Graph grouping method, apparatus, device, and storage medium that allow for positional deviation

Abstract

The present application provides a graph grouping method, apparatus, device, and storage medium that tolerate position deviation. The graph grouping method that tolerates position deviation includes: obtaining a set of graphs waiting for grouping that includes a plurality of graphs waiting for grouping; constructing a graph comparison matrix including a plurality of elements for representing the similarity between different graphs based on the set of graphs; selecting a target row that meets a preset condition from the graph comparison matrix and determining the elements located in the target row based on a preset policy; grouping a plurality of graphs waiting for grouping in the set of graphs based on the elements of the target row and the preset policy, and outputting a grouping result. According to the embodiments of the present application, an overall optimal grouping result can be realized in the graph grouping process, and the problem that the grouping result also changes due to different data input orders in the graph grouping process, further affecting the processing steps after graph grouping, can be avoided.

Description

[Technical Field]

[0001] This application relates to the semiconductor technology field, and more particularly to a graph grouping method, apparatus, device, and storage medium that allow for positional deviation. [Background technology]

[0002] In the manufacturing and detection processes of chips, positional deviations are common due to the movement accuracy of the manufacturing detector and the randomness inherent in the location of defects themselves. Therefore, graph grouping techniques that tolerate positional deviations are particularly important, especially in the application of design-to-database (D2DB) detection techniques. For example, as shown in Figure 1, the central graph is the graph awaiting grouping (i.e., the graph shown by the solid line frame in Figure 1), but due to various errors, the graph actually obtained is the graph shown by the dotted line frame in Figure 1, and as a result, they cannot be precisely divided into one group. Based on this, in the graph grouping process, it is not possible to consider only a single graph, but it is necessary to consider the set of all graphs within the positional deviation range.

[0003] Conventional graph grouping methods that allow for positional deviations include methods that add and compare points one by one. As shown in Figure 2, this method involves first comparing Graph 1 and Graph 2, that is, comparing two sets of graphs and determining whether identical graph elements exist between the two sets. If identical graphs exist between the two sets of graphs, Graph 1 and Graph 2 are grouped together. If no identical graphs exist between Graph 1 and Graph 2, Graph 1 and Graph 2 are not grouped (isolated). Then, Graph 3 is added and compared with the result of the previous step. If Graph 1 and Graph 2 are grouped together, Graph 3 is compared based on the common graph of Graph 1 and Graph 2. If Graph 1 and Graph 2 are not grouped together, Graph 1 and Graph 2 are compared separately based on Graph 3, and the result of this comparison is taken as the grouping result for Graphs 1, 2, and 3. Similarly, Graph 4 is added and compared with the grouping result of the previous three graphs (1, 2, and 3), and all graphs (including Graphs 1 to Graph n) are grouped together. Here, graphs 1, 2, 3, 4...n are a set that covers the graph itself and all graphs within a certain range of positional deviations, where n is a positive integer.

[0004] However, adopting the method of adding and comparing each of the above points presents the following problems.

[0005] Firstly, the method of adding and comparing points one by one can only yield locally optimal grouping results. For example, if there is a common graph between the first graph and the second graph, they can always be divided into one group. However, if the first graph and the Xth graph have more common graphs, but the common graphs between the first graph and the Xth graph are far outside the acceptable range set between the common graphs between the first graph and the second graph, then the Xth graph cannot be divided into one group with the first graph. In other words, the order in which points are added and compared one by one does not yield a globally optimal grouping result, but only the optimal grouping result in the current order.

[0006] Secondly, differences in the ordering of the input data affect the grouping results. As shown in the previous example, if the Xth-order graph appears as the first or second-order comparison, the grouping results will change. In other words, if the input data remains the same but the data order changes, different grouping results will occur, which has a significant impact on subsequent processing steps that depend heavily on the reliability and stability of the grouping results.

[0007] Therefore, achieving an overall optimal grouping result in the graph grouping process, and avoiding the fact that differences in the data input order in the graph grouping process will alter the grouping result and further affect subsequent processing steps, are technical challenges that should be addressed by those skilled in the art. [Overview of the Initiative]

[0008] Embodiments of the present invention provide a graph grouping method, apparatus, device, and computer-readable storage medium that allow for positional deviations, thereby achieving an overall optimal grouping result in the graph grouping process and avoiding the problem that differences in the data input order in the graph grouping process can alter the grouping result and further affect post-graph grouping processing steps.

[0009] According to a first aspect, an embodiment of the present application provides a graph grouping method that allows for positional deviation, the method being Obtaining a set of graphs waiting to be grouped, which includes graphs waiting to be grouped in multiple groups, Based on a set of graphs, construct a graph comparison matrix containing multiple elements to represent the similarity between different graphs, Based on a pre-defined policy, select a target row from the graph comparison matrix that satisfies pre-defined conditions, and determine the elements located in the target row. This includes grouping multiple graphs awaiting grouping in a graph set based on the elements of the target row and a pre-configured policy, and outputting the grouping results.

[0010] Selectively, the pre-configured policies include one of the following: greedy grouping policy, orphaned grouping policy, rich grouping policy, or weak rich grouping policy.

[0011] Selectively, if the pre-configured policy is a greedy grouping policy, then based on the pre-configured policy, it is possible to select target rows that satisfy the pre-configured conditions from the graph comparison matrix and determine the elements located in the target rows. Based on the greedy grouping policy, the row with the most non-zero elements is selected from the graph comparison matrix and determined as the target row. This includes determining the elements located in the target row based on the target row.

[0012] Selectively choosing the row with the most non-zero elements from the graph comparison matrix and determining it as the target row, based on a greedy grouping policy, Determining the number of non-zero elements in each row of the graph comparison matrix, The process involves ordering each row in a graph comparison matrix based on the number of non-zero elements, and obtaining the ordering result. This includes selecting the row with the most non-zero elements from the graph comparison matrix based on the ordering results and determining it as the target row.

[0013] Selectively, if a pre-configured policy includes an isolated grouping policy, then based on the pre-configured policy, it is possible to select target rows from the graph comparison matrix that satisfy the pre-configured conditions and determine the elements located in the target rows. Based on the isolated grouping policy, the row with the fewest non-zero elements is selected from the graph comparison matrix and determined as the target row. including determining an element located at a target row based on the target row.

[0014] Optionally, determining a target row by selecting, based on an isolation grouping policy, a row with the fewest non-zero elements from a graph comparison matrix includes: determining the number of non-zero elements included in each row in the graph comparison matrix; ordering each row in the graph comparison matrix based on the number of non-zero elements to obtain an ordering result; and determining a target row by selecting, based on the ordering result, a row with the fewest non-zero elements from the graph comparison matrix.

[0015] Optionally, when a preset policy includes a rich grouping policy, determining a target row that satisfies a preset condition from a graph comparison matrix based on the preset policy and determining an element located at the target row includes: determining a target row by selecting, based on the rich grouping policy, a row with the most non-zero elements from the graph comparison matrix; and determining an element located at the target row based on the target row.

