CN114610962A

CN114610962A - Graph data processing method, device and equipment

Info

Publication number: CN114610962A
Application number: CN202210196579.8A
Authority: CN
Inventors: 王妍岩; 邓绍婷; 费冬妮; 廖博森
Original assignee: Alipay Hangzhou Information Technology Co Ltd
Current assignee: Alipay Hangzhou Information Technology Co Ltd
Priority date: 2022-03-01
Filing date: 2022-03-01
Publication date: 2022-06-10

Abstract

The embodiment of the specification discloses a method, a device and equipment for processing graph data. The scheme comprises the following steps: displaying the graph data and a configuration interface for configuring a graph mode at the front end, wherein the graph mode is used for designating a sub-graph structure; receiving the editing operation of adding nodes and edges by a user through a configuration interface; generating a target graph mode in response to an editing operation; matching in the graph data according to the target graph mode to obtain a subgraph matched with the target graph mode in the graph data; in the graph data displayed at the front end, the region where the sub-graph is located is displayed in a highlighting mode so that a user can pay attention to the region.

Description

Graph data processing method, device and equipment

Technical Field

The present disclosure relates to the field of computer technologies, and in particular, to a method, an apparatus, and a device for processing graph data.

Background

With the development of computer and internet technologies, more and more relationship data is emerging. Compared with general statistical data, the relational data are more complex and are difficult to clearly manage and display only through traditional modes such as characters, tables and the like. Clear graph visualization can help end users and analysts to quickly learn relationship data and help locate risks and problems in the data. For example, in business scenarios such as wind control, finance, user network, and the like, graph visual analysis of data has become an important technology.

Therefore, drawing relational data into a point-line relational graph (data included in the relational graph is also referred to as graph data) in the form of nodes and edges is the most common means for graph visualization at present. When the data itself is complex, the drawn relationship graph is also complex, and the time sequence information included in the relationship data is generally hidden in the attribute of the node or edge, which makes the relationship graph more complex.

When a user wants to find a desired local structure (e.g., a sub-relationship diagram with a special structure, a local structure carrying chronological information) in the relationship diagram, the user usually searches through artificial visual search, using a graph pattern matching algorithm, and the like.

Based on this, there is a need for a graph data processing scheme that is more efficient, accurate, and able to more intuitively present to a user the layout structure that he desires.

Disclosure of Invention

One or more embodiments of the present specification provide a graph data processing method, apparatus, device, and storage medium to solve the following technical problems: there is a need for a graph data processing scheme that is more efficient, accurate, and able to more intuitively present to a user the layout structure that he desires.

To solve the above technical problem, one or more embodiments of the present specification are implemented as follows:

one or more embodiments of the present specification provide a method for processing graph data, where the graph data includes a plurality of nodes and edges between the nodes, the method including:

displaying the graph data on a configuration interface for configuring graph modes at the front end, wherein the graph modes are used for designating sub-graph structures;

receiving the editing operation of adding nodes and edges by a user through the configuration interface;

generating a target graph mode in response to the editing operation;

matching in the graph data according to the target graph mode to obtain a sub-graph matched with the target graph mode in the graph data;

and in the graph data displayed at the front end, performing saliency display on the region where the sub-graph is located so as to facilitate the user to pay attention to.

acquiring time sequence information of edges in the graph data;

generating hierarchical layout information of each node and sequence number information of each edge according to the time sequence information;

and correspondingly and hierarchically displaying the nodes in the graph data and displaying the edges in the graph data with sequence numbers at the front end according to the hierarchical layout information and the sequence number information, wherein the sequence numbers indicate corresponding time sequences.

