used to record nodes that need to be accessed; and the result set array is used to store the results of the level-first traversal and depth-first traversal.

[0048] Specifically, a visit record array DBR_visited, a queue of nodes to be visited DBR_queue, and a result set array DBR_Graph may be created respectively.

[0049] It should be noted here that the visit record array DBR_visited, the queue of nodes to be visited DBR_queue and the result set array DBR_Graph can be used not only in the hierarchical traversal part, but also in the depth-first traversal part.

[0050] Step S213: Obtain the specified number of layers m input by the user.

[0051] It should be understood that the specific value of the specified number of layers m can be set according to actual needs, and the embodiments of the present application are not limited thereto.

[0052] For example, m can be 1 or 2.

[0053] Step S214: Take out the first visited node from the queue of nodes to be visited. The first visited node can be one of the first m nodes in the tree structure.

[0054] Specifically, the current first visited node Vi can be taken out from the queue of nodes to be visited, and the current first visited node Vi can be initially set as the source node.

[0055] Step S215: Determine whether the current first visited node has been visited.

[0056] If it is determined that the current first visited node has not been visited, step S216 is executed; if it is determined that the current first visited node has been visited, step S218 is executed.

[0057] Specifically, it is possible to determine whether the current first visited node has been visited by querying the visit record array. The subsequent related steps are similar and will not be described one by one.

[0058] Step S216: Use the breadth-first search method to store the current first visited node in a pre-created access record array, access the current first visited node, and update the access record array with the message that the current first visited node has been visited.

[0059] Step S217, determining whether the designated number of layers m has been reached.

[0060] If it is determined that the specified number of layers m has been reached, step S219 is executed; if the specified number of layers m has not been reached, step S218 is executed.

[0061] Step S218: Update the current first visited node to the next node of the current first visited node, and return to step S215.

[0062] Step S219: hierarchical community mining ends.

[0063] That is, the tree structure can be initialized, and hierarchical local community mining can be started from the source node of the tree structure. The visited node data can be stored in the result set array array during the traversal. After all the child node sets (for example, leaf node sets) except the source node are visited, the number of traversed layers is increased by 1. When the number of traversed layers reaches the specified number of layers m, the remaining nodes can be subjected to deep community traversal.

[0064] See Figure 3 , Figure 3 The flowchart of the depth-first traversal part of a reordering method for graph data locality mining provided by an embodiment of the present application is shown. Specifically, Figure 3 As shown, the relevant methods of the depth-first traversal part include:

[0065] Step S311: Visit the current second visited node and identify the current second visited node using a target identifier. The current second visited node is one of the remaining nodes, and the current second visited node can be taken from a queue of nodes to be visited or an updated node. The target identifier is used to indicate that the current second visited node is a node mined by the deep community, and the specific identifier of the target identifier can be set according to actual needs, and the embodiments of the present application are not limited thereto.

[0066] Step S312: Determine whether there is an adjacent node to the current second visited node, wherein the adjacent node is a child node adjacent to the current second visited node.

[0067] If it is determined that the current second visited node has an adjacent point, step S313 is executed; if it is determined that the current second visited node does not have an adjacent point, step S316 is executed.

[0068] Step S313: Determine whether the current neighboring point has been visited.

[0069] If it is determined that the current adjacent point has not been visited, step S314 is executed; if it is determined that the current adjacent point has been visited, step S315 is executed.

[0070] Step S314: Update the current second visited node as the current adjacent node, and return to step S311.

[0071] Step S315: Update the current second visited node to the next adjacent node, and return to step S312.

[0072] Step S316: Update the current second visited node to the next node of the current second visited node.

[0073] Step S317: Determine whether the updated current second visited node is the last node.

[0074] If it is determined that the updated current second visited node is not the last node, step S318 is executed; if it is determined that the current second visited node is the last node, step S319 is executed.

[0075] Step S318: Determine whether the updated current second visited node has been visited.