[0016] Optionally, determining a target row by selecting, based on a rich grouping policy, a row with the most non-zero elements from a graph comparison matrix includes: determining the number of non-zero elements included in each row in the graph comparison matrix; ordering each row in the graph comparison matrix based on the number of non-zero elements to obtain an ordering result; and determining a target row by selecting, based on the ordering result, a row with the most non-zero elements from the graph comparison matrix.

[0017] Selectively, if the pre-configured policy is a greedy grouping policy, then based on the elements of the target row and the pre-configured policy, multiple graphs awaiting grouping in the graph set can be grouped and the grouping results output. Based on the elements of the target row, determine the non-zero elements in the target row, Based on non-zero elements, select rows in the graph comparison matrix other than the target row that contain the target element whose value is equal to the non-zero element, This includes, based on rows containing target elements and target rows, creating a single graph for the rows containing target elements and the graphs corresponding to non-zero elements in the target rows, and outputting the grouping results.

[0018] A graph grouping method that selectively allows such positional deviations is: This involves determining whether all graphs in a set of graphs that are awaiting grouping will be grouped, If the decision is negative, then the graphs that are not yet grouped and awaiting grouping will be determined from the set of graphs. This further includes performing grouping according to a greedy grouping policy based on a graph of items awaiting grouping.

[0019] Selectively, if the pre-configured policy is an orphaned grouping policy, the following is possible: Grouping multiple graphs awaiting grouping in the graph set based on the elements of the target row and the pre-configured policy, and outputting the grouping results. Based on the elements of the target row and the orphaned grouping policy, the graphs corresponding to the non-zero elements included in the target row will be made into the same graph, This includes selecting ungrouped graphs from a set of graphs based on graphs included in the same graph, grouping them, and outputting the grouping results until all graphs in the set that are awaiting grouping are grouped.

[0020] Selectively, if the pre-configured policy is a rich grouping policy, then based on the elements of the target row and the pre-configured policy, multiple graphs awaiting grouping in the graph set can be grouped and the grouping results output. Based on the elements of the target row and the rich grouping policy, the graphs corresponding to non-zero elements included in the target row will be made into the same graph, This includes selecting ungrouped graphs from a set of graphs based on graphs included in the same graph, grouping them, and outputting the grouping results until all graphs in the set that are awaiting grouping are grouped.

[0021] According to a second aspect, an embodiment of the present application provides a graph grouping device that allows for positional deviation, the device, A module for obtaining a set of graphs waiting to be grouped, which includes multiple graphs waiting to be grouped, A construction module for building a graph comparison matrix containing multiple elements to represent the similarity between different graphs, based on a set of graphs, A selection module for selecting target rows that satisfy predefined conditions from a graph comparison matrix based on a predefined policy, and for determining the elements located in the target rows, It includes a grouping module for grouping multiple graphs awaiting grouping in a graph set based on the elements of the target row and a pre-configured policy, and for outputting the grouping results.

[0022] Selectively, the pre-configured policies include one of the following: greedy grouping policy, orphaned grouping policy, rich grouping policy, or weak rich grouping policy.

[0023] Selectively, if the pre-configured policy is a greedy grouping policy, the selection module will: Based on the greedy grouping policy, a first selection unit is used to select the row with the most non-zero elements from the graph comparison matrix and determine it as the target row, Includes a first decision unit for determining the elements located in the target row, based on the target row.

[0024] Selectively, the first selected unit is, Determining the number of non-zero elements in each row of the graph comparison matrix, The process involves ordering each row in a graph comparison matrix based on the number of non-zero elements, and obtaining the ordering result. It is used to select the row with the most non-zero elements from the graph comparison matrix based on the ordering result and determine it as the target row.

[0025] Selectively, if the pre-configured policy includes an orphaned grouping policy, the selection module will: Based on the isolated grouping policy, a second selection unit is used to select the row with the fewest non-zero elements from the graph comparison matrix and determine it as the target row, It includes a second decision unit for determining the elements located in the target row, based on the target row.

[0026] Selectively, the second selected unit is, Determining the number of non-zero elements in each row of the graph comparison matrix, The process involves ordering each row in a graph comparison matrix based on the number of non-zero elements, and obtaining the ordering result. It is used to select the row with the fewest non-zero elements from the graph comparison matrix based on the ordering result and determine it as the target row.

[0027] Selectively, if the pre-configured policy includes a rich grouping policy, the selection module will: Based on the rich grouping policy, a third selection unit is used to select the row with the most non-zero elements from the graph comparison matrix and determine it as the target row, It includes a third decision unit that determines the elements located in the target row based on the target row.

[0028] Selectively, the third selected unit is, Determining the number of non-zero elements in each row of the graph comparison matrix, The process involves ordering each row in a graph comparison matrix based on the number of non-zero elements, and obtaining the ordering result. It is used to select the row with the most non-zero elements from the graph comparison matrix based on the ordering result and determine it as the target row.

[0029] Selectively, if the pre-configured policy is a greedy grouping policy, the grouping module will: Based on the elements of the target row, determine the non-zero elements in the target row, Based on non-zero elements, select rows in the graph comparison matrix other than the target row that contain the target element whose value is equal to the non-zero element, It is used to group rows containing target elements and target rows into the same graph based on the target row itself, and to output the grouping result.

[0030] A graph grouping device that selectively allows such positional deviations further, This involves determining whether all graphs in a set of graphs that are awaiting grouping will be grouped, If the decision is negative, then the graphs that are not yet grouped and awaiting grouping will be determined from the set of graphs. This is used to perform grouping according to a greedy grouping policy based on a graph of data awaiting grouping.

[0031] Selectively, if the pre-configured policy is an orphaned grouping policy, the grouping module will: Based on the elements of the target row and the orphaned grouping policy, the graphs corresponding to the non-zero elements included in the target row will be made into the same graph, This is used to select ungrouped graphs from a set of graphs based on the graphs contained within the same graph, group them, and output the grouping results until all graphs in the set that are awaiting grouping are grouped.

[0032] Selectively, if the pre-configured policy is a rich grouping policy, the grouping module will: Based on the elements of the target row and the rich grouping policy, the graphs corresponding to non-zero elements included in the target row will be made into the same graph, This is used to select ungrouped graphs from a set of graphs based on the graphs contained within the same graph, group them, and output the grouping results until all graphs in the set that are awaiting grouping are grouped.

[0033] According to a third aspect, an embodiment of the present application provides a graph grouping device that allows positional deviation, the device comprising a processor and memory for storing computer program instructions. When the processor executes a computer program instruction, it implements a graph grouping method that allows for the positional deviation described in the first embodiment.

[0034] According to a fourth aspect, an embodiment of the present application provides a computer-readable storage medium in which computer program instructions are stored, and when the computer program instructions are executed by a processor, a graph grouping method that allows positional deviations as described in any of the first aspects is realized.