One or more embodiments of the present specification provide a graph data processing apparatus, where the graph data includes a plurality of nodes and edges between the nodes, the apparatus including:

the front-end display module is used for displaying the graph data on a configuration interface for configuring a graph mode at the front end, wherein the graph mode is used for specifying a sub-graph structure;

the editing module receives the editing operation of adding nodes and edges by a user through the configuration interface;

the first generation module responds to the editing operation and generates a target graph mode;

the matching module is used for matching in the graph data according to the target graph mode to obtain a subgraph matched with the target graph mode in the graph data;

and the prominent displaying module is used for prominently displaying the area where the subgraph is located in the graph data displayed at the front end so as to facilitate the user to pay attention to.

the acquisition module acquires the time sequence information of the edges in the graph data;

the second generation module generates the hierarchical layout information of each node and the sequence number information of each edge according to the time sequence information;

and the level display module correspondingly displays the nodes in the graph data in a level mode at the front end and displays the edges in the graph data in a sequence mode according to the level layout information and the sequence number information, and the sequence number indicates the corresponding time sequence.

at least one processor; and (c) a second step of,

a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,

the memory stores instructions executable by the at least one processor to cause the at least one processor to:

displaying the graph data on a configuration interface for configuring a graph mode on the front end, wherein the graph mode is used for specifying a sub-graph structure;

generating a target graph mode in response to the editing operation;

and in the graph data displayed at the front end, performing saliency display on the region where the subgraph is located so as to facilitate the user to pay attention to.

One or more embodiments of the present specification provide a graph data processing apparatus, the graph data including a plurality of nodes and edges between the nodes, the apparatus including:

at least one processor; and the number of the first and second groups,

the memory stores instructions executable by the at least one processor to enable the at least one processor to:

acquiring time sequence information of edges in the graph data;

At least one technical scheme adopted by one or more embodiments of the specification can achieve the following beneficial effects:

a user can design a special graph structure mode interested by the user through simple configuration, and then sub-graphs with similar structures are automatically matched in graph data. When the user knows the graph pattern of interest, the pattern can be used directly for calculation; when the user does not know the specific graph mode, the scheme can be used for exploring and discovering the interested mode structure.

Compared with the traditional observation by human eyes, the automatic matching is carried out by the corresponding matching algorithm, so that the matching efficiency and accuracy can be improved. Compared with the traditional method only adopting a pattern matching algorithm, the method supports the user to edit and configure the pattern according to the self scene requirement, and can correct the pattern in real time.

When the user knows the image pattern that the user wants to find, the user can directly edit the desired target image pattern in a self-defining way in the pattern editor and match the target image pattern. When the user does not know the specific mode structure and conditions, the mode editor can be used, and the graph mode which can help the user to find the concerned information can be finally found out through multiple times of editing, modifying and switching the mode. After different image data are switched, the known target image patterns can still be used for matching, and interesting patterns on different data are found.

The result calculated by the pattern matching algorithm can be presented more prominently (for example, by sub-image highlighting, table and main map linkage and the like), so that the display interface is more intuitive, and the user has more opportunities to explore unknown special patterns.

Drawings

In order to more clearly illustrate the embodiments of the present specification or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments described in the present specification, and for those skilled in the art, other drawings can be obtained according to the drawings without inventive labor.

FIG. 1 is a flow diagram illustrating a graph schema-based graph data processing method according to one or more embodiments of the present disclosure;

fig. 2 is a schematic diagram of a front-end interface in an application scenario according to one or more embodiments of the present disclosure;

FIG. 3 is a schematic diagram of a creation mode in an application scenario, according to one or more embodiments of the present disclosure;

FIG. 4 is a schematic diagram of generating a target graph schema in the form of adding nodes in an application scenario, according to one or more embodiments of the present disclosure;

FIG. 5 is a schematic diagram of generating a target graph schema in the form of an added template in an application scenario, according to one or more embodiments of the present disclosure;

fig. 6 is a schematic flowchart of a graph data processing method based on graph schema in an application scenario, according to one or more embodiments of the present disclosure;

FIG. 7 is a flowchart illustrating a method for processing graph data based on timing information according to one or more embodiments of the present disclosure;