[0076] If it is determined that the current second visited node has been visited, step S315 is executed; if it is determined that the current second visited node has not been visited, the process returns to step S311.

[0077] Step S319, end.

[0078] That is to say, during the deep community traversal, the data of the visited nodes are also stored in the result set array until all nodes are visited, and then the algorithm ends.

[0079] It should be noted here that the starting node set of the deep community traversal is the leaf node set at the end of the last traversal at the end of the hierarchical community traversal, that is, the starting node set of the depth-first traversal is the child node set of the m-th layer node.

[0080] It should be noted here that since the traversal data in the embodiments of the present application can be cached (for example, the traversal results are cached through a result set array; for another example, the visited nodes are recorded through an access record array, etc.), a large number of repeated traversal processes can be avoided, and the traversal time can be greatly reduced by first performing hierarchical community mining and deep community mining in order.

[0081] In order to facilitate understanding of the embodiments of the present application, a description is given below using specific embodiments.

[0082] Specifically, see Figures 4a to 4c , which shows the algorithm process of DBR reordering of one layer of sub-nodes. Figure 4a As shown, there are 16 nodes in the graph, and they are sorted by one level of child nodes and converted into a tree structure. The result is as follows Figure 4b shown.

[0083] And, starting from node 1, there are five layers of data. If we start from the source node and perform depth-first traversal, the result is as follows: Figure 4c As shown in the figure, node 1 is traversed downwards one level using the hierarchical community mining method, visiting nodes 5, 8, and 12. Then, a depth-first traversal is performed downwards from the child nodes of node 1, with node 5 visiting nodes 2, 11, and 14, node 8 visiting nodes 3, 7, and 15, and node 12 visiting nodes 4, 9, 13, 10, 16, and 6, traversing the entire graph in sequence.

[0084] Also, see Figure 5a , which is a tree structure transformed by sorting two-layer child nodes. Also, starting from node 1, traversal is divided into two layers. The first two layers are traversed according to the hierarchical community. The traversal sequence is from node 1 to nodes 5, 8, and 12. Then, nodes 2, 14, 3, and 4 are visited. After the two layers are completed, a single-branch depth traversal is performed downward according to the second-layer child nodes. First, the child node 11 of node 2 is visited. Then, other branches are traversed downward according to the node order of the second layer, and nodes 7, 15, 9, 13, 10, 16, and 6 are visited in sequence until the entire graph is traversed. The result is as follows Figure 5b shown.

[0085] Also, see Figures 6a to 6c, which shows the DBR sorting algorithm process diagram for the two-layer child nodes of the 64-node graph community. Starting from node 1, the traversal is divided into two layers. The first two layers are traversed according to the hierarchical community. The traversal sequence is from node 1 to visit nodes 2, 3 and 4, and then to visit nodes 5, 6, 7, 8 and 9. After the two layers are completed, the depth traversal of the branches is carried out downward according to the second-layer child nodes. First, the child nodes of 5 are visited, and then the other branches are traversed downward according to the node order of the second layer. Figure 6b As shown, until the entire graph is traversed, the result is as follows Figure 6c shown.

[0086] It should be understood that the above data protection method is only exemplary, and those skilled in the art can make various modifications based on the above method, and the modified scheme also falls within the scope of protection of this application.

[0087] The present application provides a storage medium having a computer program stored thereon. When a processor executes the computer program, the method described in the embodiment is executed.

[0088] The present application also provides an electronic device, which includes a processor, a memory, and a computer program stored in the memory, wherein the processor executes the computer program to perform the method described in the embodiment.

[0089] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0090] The present invention is described with reference to flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions.

[0091] It should be noted that, in the claims, any reference signs placed between brackets shall not be construed as limiting the claims. The word "comprising" does not exclude the presence of components or steps not listed in the claims. The word "a" or "an" preceding a component does not exclude the presence of a plurality of such components. The invention may be implemented by means of hardware comprising several different components and by means of a suitably programmed computer. In claims enumerating several means, several of these means may be embodied by one and the same hardware. The use of the words first, second, third etc. is for convenience only and does not indicate any order. These words may be understood as part of the component name.