[0035] According to the graph grouping method, apparatus, device, and computer-readable storage medium that allow positional deviation in the embodiment of the present application, when grouping graphs, first, a set of graphs awaiting grouping, including multiple graphs awaiting grouping, is obtained; second, a graph comparison matrix is ​​constructed based on the set of graphs, including multiple elements to represent the similarity between different graphs; and third, a target row that satisfies pre-set conditions is selected from the graph comparison matrix based on a pre-set policy, and the elements located in the target row are determined; and finally, the multiple graphs awaiting grouping in the set of graphs are grouped based on the elements of the target row and the pre-set policy, and the grouping result is output. This makes it possible to achieve an overall optimal grouping result in the graph grouping process and avoid the problem in the graph grouping process where the grouping result changes due to different data input orders, which further affects processing steps after graph grouping. [Brief explanation of the drawing]

[0036] To more clearly describe specific embodiments of the present application or technical concepts in the prior art, the following briefly introduces the drawings that may be used in the description of specific embodiments or the prior art. Clearly, the drawings in the following description are some embodiments of the present application, and those skilled in the art can obtain other drawings based on these without any creative effort. [Figure 1] This is a schematic diagram of a graph awaiting grouping related to the conventional technology. [Figure 2] This is a schematic flowchart for grouping graphs based on the method of adding and comparing each point related to the conventional technology. [Figure 3a] This is a schematic flowchart of a graph grouping method that allows for positional deviation, according to one embodiment of the present invention. [Figure 3b] This is a schematic flowchart of a method for grouping graphs awaiting grouping based on a greedy grouping policy, according to one embodiment of the present invention. [Figure 3c] This is a schematic flowchart of a method for grouping graphs awaiting grouping based on a greedy grouping policy, according to one embodiment of the present invention. [Figure 3d] This is a schematic flowchart of a method for grouping graphs awaiting grouping based on an isolated grouping policy, according to one embodiment of the present invention. [Figure 3e] This is a schematic flowchart of a method for grouping graphs awaiting grouping based on a rich grouping policy, according to one embodiment of the present invention. [Figure 4] This is a schematic diagram of a graph grouping device that allows for positional deviation according to one embodiment of the present invention. [Figure 5] This is a schematic diagram of a graph grouping device that allows for positional deviation according to one embodiment of the present invention. [Modes for carrying out the invention]

[0037] The following describes in detail the features and exemplary embodiments of the present application in each aspect, and further details of the present application with reference to the drawings and specific embodiments in order to clarify the purpose, technical proposal and advantages of the present application. It should be understood that the specific embodiments described herein are for interpretive purposes only and do not limit the present application. Those skilled in the art may implement the present application without needing some of these specific details. The following description of embodiments is merely to provide examples of the present application for better understanding.

[0038] In this text, relational terms such as “first” and “second” are merely used to distinguish one entity or operation from another, and do not necessarily require or imply that there is any actual relationship or order between these entities or operations. Furthermore, the terms “includes,” “equip,” “have,” or any other variation thereof are intended to cover non-exclusive inclusion, and a process, method, article, or device that includes a set of elements includes not only those elements but also other elements not explicitly listed, or further elements specific to such a process, method, article, or device. Unless further restrictions are imposed, elements limited by the phrases “includes…,” “equipment…,” or “have…” do not preclude the existence of other identical elements in a process, method, article, or device that includes the element.

[0039] In the manufacturing and detection processes of chips, positional deviations are common due to the movement accuracy of the manufacturing detector and the randomness inherent in the location of defects themselves. Therefore, graph grouping techniques that tolerate positional deviations are particularly important, especially in the application of design-to-database (D2DB) detection techniques. For example, as shown in Figure 1, the central graph is the graph awaiting grouping (i.e., the graph shown by the solid line frame in Figure 1), but due to various errors, the graph actually obtained is the graph shown by the dotted line frame in Figure 1, and as a result, they cannot be precisely divided into one group. Based on this, in the graph grouping process, it is not possible to consider only a single graph, but it is necessary to consider the set of all graphs within the positional deviation range.

[0040] Conventional graph grouping methods that allow for positional deviations include methods that add and compare points one by one, but adopting this method has the following problems.

[0041] Firstly, the comparison method of adding points one by one can only yield locally optimal grouping results. For example, if there is a common graph between the first graph and the second graph, they will always be divided into one group. However, if the first graph and the Xth graph have more common graphs, but the common graphs between the first graph and the Xth graph are far outside the ultra-high tolerance range between the common graphs of the first and second graphs, then the Xth graph will not be grouped with the first graph. In other words, the comparison order of adding points one by one will only yield the optimal grouping result in the current order, not the globally optimal grouping result.

[0042] Secondly, differences in the ordering of the input data affect the grouping results. As shown in the previous example, if the Xth-order graph appears as the first or second-order comparison, the grouping results will change. In other words, if the input data remains the same but the data order changes, different grouping results will occur, which significantly impacts subsequent processing steps that are highly dependent on the grouping results and have reliability and stability.

[0043] To solve the problems of the prior art, embodiments of the present application provide a graph grouping method, apparatus, device, and computer-readable storage medium that allow for positional deviation. The graph grouping method that allows for positional deviation according to embodiments of the present application will be described first.

[0044] Figure 3a is a schematic flowchart of a graph grouping method that allows for positional deviation according to one embodiment of the present invention. As shown in Figure 3a, the method includes the following steps S301, S302, S303, and S304.

[0045] In step S301, a set of graphs waiting to be grouped is obtained, which includes multiple graphs waiting to be grouped.

[0046] In step S302, a graph comparison matrix is ​​constructed based on the set of graphs, containing multiple elements to represent the similarity between different graphs.

[0047] In the embodiments of the present application, in actual applications, since most of the graphs waiting for grouping may not be related, in this case, if the conventional method is adopted to store for some of the graphs waiting for grouping, a large amount of storage space may be wasted. To avoid this problem, the present application adopts a graph comparison matrix method to represent the comparison information of all graphs waiting for grouping, thereby saving storage space.

[0048] As shown in Table 1, in the embodiments of the present application, assuming that a plurality of graphs waiting for grouping in a graph set include Graph 1, Graph 2, Graph 3,..., Graph n, based on Graph 1, Graph 2, Graph 3,..., Graph n, an n×n graph comparison matrix shown in Table 1 can be constructed. The graph comparison matrix is M 12 , M 12 , M 13 , M 14 , M 21 , M 22 , M 23 , M 24 , …, M nn and includes elements such as.

[0049]

Table 1

[0050] Here, M 11 may be used to represent the similarity between Graph 1 and Graph 1, M 12 may be used to represent the similarity between Graph 1 and Graph 2, M 13 may be used to represent the similarity between Graph 1 and Graph 3, M 1n may be used to represent the similarity between Graph 1 and Graph n, M 21 may be used to represent the similarity between Graph 2 and Graph 1, M 22 may be used to represent the similarity between Graph 2 and Graph 2, M23 This may be used to represent the similarity between Graph 2 and Graph 3, M 2n This is used to represent the similarity between graph 2 and graph n, and M nn This can also be used to represent the similarity between graph n and graph n, but the explanation is omitted here.