FIG. 8 is an illustration of a timing analysis configuration interface in an application scenario, as provided by one or more embodiments of the present disclosure;

FIG. 9 is a schematic diagram of the time sequence analysis effect in an application scenario according to one or more embodiments of the present disclosure;

FIG. 10 is a schematic diagram illustrating a filtering of time ranges on an interaction timeline in an application scenario, according to one or more embodiments of the present disclosure;

fig. 11 is a schematic diagram of filtering time points on an interaction timeline in an application scenario, according to one or more embodiments of the present disclosure;

fig. 12 is a schematic diagram of a list analysis interface of an analysis result interface in an application scenario, according to one or more embodiments of the present disclosure;

fig. 13 is a schematic diagram of a chart analysis interface of an analysis result interface in an application scenario, according to one or more embodiments of the present disclosure;

fig. 14 is a flowchart illustrating a method for processing graph data based on timing information in an application scenario, according to one or more embodiments of the present disclosure;

FIG. 15 is a schematic diagram of a diagram schema-based diagram data processing apparatus according to one or more embodiments of the present disclosure;

fig. 16 is a schematic structural diagram of a graph data processing apparatus based on timing information according to one or more embodiments of the present disclosure.

Detailed Description

The embodiment of the specification provides a graph data processing method, a graph data processing device, graph data processing equipment and a storage medium.

In order to make those skilled in the art better understand the technical solutions in the present specification, the technical solutions in the embodiments of the present specification will be clearly and completely described below with reference to the drawings in the embodiments of the present specification, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all embodiments. All other embodiments, which can be obtained by a person skilled in the art without making any inventive step based on the embodiments of the present disclosure, shall fall within the scope of protection of the present application.

Before processing the relationship graph, the relationship graph is stranger to the user. Although the user has the purpose of exploring the graph structure, it is unknown how to explore and what structure to explore. The first method is obviously inapplicable to complex graph data scenes through manual visual search, and human eyes are difficult to find subgraphs of specific structures in complicated graphs, so that the efficiency is low and the error rate is high. For the second way of using the graph pattern matching algorithm, the graph pattern is generally built in the system, so that it is difficult for a user to modify and edit the graph pattern in real time, and it is also difficult to configure the attribute conditions of nodes and edges in the pattern as required, and only the built-in pattern can be used. However, the graph data and the graph scenes are varied, and graph modes are different under different scenes and requirements and are difficult to enumerate, so that the second mode has low degree of freedom and poor expandability, and is difficult to be universal for various scenes and requirements. In addition, most users may not have prior knowledge of the data itself and know what the specific structural patterns that need to be focused on in the current scenario, and both of these ways are difficult to help users explore and summarize the unknown structural patterns.

Fig. 1 is a flow diagram illustrating a graph data processing method based on a graph schema according to one or more embodiments of the present disclosure. The method can be applied to different business fields, such as a wind control field (such as risk patterns of gambling, money laundering and the like in a wind control scene), an internet financial business field, an online social field (such as a special community pattern in a social network), an e-commerce business field, an instant messaging business field, a game business field, a public business field and the like. The process can be executed by computing equipment in the corresponding field (such as an intelligent customer service server or an intelligent mobile terminal corresponding to the payment service, and the like), and certain input parameters or intermediate results in the process allow manual intervention and adjustment to help improve the accuracy.

The process in fig. 1 may include the following steps:

s102: and displaying the graph data on a configuration interface for configuring a graph mode at the front end, wherein the graph mode is used for specifying a sub-graph structure.

Graph schema refers to a structural relationship of certain node and edge components. For example, there is an edge between two nodes, which is a simple graph model. The graph schema is a general schema, and does not refer to a specific local graph structure in the graph data, and whether the graph schema contains specific node or edge information or not can be configured according to specific requirements of users.