[0092] In addition, it should be noted that, in the description of this specification, the description of the terms "one embodiment", "some embodiments", "embodiment", "example", "specific example" or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification and the features of different embodiments or examples, unless they are contradictory.

[0093] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments after learning the basic creative concept. Therefore, the claims should be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0094] Obviously, those skilled in the art may make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. Thus, if such modifications and variations fall within the scope of the claims and their equivalents, the present invention shall also include such modifications and variations.

Claims

1. A re-ranking method for graph data locality mining, characterized in that: include: Step S1, converting the original image dataset into its corresponding tree structure; Step S2, performing a hierarchical traversal on the first m layers of nodes of the tree structure based on a deep branch reordering algorithm to obtain a hierarchical traversal result; wherein m is a specified number of layers, and m is a positive integer less than the total number of layers of the tree structure; Step S3, performing a depth-first traversal on the remaining nodes of the tree structure except the first m layers of nodes based on the depth branch reordering algorithm to obtain a depth-first traversal result; Step S4, storing the hierarchical traversal result and the depth-first traversal result into the same result set array to obtain a reordering result of the original image data set; The step S2 comprises the following steps: Step S21, determining whether the current first visited node has been visited; wherein the current first visited node is a node in the first m layers of nodes; Step S22: If it is determined that the current first visited node has not been visited, the current first visited node is stored in a pre-created access record array using a breadth-first search method, and the current first visited node is visited; wherein the access record array is used to record nodes that have been visited; Step S23: If it is determined that the current first visited node has been visited, then the current first visited node is updated to the next node of the current first visited node, and the process returns to step S21; The step S3 comprises the following steps: Step S31, accessing the current second visited node; wherein the current second visited node is one of the remaining nodes; Step S32, determining whether the current second visited node has an adjacent node; wherein the adjacent node is a child node adjacent to the current second visited node; Step S33: If it is determined that the current second visited node has the adjacent point, the current second visited node is updated according to the adjacent point; Step S34: If it is determined that the current second visited node does not have the adjacent point, the current second visited node is updated to the next node of the current second visited node.

2. The reordering method according to claim 1, wherein: After visiting the current first visited node, step S22 further includes the following steps: Step S221, determining whether the specified number of layers m has been reached; Step S222: If the specified number of layers m has not been reached, the current first visited node is updated to the next node of the current first visited node, and the process returns to step S21. Step S223: If the specified number of layers m is reached, execute step S3.

3. The reordering method according to claim 1, wherein: There are multiple adjacent points; and updating the current second visited node according to the adjacent points includes the following steps: Step S331, determining whether the current neighboring point has been visited; Step S332: If it is determined that the current adjacent node has been visited, the current second visited node is updated as the next adjacent node, and the process returns to step S32; Step S333: If it is determined that the current adjacent point has not been visited, the current second visited node is updated as the current adjacent point, and the process returns to step S31.

4. The reordering method according to claim 1, wherein: After step S34, the reordering method further includes: Step S35, determining whether the updated current second visited node is the last node; Step S36: If it is determined that the updated current second visited node is the last node, the depth-first traversal ends.

5. The reordering method according to claim 4, characterized in that: The reordering method further comprises: Step S37: If it is determined that the updated current second visited node is not the last node, determine whether the updated current second visited node has been visited; Step S38: If it is determined that the updated current second visited node has been visited, then the updated current second visited node is updated to the next node of the updated current second visited node, and the process returns to step S35; Step S39: If it is determined that the updated current second visited node has not been visited, return to step S31.

6. The reordering method according to claim 1, wherein: The starting node set of the depth-first traversal in step S3 is the child node set of the m-th layer node.

Images