[0051] In step S303, based on a pre-configured policy, a target row that satisfies a pre-configured condition is selected from the graph comparison matrix, and the elements located in the target row are determined.

[0052] In the embodiments of the present invention, the pre-configured policy may include one of the following: a greedy grouping policy, an isolated grouping policy, a rich grouping policy, or a weak rich grouping policy.

[0053] In one selectable embodiment, taking the example that the pre-configured policy is a greedy grouping policy, a target row that satisfies the pre-configured conditions is selected from the graph comparison matrix based on the greedy grouping policy, and the elements located in the target row are determined. The greedy grouping policy is primarily reflected in the row selection of the graph comparison matrix.

[0054] Specifically, based on the greedy grouping policy, the row with the most non-zero elements can be selected from the graph comparison matrix and designated as the target row. Based on the target row, the elements located in the target row can then be determined. This ensures that the subsequent graph grouping results are more representative because the selected row has the most non-zero elements.

[0055] If you selectively choose the row with the most non-zero elements from the graph comparison matrix and determine it as the target row, based on a greedy grouping policy, (1) The step of determining the number of non-zero elements in each row of the graph comparison matrix, (2) A step of ordering each row in the graph comparison matrix based on the number of non-zero elements and obtaining the ordering result, (3) Based on the ordering results, the step of selecting the row with the most non-zero elements from the graph comparison matrix and determining it as the target row.

[0056] For example, using the information shown in Table 1 above, if we assume that the number of non-zero elements in the first row of the graph comparison matrix is ​​2, the number of non-zero elements in the second row is 4, the number of non-zero elements in the i-th row is A, ..., the number of non-zero elements in the n-th row is 1, and A is the maximum number of non-zero elements, then we can order each row in the graph comparison matrix based on the number of non-zero elements to obtain the ordering result, and based on the ordering result, we can select the row with the most non-zero elements from the graph comparison matrix and determine it as the target row.

[0057] As shown in Table 2, in the embodiments of the present invention, the row with the most non-zero elements (i.e., the i-th row, the grayed-out row in Table 2) can be selected from the graph comparison matrix and determined as the target row.

[0058] [Table 2]

[0059] In one selectable embodiment, taking the example that the pre-configured policy is an orphaned grouping policy, a target row that satisfies the pre-configured conditions is selected from the graph comparison matrix based on the orphaned grouping policy, and the elements located in the target row are determined.

[0060] Here, the isolated grouping policy, in the case of precise grouping, can prioritize grouping smaller sets of graphs into one group, thereby ensuring the uniqueness of the graphs. It can be applied to several special cases (non-periodic regions, e.g., metal / poly layers, or regions with small area), and can ensure that all sets of graphs within each grouped group have identical graphs within a range of positional deviation.

[0061] Specifically, based on the isolated grouping policy, the row with the fewest non-zero elements can be selected from the graph comparison matrix and determined as the target row, and based on the target row, the elements located in the target row can be determined.

[0062] In the embodiment of the present application, if, based on the isolated grouping policy, the row with the fewest non-zero elements is selected from the graph comparison matrix and determined as the target row, (1) The step of determining the number of non-zero elements in each row of the graph comparison matrix, (2) A step of ordering each row in the graph comparison matrix based on the number of non-zero elements and obtaining the ordering result, (3) Based on the ordering results, the step of selecting the row with the fewest non-zero elements from the graph comparison matrix and determining it as the target row.

[0063] In one selectable embodiment, taking the example that the pre-configured policy is a rich grouping policy, a target row that satisfies the pre-configured conditions is selected from the graph comparison matrix based on the rich grouping policy, and the elements located in the target row are determined.

[0064] Here, the rich grouping policy requires that all graph sets within each group after grouping have identical graphs within the range of positional deviation. Furthermore, to ensure accurate grouping, as many graphs as possible can be grouped together, thereby giving the selected graphs a more representative meaning and providing reference for OPC and defect detection.

[0065] Specifically, based on the rich grouping policy, the row with the most non-zero elements can be selected from the graph comparison matrix and determined as the target row, and based on the target row, the elements located in the target row can be determined.

[0066] If you want to selectively select the row with the most non-zero elements from the graph comparison matrix and determine it as the target row based on a rich grouping policy, first determine the number of non-zero elements in each row of the graph comparison matrix, then order each row of the graph comparison matrix based on the number of non-zero elements to obtain the ordering result, and finally, based on the ordering result, select the row with the most non-zero elements from the graph comparison matrix and determine it as the target row.

[0067] In step S304, based on the elements of the target row and a pre-configured policy, multiple graphs awaiting grouping in the graph set are grouped, and the grouping results are output.

[0068] In one selectable embodiment, if the pre-configured policy is a greedy grouping policy, grouping multiple graphs awaiting grouping in a graph set based on the elements of the target row and the pre-configured policy, and outputting the grouping result, includes determining non-zero elements in the target row based on the elements of the target row, selecting rows from rows other than the target row in the graph comparison matrix that contain target elements whose numerical values ​​are equal to the non-zero elements based on the non-zero elements, and making the graphs corresponding to the rows containing target elements and the non-zero elements in the target row the same graph based on the rows containing target elements and the target row, and outputting the grouping result.

[0069] For example, based on the elements of the target row, if there are non-zero elements in the target row, then M i1 M ii M in Includes and M i1 The value of is a, and M ii The value of is b, and M in Assuming the value of is c, select rows containing target elements such as a, b, or c from the rows other than the target row in the graph comparison matrix. Then, based on the rows containing target elements and the target row, make the graphs corresponding to the non-zero elements in the rows containing target elements and the target row the same graph, and output the grouping result.

[0070] For example, as shown in Table 3 below, if rows other than the target row in the graph comparison matrix contain target elements such as a, b, or c, then M 11 and M 21 Assuming we select rows containing elements such as M, we can further determine the non-zero elements contained in these two rows, for example, if the non-zero elements contained in these two rows are M 11 (M 11 =a), M 1n (M 1n =b) and M 2i (M 2iAssuming that = c), the graphs of the rows containing the target elements and the graphs corresponding to the non-zero elements in the target rows can be the same graph, that is, graph 1, graph i, and graph n can be the same graph.

[0071] [Table 3]

[0072] Selectively, based on the row containing the target element and the target row, the graphs corresponding to the non-zero elements in the row containing the target element and the target row are made into the same graph. Then, it is determined whether all the graphs awaiting grouping in the graph set will achieve grouping. If the determination is no, the graphs awaiting grouping that have not been grouped are determined from the graph set, and grouping is performed according to the greedy grouping policy based on the graphs awaiting grouping.