Fig. 2 is a schematic diagram of a front-end interface in an application scenario, where the diagram data and the configuration interface are arranged in different front-end areas for presentation according to one or more embodiments of the present disclosure. The method includes the steps that graph data are displayed in a front-end interface, when a user triggers the front-end interface (for example, the user clicks a certain switch, a certain node or an edge in the front-end interface), a drawer in a right-side area is popped out, a configuration interface of a graph mode is displayed in the drawer, and meanwhile the graph data are moved to a left-side area in a translation, reduction and other modes, so that the user can edit in the configuration interface according to the graph data.

S104: and receiving the editing operation of adding nodes and edges by the user through the configuration interface.

Based on the requirement of the user, the configuration interface can select a creating mode, a clone mode and other modes to edit the graph mode. After the user selects the creation mode or the clone mode in the configuration interface, the user clicks an edit button to enter a mode editor provided by the configuration interface.

And if the user selects the creation mode, creating a null mode for the user, and editing the null mode by the user to obtain a target graph mode. And if the user selects the clone mode, displaying a plurality of existing graph modes to the user, and editing according to the graph mode selected by the user to obtain the target graph mode when the user selects the graph modes.

Still taking fig. 2 as an example, the user clicks "new mode" to enter the creation mode. If the edited mode 1 exists currently, the user clones the edited mode by clicking the copy mode-1, enters a clone mode, copies the existing mode 1 selected by the user into the current mode editor, and further edits the mode. When the user does not know the required specific mode structure and conditions, the user can quickly edit and modify for many times through the clone mode so as to finally debug and obtain the required target graph mode, and the editing efficiency is improved.

Specifically, fig. 3 and fig. 4 are schematic diagrams of generating a target graph pattern in the form of adding nodes and adding templates in an application scenario, respectively, provided by one or more embodiments of the present specification. The mode editor provides a plurality of editing modes for users, such as the mode editor is shown in the left area in fig. 3 and 4, and comprises two modes of adding nodes and adding templates. In the editing process of the user, the user can add a single node at a time in the form of adding nodes, or add a graph mode template consisting of a plurality of nodes and edges at a time in the form of adding templates. When the user adds a node (referred to herein as a first editing operation) by "adding a node" or "adding a template", editing can be performed in the schema editor by specifying a type (for example, as shown in fig. 3, in the movie industry-related graph data, the type includes Action, USA, PG-13, and the like), an attribute condition (for example, description information description, movie soundtrack file Score, movie show time file Runtime, and the like) for the added node.

In addition to the first editing operation, the user may also perform an add operation (referred to herein as a second editing operation) on an edge between the newly added nodes at the schema editor, such as right clicking on a node in which an edge extends in a drag-and-drop manner. In the second editing operation, types (for example, as shown in fig. 3, in the graph data related to the movie industry, the types include movie rating, movie company, movie lead actor, movie category, movie country of origin, and the like) and attribute conditions may also be specified for the edges, and are edited and added, which is similar to the first editing operation and will not be described herein again.

If the user selects the creation mode, the first editing operation and the second editing operation can be directly executed, if the user selects the clone mode, new nodes and edges can be directly added on the basis of cloning, and the graph mode template of the selected clone can also be edited to modify the nodes and edges contained in the graph mode template.

S106: and generating a target graph mode in response to the editing operation.

After the user performs the first editing operation, the second editing operation, and other editing operations (for example, deleting existing nodes and edges, or changing the representation modes, the positions, and the like of the nodes and the edges in the graph mode), the target graph mode is obtained in the mode editor, the obtained target graph mode can be returned to the main interface by saving, and the user performs the next matching operation according to the target graph mode in the main interface.

S108: and matching in the graph data according to the target graph mode to obtain a subgraph matched with the target graph mode in the graph data.

Fig. 5 is a schematic diagram of a matching result in an application scenario, provided by one or more embodiments of the present disclosure, that is, after a user returns to a main interface, the user can click a "match" button, and then match the target graph pattern edited in the pattern editor. In the process of matching, subgraphs similar to the target graph mode are searched in the graph data of the left area. Whether the similarity exists or not can be considered from multiple dimensions, such as the structure of the target graph mode, the types of nodes and edges contained in the target graph mode, attribute conditions and the like. The matching process can use matching algorithms such as Ullmann, VF2, GraphQL, GADDI, and Spath.