[0073] When grouping graphs awaiting grouping by adopting the greedy grouping policy provided by the embodiments of the present invention, it is possible to select the row with the most non-zero elements when selecting the target row, thereby ensuring that the grouping is more representative. Subsequently, when selecting rows containing target elements from rows other than the target row in the graph comparison matrix based on non-zero elements, it is possible to group together as many directly or indirectly related graphs as possible, in order to cover rows where all non-zero elements in the first selected target row are located.

[0074] Furthermore, by adopting the greedy grouping policy according to the embodiment of this application and grouping graphs awaiting grouping, it is possible to group more graphs into one group with a smaller positional deviation and reduce the number of comparisons, while at the same time, the positional deviation range can be well controlled to achieve the overall optimal grouping result at the fastest speed.

[0075] For example, assuming that graphs 2, 3, and 4 are already grouped together as shown in Figure 3b, if we want to know whether graph 1 is in the same group as graphs 2, 3, and 4, we can arbitrarily select one graph from graphs 2, 3, and 4 based on the greedy grouping policy and compare it with graph 1. If there is a common graph, graphs 1, 2, 3, and 4 will be grouped together; otherwise, they cannot be grouped together, which significantly reduces the number of comparisons required.

[0076] In one selectable embodiment, if the pre-configured policy is an orphaned grouping policy, grouping multiple graphs awaiting grouping in the graph set based on the elements of the target row and the pre-configured policy, and outputting the grouping results, includes making the graphs corresponding to non-zero elements included in the target row the same graph based on the elements of the target row and the orphaned grouping policy, selecting ungrouped graphs from the graph set and grouping them based on the graphs included in the same graph, and outputting the grouping results until all multiple graphs awaiting grouping in the graph set are grouped.

[0077] By adopting the isolated grouping policy according to the embodiment of this application and grouping graphs awaiting grouping, it is possible to obtain an overall optimal grouping result while ensuring diversity in the groups and making specific regions (e.g., aperiodic regions / small area regions) stand out.

[0078] In one selectable embodiment, if the pre-configured policy is a rich grouping policy, grouping multiple graphs awaiting grouping in the graph set based on the elements of the target row and the pre-configured policy, and outputting the grouping results, includes making the graphs corresponding to non-zero elements included in the target row the same graph based on the elements of the target row and the rich grouping policy, selecting ungrouped graphs from the graph set and grouping them based on the graphs included in the same graph, and outputting the grouping results until all multiple graphs awaiting grouping in the graph set are grouped.

[0079] By adopting the rich grouping policy according to the embodiment of this application and grouping graphs awaiting grouping, it is possible to obtain an overall optimal grouping result, while also ensuring that all group members within the grouped group have a common graph, thus guaranteeing that the resulting graph is the most representative graph.

[0080] As described above, the graph grouping method that allows positional deviation according to the embodiment of the present application first obtains a set of graphs awaiting grouping, which includes multiple graphs awaiting grouping; second, constructs a graph comparison matrix based on the set of graphs, which includes multiple elements to represent the similarity between different graphs; and, based on a pre-set policy, selects a target row from the graph comparison matrix that satisfies pre-set conditions and determines the elements located in the target row; and finally, groups the multiple graphs awaiting grouping in the set of graphs based on the elements of the target row and the pre-set policy, and outputs the grouping result. This makes it possible to achieve an overall optimal grouping result in the graph grouping process and avoids the problem in the graph grouping process where the grouping result changes due to different data input orders, which further affects processing steps after graph grouping.

[0081] The following describes in detail the graph grouping method that allows for positional deviation according to the embodiment of this application, tailored to actual application scenarios.

[0082] As shown in Figure 3c, this is a schematic flowchart of a method for grouping graphs awaiting grouping based on a greedy grouping policy according to an embodiment of the present invention, the method comprising steps 1 to 6. Step 1: Construct a set of n graphs (graphs 1, 2, ..., n) based on n graphs awaiting grouping (also referred to as the predetermined n graphs in Figure 3c) and positional deviation parameters. Step 2: Compare pairs of graphs from each set to construct a graph comparison matrix, where each matrix element represents the similarity between the two graphs being compared. If there are no common graphs between the two sets of graphs, the matrix element is zero. Step 3: Select a row in the graph comparison matrix based on the selection policy. In this embodiment, the selection policy may be a greedy grouping policy. Step 4: Further select all rows in which non-zero elements are located in the rows selected in Step 3 (generally multiple rows; if a non-zero element appears only in that row, do not select other rows). Step 5: Group the graphs containing non-zero elements in the rows selected in Steps 3 and 4 into one group. Step 6: Determine whether all graphs are grouped or not. If not, repeat steps 3-5. If all are grouped, output the grouping result.

[0083] As shown in Figure 3d, this is a schematic flowchart of a method for grouping graphs awaiting grouping based on an isolated grouping policy according to an embodiment of the present invention, the method comprising the following steps 1 to 5. Step 1: Construct a set of n graphs (graphs 1, 2, ..., n) based on the n graphs awaiting grouping (also referred to as the predetermined n graphs in Figure 3d) and the positional deviation parameter. Step 2: Compare pairs of graph sets together to construct a graph comparison matrix, where each matrix element represents the similarity between the two compared graphs. If there are no common graphs between the two graph sets, the matrix element is zero. Step 3: Select a row in the graph comparison matrix based on the isolated grouping policy. Specifically, for the graph comparison matrix, based on the row ordering, we can select the row with the fewest non-zero elements according to the number of non-zero elements in each row. Step 4: Group the graphs containing non-zero elements in the selected rows into one group. Selectively, after performing Step 4, the grouped groups can be deleted, which not only speeds up the grouping process but also eliminates grouping interference. Step 5: Determine whether all graphs are grouped or not. If not, repeat steps 3 and 4. If all graphs are grouped, output the grouping result.

[0084] As shown in Figure 3e, this is a schematic flowchart of a method for grouping graphs awaiting grouping based on a rich grouping policy according to an embodiment of the present invention, and includes the following steps 1 to 5. Step 1: Construct a set of n graphs (graphs 1, 2, ..., n) based on the n graphs awaiting grouping (also referred to as the predetermined n graphs in Figure 3d) and the positional deviation parameter. Step 2: Compare pairs of graph sets together to construct a graph comparison matrix, where each matrix element represents the similarity between the two compared graphs. If there are no common graphs between the two graph sets, the matrix element is zero. Step 3: Select a row in the graph comparison matrix based on the rich strategy. Specifically, for the graph comparison matrix, based on the row ordering, we can select the row with the most non-zero elements according to the number of non-zero elements in each row, thereby ensuring that the graphs within this group are the richest and most representative. Step 4: Group the graphs containing non-zero elements in the selected rows. After selectively grouping the graphs containing non-zero elements in the selected rows, the grouped groups can be deleted, which not only speeds up the grouping process but also eliminates grouping interference. Step 5: Determine whether all graphs are grouped or not. If not, repeat steps 3 and 4. If all graphs are grouped, output the grouping result.

[0085] Figure 4 shows a schematic diagram of the structure of a graph grouping device that allows for positional deviation according to an embodiment of the present application. As shown in Figure 4, the device is A retrieval module 401 for obtaining a set of graphs waiting to be grouped, which includes multiple graphs waiting to be grouped, A construction module 402 for constructing a graph comparison matrix containing multiple elements to represent the similarity between different graphs based on a set of graphs, A selection module 403 for selecting target rows that satisfy pre-defined conditions from a graph comparison matrix based on a pre-defined policy, and for determining the elements located in the target rows, It includes a grouping module 404 for grouping multiple graphs awaiting grouping in a graph set based on the elements of the target row and a pre-configured policy, and for outputting the grouping results.

[0086] Selectively, the pre-configured policies include one of the following: greedy grouping policy, orphaned grouping policy, rich grouping policy, or weak rich grouping policy.

[0087] Selectively, if the pre-configured policy is a greedy grouping policy, the selection module 403 will: Based on the greedy grouping policy, a first selection unit is used to select the row with the most non-zero elements from the graph comparison matrix and determine it as the target row, Includes a first decision unit for determining the elements located in the target row, based on the target row.

[0088] Selectively, the first selected unit is, Determining the number of non-zero elements in each row of the graph comparison matrix, The process involves ordering each row in a graph comparison matrix based on the number of non-zero elements, and obtaining the ordering result. It is used to select the row with the most non-zero elements from the graph comparison matrix based on the ordering result and determine it as the target row.

[0089] Selectively, if the pre-configured policy includes an orphaned grouping policy, the selection module 403 will: Based on the isolated grouping policy, a second selection unit is used to select the row with the fewest non-zero elements from the graph comparison matrix and determine it as the target row, It includes a second decision unit for determining the elements located in the target row, based on the target row.

[0090] Selectively, the second selected unit is, Determining the number of non-zero elements in each row of the graph comparison matrix, The process involves ordering each row in a graph comparison matrix based on the number of non-zero elements, and obtaining the ordering result. It is used to select the row with the fewest non-zero elements from the graph comparison matrix based on the ordering result and determine it as the target row.

[0091] Selectively, if the pre-configured policy includes a rich grouping policy, the selection module 403 will: Based on the rich grouping policy, a third selection unit is used to select the row with the most non-zero elements from the graph comparison matrix and determine it as the target row, It includes a third decision unit that determines the elements located in the target row based on the target row.

[0092] Selectively, the third selected unit is, Determining the number of non-zero elements in each row of the graph comparison matrix, The process involves ordering each row in a graph comparison matrix based on the number of non-zero elements, and obtaining the ordering result. It is used to select the row with the most non-zero elements from the graph comparison matrix based on the ordering result and determine it as the target row.

[0093] Selectively, if the pre-configured policy is a greedy grouping policy, the grouping module 404 will: Based on the elements of the target row, determine the non-zero elements in the target row, Based on non-zero elements, select rows in the graph comparison matrix other than the target row that contain the target element whose value is equal to the non-zero element, It is used to group rows containing target elements and target rows into the same graph based on the target row itself, and to output the grouping result.

[0094] A graph grouping device that selectively allows such positional deviations further, This involves determining whether all graphs in a set of graphs that are awaiting grouping will be grouped, If the decision is negative, then the graphs that are not yet grouped and awaiting grouping will be determined from the set of graphs. This is used to perform grouping according to a greedy grouping policy based on a graph of data awaiting grouping.

[0095] Selectively, if the pre-configured policy is an orphaned grouping policy, the grouping module 404 will: Based on the elements of the target row and the orphaned grouping policy, the graphs corresponding to the non-zero elements included in the target row will be made into the same graph, This is used to select ungrouped graphs from a set of graphs based on the graphs contained within the same graph, group them, and output the grouping results until all graphs in the set that are awaiting grouping are grouped.

[0096] Selectively, if the pre-configured policy is a rich grouping policy, the grouping module 404 will: Based on the elements of the target row and the rich grouping policy, the graphs corresponding to non-zero elements included in the target row will be made into the same graph, This is used to select ungrouped graphs from a set of graphs based on the graphs contained within the same graph, group them, and output the grouping results until all graphs in the set that are awaiting grouping are grouped.

[0097] Each module / unit in the apparatus shown in Figure 4 has the function to realize each step in Figure 3a and can achieve the corresponding technical effects; for the sake of brevity, a detailed explanation is omitted here.

[0098] Figure 5 is a schematic diagram of a graph grouping device that allows for positional deviation according to an embodiment of the present application.

[0099] A graph grouping device that allows positional deviation may include a processor 501 and a memory 502 for storing computer program instructions.

[0100] Specifically, the processor 501 may include a central processing unit (CPU) or an application-specific integrated circuit (ASIC), or it may be configured as one or more integrated circuits that implement the embodiments of the present application.

[0101] Memory 502 may include large-capacity memory for data or instructions. For example, memory 502 may include a hard disk drive (HDD), a flexible disk drive, flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a USB (Universal Serial Bus) drive, or a combination of two or more of these. Where appropriate, memory 502 may include removable or non-removable (or fixed) media. Where appropriate, memory 502 may be located inside or outside a graph grouping device that tolerates positional deviation. In certain embodiments, memory 502 may be non-volatile solid-state memory.

[0102] In one embodiment, the memory 502 may be a read-only memory (ROM). In one embodiment, the ROM may be a masked ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically rewritable ROM (EAROM), or a flash memory, or a combination of two or more of these.

[0103] The processor 501 realizes the graph grouping method of any of the above embodiments by reading and executing computer program instructions stored in memory 502.

[0104] In one example, a graph grouping device that tolerates positional deviation may further include a communication interface 503 and a bus 510. As shown in Figure 5, the processor 501, memory 502, and communication interface 503 are connected via the bus 510 and complete communication between them.

[0105] The communication interface 503 is primarily configured to implement communication between each module, apparatus, unit and / or device in the embodiments of the present application.

[0106] Bus 510 includes hardware, software, or both, and connects components of a graph grouping device that tolerates positional deviations. For example, the bus may include, but is not limited to, an Accelerator Graphics Port (AGP) or other graphics bus, an Enhanced Industrial Standards Architecture (EISA) bus, a Front-End Bus (FSB), an Ultra-High Transmission (HT) interconnect, an Industrial Standards Architecture (ISA) bus, an Infinite Bandwidth interconnect, a Low Pin-in-Number (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Accessory (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses or a combination of two or more. Where appropriate, bus 510 may include one or more buses. Where appropriate, bus 510 may include one or more buses. While the embodiments of this application describe or illustrate specific buses, this application considers any suitable bus or interconnect.

[0107] Furthermore, in combination with the graph grouping method that allows positional deviation in the above embodiments, the embodiments of the present application can be realized by providing a computer-readable storage medium. A computer program instruction is stored in the computer-readable storage medium, and when the computer program instruction is executed by a processor, one of the graph grouping methods with an allowable positional deviation in the above embodiments is realized.