In addition to finding similar subgraphs, the matching process may also be to find other types of subgraphs according to the target graph data, for example, to find complementary subgraphs, which may refer to graph structures between the target graph pattern and the subgraphs, which can form graph structures in a preset template, or contain types, attribute conditions, and the like of nodes and edges that are missing from each other.

S110: and in the graph data displayed at the front end, performing saliency display on the region where the sub-graph is located so as to facilitate the user to pay attention to.

The highlighting presentation may be to present a matching result list below the right area, and when a user clicks one of the lists, in the graph data on the left side, a sub-graph corresponding to the user clicked list is presented in a highlighted form. As shown in fig. 5, which illustrates a graph data processing scheme within the movie industry. Connecting a node of a 'movie' type to a node of a 'movie company' type through an edge of the 'movie company' type, wherein the 'movie' node satisfies a score (filmScore) attribute of more than 5, thereby obtaining a target graph mode. Through a matching algorithm, two similar sub-graphs are obtained through calculation in the graph data on the left side, a corresponding matching result list is shown on the lower portion of the right side, and highlight display is carried out in the graph data on the left side. Of course, the highlighting display may also include other forms, such as thickening the sub-graph, changing the color of the sub-graph, displaying the sub-graph in a partially enlarged manner, and so on, and as long as it can be distinguished from other surrounding graph structures, the highlighting display can be regarded as the highlighting display.

Compared with the traditional observation by human eyes, the automatic matching is carried out by the corresponding matching algorithm, so that the matching efficiency and accuracy can be improved. Compared with the traditional method only using the pattern matching algorithm, the method supports the user to edit and configure the pattern according to the self scene requirement, and can correct the pattern in real time. And the editing efficiency can be improved by adding a template, cloning the existing mode and the like.

When the user knows the image pattern that the user wants to find, the user can directly edit the desired target image pattern in a self-defining way in the pattern editor and match the target image pattern. When the user does not know the specific mode structure and conditions, the mode editor can be used for editing, modifying and switching the modes for multiple times, and finally, the graph mode which can help the user to find the concerned information can be searched. After different image data are switched, the known target image patterns can still be used for matching, and interesting patterns on different data are found.

The results computed by the pattern matching algorithm can also be presented more prominently (e.g., by sub-graph highlighting, table-to-main graph linkage, etc.), giving the user more opportunities to explore unknown special patterns.

In one or more embodiments of the present specification, fig. 6 is a schematic flowchart of a graph data processing method based on graph patterns in an application scenario, where after a user starts using the method, the user first selects a desired pattern from an innovative modeling pattern and a clone pattern, then enters a pattern editor, edits nodes, edges, types and attribute conditions included in the nodes and edges in the pattern editor by a first editing operation, a second editing operation, and the like in a manner of adding nodes or adding templates to obtain a target graph pattern, and matches the graph data by the target graph pattern to obtain a desired sub-graph, and generates a related result list and a statistical graph.

In one or more embodiments of the present description, the information in the graph data that includes the timing dimension is significant in many scenarios. For example, in a fund analysis scenario, the time on the edge generally represents the time when the fund is circulated, and the time-series analysis may help the user to understand the development of the fund network over time, discover the destination of the fund over time, and the like. In the social relationship network, the time information on the node may represent the time of one user registering an account, the time information on the side may represent the time of two users plus friends, and the like, and the time sequence analysis can help the users to know how the social community is organized and established, so that people can be encircled, decisions can be made, popularization can be performed, and the like.