[0108] Clearly, this application is not limited to the specific configurations and processes described and illustrated above. For simplicity, a detailed description of known methods is omitted here. In the above embodiments, several specific steps are described or shown as examples. However, the process of the method of this application is not limited to the specific steps described and shown. Those skilled in the art, after understanding the spirit of this application, can make various changes, modifications and additions, or change the order between steps.

[0109] The functional blocks shown in the above configuration block diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they may include, for example, electronic circuits, dedicated integrated circuits (ASICs), appropriate firmware, plug-ins, or functional cards. When implemented in software, the elements of this application are programs or code segments for performing the required tasks. The programs or code segments may be stored in a machine-readable medium or transmitted to a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable mediums include electronic circuits, semiconductor memory devices, ROMs, flash memory, erasable ROMs (EROMs), flexible disks, CD-ROMs, optical disks, hard disks, optical fiber media, and radio frequency (RF) links. Code segments can be downloaded via computer networks such as the Internet or intranets.

[0110] It should be noted that the exemplary embodiments referenced in this application describe several methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the steps described above. That is, the steps may be performed in the order referenced in the embodiments, in a different order than that of the embodiments, or some steps may be performed simultaneously.

[0111] Various aspects of this application have been described with reference to flowcharts and / or block diagrams of methods, apparatus (systems) and computer program products relating to embodiments of this application. Each block in the flowcharts and / or block diagrams and each combination of blocks in the flowcharts and / or block diagrams may be implemented by computer program instructions. These computer program instructions are provided to a processor of a general-purpose computer, a dedicated computer, or other programmable data processing device to generate a machine such that these instructions, executed by the processor of the computer or other programmable data processing device, enable the implementation of one or more blocks in the flowcharts and / or block diagrams that are specified in the flowcharts and / or block diagrams. Such a processor may, but is not limited to, a general-purpose processor, a dedicated processor, a special application processor, or a field-programmable gate array. Each block in the block diagrams and / or flowcharts and each combination of blocks in the block diagrams and / or flowcharts may be implemented by dedicated hardware that performs a specified function or operation, or by a combination of dedicated hardware and computer instructions.

[0112] The above describes only specific embodiments of the present application. As will be obvious to those skilled in the art, for the sake of brevity and ease of explanation, the specific operating processes of the systems, modules, and units described above can be referenced to the corresponding processes in the embodiments of the aforementioned methods and will not be described here. It should be understood that the scope of protection of this application is not limited thereto. Those skilled in the art will readily conceive of various equivalent modifications or substitutions within the technical scope disclosed herein, and all such modifications or substitutions should fall within the scope of protection of this application.

Claims

1. Obtain a set of graphs waiting to be grouped, which includes graphs waiting to be grouped in multiple groups. Based on the aforementioned set of graphs, a graph comparison matrix is ​​constructed that includes multiple elements to represent the similarity between different graphs. Based on a pre-configured policy, select a target row from the graph comparison matrix that satisfies pre-configured conditions, and determine the element located in the target row. The process includes grouping the multiple graphs awaiting grouping in the graph set based on the elements of the target row and the pre-configured policy, and outputting the grouping result. The aforementioned pre-configured policy includes one of the following: greedy grouping policy, orphaned grouping policy, rich grouping policy, or weak rich grouping policy. If the aforementioned pre-configured policy is the greedy grouping policy, then based on the pre-configured policy, selecting a target row that satisfies the pre-configured conditions from the graph comparison matrix and determining the element located in the target row is: Based on the aforementioned greedy grouping policy, the row with the most non-zero elements is selected from the graph comparison matrix and determined to be the target row. The process includes determining the element located in the target row based on the target row, A graph grouping method that allows for positional deviation, characterized by the following features.

2. Based on the aforementioned greedy grouping policy, selecting the row with the most non-zero elements from the graph comparison matrix and determining it as the target row is: Determining the number of non-zero elements in each row of the aforementioned graph comparison matrix, Based on the number of non-zero elements, each row in the graph comparison matrix is ​​ordered, and the ordering result is obtained. This includes selecting the row with the most non-zero elements from the graph comparison matrix based on the ordering result and determining it as the target row, A graph grouping method that allows for positional deviation as described in claim 1, characterized in that it is a graph grouping method that allows for positional deviation as described in claim 1.

3. If the aforementioned pre-configured policy includes the isolated grouping policy, then selecting a target row that satisfies the pre-configured conditions from the graph comparison matrix and determining the element located in the target row based on the pre-configured policy is as follows: Based on the isolated grouping policy, the row with the fewest non-zero elements is selected from the graph comparison matrix and determined to be the target row. The process includes determining the element located in the target row based on the target row, A graph grouping method that allows for positional deviation as described in claim 1, characterized in that it is a graph grouping method that allows for positional deviation as described in claim 1.

4. Based on the isolated grouping policy, selecting the row with the fewest non-zero elements from the graph comparison matrix as the target row is: Determining the number of non-zero elements in each row of the aforementioned graph comparison matrix, Based on the number of non-zero elements, each row in the graph comparison matrix is ​​ordered, and the ordering result is obtained. This includes selecting the row with the fewest non-zero elements from the graph comparison matrix based on the ordering result and determining it as the target row, A graph grouping method that allows for positional deviation, as described in claim 3.

5. If the aforementioned pre-configured policy includes the rich grouping policy, then, based on the pre-configured policy, selecting a target row from the graph comparison matrix that satisfies the pre-configured conditions and determining the element located in the target row is: Based on the rich grouping policy, the row with the most non-zero elements is selected from the graph comparison matrix and determined to be the target row. The process includes determining the element located in the target row based on the target row, A graph grouping method that allows for positional deviation as described in claim 1, characterized in that it is a graph grouping method that allows for positional deviation as described in claim 1.

6. Based on the rich grouping policy, selecting rows containing non-zero elements from the graph comparison matrix and determining them as target rows is: Determining the number of non-zero elements in each row of the aforementioned graph comparison matrix, Based on the number of non-zero elements, each row in the graph comparison matrix is ​​ordered, and the ordering result is obtained. This includes selecting the row with the most non-zero elements from the graph comparison matrix based on the ordering result and determining it as the target row, A graph grouping method that allows for positional deviation, as described in claim 5.

7. If the pre-configured policy is the greedy grouping policy, then grouping the multiple graphs awaiting grouping in the graph set based on the elements of the target row and the pre-configured policy, and outputting the grouping result, Based on the elements of the target row, determine the non-zero elements in the target row, Based on the non-zero element, select a row from the rows other than the target row in the graph comparison matrix that contains a target element whose numerical value is equal to the non-zero element; The process includes: making the graphs corresponding to the non-zero elements in the rows containing the target elements and the target rows the same graph based on the target rows, and outputting the grouping result; A graph grouping method that allows for positional deviation as described in claim 1, characterized in that it is a graph grouping method that allows for positional deviation as described in claim 1.