For the time sequence information, it is usually presented to the user in a way of marking time character strings on nodes or sides, or a filter with a time axis as a time attribute. The first way of marking time strings on nodes or edges is only applicable to small-scale data of a plurality of points. The time text is generally long and may include year, month, day, hour, minute, second, even millisecond, the ultra-long text makes the user difficult to read quickly, and the phenomena of shielding and overlapping are likely to occur in the drawing of the relation graph, which greatly affects the readability of the graph. For the second approach with a timeline, a timeline similar to a progress bar is generally placed below the graph, and the user can filter out the nodes and edges in the graph that satisfy the time range by selecting the time range. Although this "dynamic" approach can help the user to observe the evolution process of the whole graph over time, it is difficult for the user to compare the time in a complete data set at a time.

In addition, the nodes and edges in the graph have some timing-related quantity attributes. For example, the time of transfer and the amount of money transferred in the money network; the time when two users in the social network add to a friend, the friend intimacy degree and the like. This information typically needs to be observed along with the timing information for analysis purposes. It is generally difficult to achieve this result in both of the above ways.

Based on this, fig. 7 is a flowchart of a data processing method based on a time sequence information diagram according to one or more embodiments of the present disclosure. The method can be applied to different business fields, such as a wind control field (such as risk modes of gambling, money laundering and the like in a wind control scene), an internet financial business field, an online social field (such as a special community mode in a social network), an e-commerce business field, an instant messaging business field, a game business field, a public business field and the like. The process can be executed by computing equipment in the corresponding field (such as an intelligent customer service server or an intelligent mobile terminal corresponding to the payment service, and the like), and certain input parameters or intermediate results in the process allow manual intervention and adjustment to help improve the accuracy.

The scheme shown in fig. 7 may be applied to the scheme shown in fig. 1, for example, the scheme is applied to the graph data displayed in the front end in step S102, so that the intuitiveness of the presentation of the timing information when the front end displays the graph data can be improved. Due to the particularity and importance of the time sequence information in the graph data, the graph data processing scheme based on the time sequence information is explained in detail below, so that the obtaining difficulty of the time sequence information in the graph data is reduced by improving the display intuitiveness of the time sequence information, and the generation efficiency and the accuracy of the time sequence information in the graph data can be improved when the target graph data is edited.

The flow in fig. 7 may include the following steps:

s702: and acquiring time sequence information of edges in the graph data.

Timing information refers to relevant information with a time attribute. The node and the edge may be values including a time attribute, which is a specific time value, for example, 1 month and 1 day in 2020, 10 o' clock and 20 min and 30 sec, or may be values of quantity attributes related to the time attribute, for example, the node specifies the occurrence frequency of an event in a certain time attribute, and the quantity of resources allocated according to the time attribute.

FIG. 8 is an illustration of a timing analysis configuration interface in an application scenario, as provided by one or more embodiments of the present disclosure; a timing analysis configuration interface is presented in the front end, which may be presented in the right region of the front end similar to the graph-pattern matching interface. To meet the different requirements of different scene data structures and attribute names, the user can select the time attribute name (e.g., payDate in fig. 8) and the quantity attribute name (e.g., quantity in fig. 8) in the graph data in the time sequence analysis configuration interface. If the user is not configured, some fields in the graph data can be automatically recommended to be used as the time attribute names and the quantity attribute names.

The time attribute and the quantity attribute of the edge can be analyzed in the graph data through the time attribute name and the quantity attribute name selected by the user, so that the value of the time attribute conforming to the selected time attribute name and the value of the quantity attribute conforming to the quantity attribute name can be found in the graph data.

Further, for the value of the quantitative attribute, it may be difficult to quickly and intuitively embody the actual meaning it represents by only a number. Based on this, as shown in fig. 8, a button "map edge thickness by number attribute" is provided in the time-series analysis configuration interface, and if the user clicks the button, elements with number attributes in the graph data are displayed in edges of the graph data at different thickness degrees according to the value of their own number attributes when displaying nodes and edges.