8. The aforementioned method, Determining whether all graphs in the aforementioned set of graphs that are awaiting grouping will achieve grouping, If the decision is negative, then the graphs that are not grouped and awaiting grouping are determined from the aforementioned set of graphs. The further includes performing grouping according to the greedy grouping policy based on the graph of groups awaiting grouping, A graph grouping method that allows for positional deviation, as described in claim 7.

9. If the pre-configured policy is the orphaned grouping policy, then grouping the multiple graphs awaiting grouping in the graph set based on the elements of the target row and the pre-configured policy, and outputting the grouping result, Based on the elements of the target row and the isolated grouping policy, the graphs corresponding to the non-zero elements included in the target row are made into the same graph, This includes selecting ungrouped graphs from the graph set based on the graphs included in the same graph, grouping them, and outputting the grouping results until all of the multiple graphs awaiting grouping in the graph set are grouped. A graph grouping method that allows for positional deviation as described in claim 1, characterized in that it is a graph grouping method that allows for positional deviation as described in claim 1.

10. If the pre-configured policy is the rich grouping policy, then grouping the multiple graphs awaiting grouping in the graph set based on the elements of the target row and the pre-configured policy, and outputting the grouping result, Based on the elements of the target row and the rich grouping policy, the graphs corresponding to the non-zero elements included in the target row are made into the same graph, This includes selecting ungrouped graphs from the graph set based on the graphs included in the same graph, grouping them, and outputting the grouping results until all of the multiple graphs awaiting grouping in the graph set are grouped. A graph grouping method that allows for positional deviation as described in claim 1, characterized in that it is a graph grouping method that allows for positional deviation as described in claim 1.

11. A module for obtaining a set of graphs waiting to be grouped, which includes graphs waiting to be grouped, A construction module that constructs a graph comparison matrix containing multiple elements to represent the similarity between different graphs based on the aforementioned set of graphs, A selection module for selecting a target row from the graph comparison matrix that satisfies pre-defined conditions based on a pre-defined policy, and for determining the element located in the target row, A grouping module for grouping the multiple graphs awaiting grouping in the graph set based on the elements of the target row and the pre-configured policy, and for outputting the grouping result, is included. The aforementioned pre-configured policy includes one of the following: greedy grouping policy, orphaned grouping policy, rich grouping policy, or weak rich grouping policy. If the aforementioned pre-configured policy is the Greedy Grouping Policy, the Selection Module, A first selection unit for selecting the row with the most non-zero elements from the graph comparison matrix and determining it as the target row, based on the greedy grouping policy, A first determination unit for determining the element located in the target row based on the target row, A graph grouping device that allows for positional deviation, characterized by the above.

12. The first selection unit described above is, Determining the number of non-zero elements in each row of the aforementioned graph comparison matrix, Based on the number of non-zero elements, each row in the graph comparison matrix is ​​ordered, and the ordering result is obtained. Based on the ordering result, the row with the most non-zero elements is selected from the graph comparison matrix and determined to be the target row, and is used for this purpose. A graph grouping device that allows positional deviation according to claim 11, characterized in that

13. If the aforementioned pre-configured policy includes the aforementioned orphaned grouping policy, the selection module shall A second selection unit for selecting the row with the fewest non-zero elements from the graph comparison matrix and determining it as the target row, based on the isolated grouping policy, A second determination unit for determining the element located in the target row based on the target row, including, A graph grouping device that allows positional deviation according to claim 11, characterized in that

14. The second selection unit is, Determining the number of non-zero elements in each row of the aforementioned graph comparison matrix, Based on the number of non-zero elements, each row in the graph comparison matrix is ​​ordered, and the ordering result is obtained. Based on the ordering result, the row with the fewest non-zero elements is selected from the graph comparison matrix and determined to be the target row, and is used for this purpose. A graph grouping device that allows positional deviation as described in claim 13, characterized in that it is a graph grouping device that allows positional deviation as described in claim 13.

15. If the aforementioned pre-configured policy includes the rich grouping policy, the selection module shall A third selection unit for selecting the row with the most non-zero elements from the graph comparison matrix and determining it as the target row, based on the rich grouping policy, A third determination unit that determines the element located in the target row based on the target row, A graph grouping device that allows positional deviation according to claim 11, characterized in that

16. The aforementioned third selection unit is, Determining the number of non-zero elements in each row of the aforementioned graph comparison matrix, Based on the number of non-zero elements, each row in the graph comparison matrix is ​​ordered, and the ordering result is obtained. Based on the ordering result, the row with the most non-zero elements is selected from the graph comparison matrix and determined to be the target row, and is used for this purpose. A graph grouping device that allows positional deviation according to claim 15, characterized in that

17. If the aforementioned pre-configured policy is the greedy grouping policy, the grouping module shall Based on the elements of the target row, determine the non-zero elements in the target row, Based on the non-zero element, select a row from the rows other than the target row in the graph comparison matrix that contains a target element whose numerical value is equal to the non-zero element; The following is used to make the graphs corresponding to the non-zero elements in the rows containing the target elements and the target rows the same graph, and to output the grouping result, based on the rows containing the target elements and the target rows: A graph grouping device that allows positional deviation according to claim 11, characterized in that

18. The device further, Determining whether all graphs in the aforementioned set of graphs that are awaiting grouping will achieve grouping, If the decision is negative, the graphs that are not grouped and awaiting grouping are determined from the set of graphs. Based on the graph of groups awaiting grouping, grouping is performed according to the greedy grouping policy, and is used for this purpose. A graph grouping device that allows positional deviation according to claim 17, characterized in that

19. If the aforementioned pre-configured policy is the isolated grouping policy, the grouping module shall Based on the elements of the target row and the isolated grouping policy, the graphs corresponding to the non-zero elements included in the target row are made into the same graph, This is used to select and group ungrouped graphs from the graph set based on the graphs included in the same graph, and to output the grouping results until all of the multiple graphs awaiting grouping in the graph set are grouped. A graph grouping device that allows positional deviation according to claim 11, characterized in that

20. If the aforementioned pre-configured policy is the rich grouping policy, the grouping module shall Based on the elements of the target row and the rich grouping policy, the graphs corresponding to the non-zero elements included in the target row are made into the same graph, This is used to select and group ungrouped graphs from the graph set based on the graphs included in the same graph, and to output the grouping results until all of the multiple graphs awaiting grouping in the graph set are grouped. A graph grouping device that allows positional deviation according to claim 11, characterized in that

21. A graph grouping device that allows for positional deviation, It includes a processor and memory for storing computer program instructions, When the processor executes the computer program instruction, it implements a graph grouping method that allows positional deviations according to any one of claims 1 to 10. A graph grouping device that allows for positional deviation, characterized by the above features.

22. A computer-readable storage medium, Computer program instructions are stored, When the computer program instruction is executed by the processor, a graph grouping method that allows positional deviation as described in any one of claims 1 to 10 is realized. A computer-readable storage medium characterized by the following features.

Images