Fig. 9 is a schematic diagram of a time sequence analysis effect in an application scenario, provided by one or more embodiments of the present specification, that is, a width of a side marked with "2" in the middle is the thickest, and widths of sides marked with "1" and "5" on two sides are the thinnest, so that even if a value of a corresponding quantity attribute is not marked in the side, a user can determine the value in graph data intuitively, thereby improving user experience.

S704: and generating the hierarchical layout information of each node and the sequence number information of each edge according to the time sequence information.

The hierarchical layout information mainly represents the hierarchical relationship among the nodes, and the hierarchical relationship generally refers to the upper and lower hierarchical relationship among the nodes, for example, as shown in fig. 8, a root node is a movie node corresponding to a certain movie "X-Men 2" and has the highest hierarchical level, and nodes of the movie, such as "movie director", "movie country of sale", "movie company of sale", and the like, respectively belong to nodes at the next level of the movie node and are connected to the root node by edges.

The sequence number information indicates the time sequence relationship of each side in the time sequence information, and the smaller the sequence number is, the more ahead the time attribute corresponding to the side is. As shown in fig. 8, in addition to "map edge thickness by number attribute", the user can also configure whether "sort from 0 to N by time attribute". If the user clicks the button, the final presentation result is as shown in fig. 9, the series marked with "0" is determined as the creator of a certain movie first, and then the actor marked with "1" is determined as the movie, so that the sequence number information corresponding to each side is determined in sequence finally according to the time sequence information, and the presentation interface of the map data is more intuitive and understandable.

S706: and correspondingly and hierarchically displaying the nodes in the graph data and displaying the edges in the graph data with sequence numbers at the front end according to the hierarchical layout information and the sequence number information, wherein the sequence numbers indicate corresponding time sequences.

The hierarchical display means that the time sequence information can be displayed in a display interface in a form different from the traditional mode, and a user can acquire the time sequence information more intuitively.

The hierarchical layout information and the sequence number information are obtained through the time sequence information analysis, and the nodes and the edges can be displayed in the graph data in a hierarchical mode, so that a user can obtain the related content of the time sequence information more visually. Compared with the first traditional mode, the time information text is not directly marked on the elements in the graph data, the screen canvas space can be effectively saved, visual confusion caused by overlong and excessive texts is prevented, the readability and the analyzability are higher, and the deeper results brought by time sequence information can be further summarized and analyzed.

In one or more embodiments of the present specification, an edge in graph data is a directed edge, which can point to a chronological precedence relationship between nodes, so as to perform analysis on a chronological order.

The starting point of the earliest node or the earliest edge on the graph data is marked as a starting point node and represents the starting point in the time sequence network; all nodes without edges will be marked as end nodes, representing end points in the time sequence network, as shown in fig. 9, a start node can be marked as a start node, and a final node can be marked as an end node, so that a user can quickly know whether the nodes are the start node or the end node. And each terminal node and the starting node are connected in series by corresponding directed edges to form a plurality of directed paths.

The node hierarchy is divided based on the directed path, the starting point node is used as the first node hierarchy, and then the farther the distance is according to the distance between other nodes and the starting point node, the later the node hierarchy is, so that the node hierarchy can be divided more quickly and clearly. Of course, all nodes in the same directed path may also be divided into the same node hierarchy. Based on different requirements of users, the node hierarchy is divided in different ways.

Further, as shown in fig. 8, when performing the time-series analysis, not only the corresponding attributes can be selected, but also different sides can be grouped, so as to further improve the intuitiveness when performing the time-series analysis.

In each directed path, a plurality of upstream sides of a specified side are determined, the specified side is a side selected by a user or a side selected by default, and the upstream sides represent sides of a timing information before. A target upstream side is selected among the plurality of upstream sides according to the current scene and the value of the time attribute. If the quantity attribute of the upstream edge of the target is enough to be allocated for the quantity attribute corresponding to the specified edge, the quantity attribute is divided into the same group and is displayed according to the same style, for example, according to the same color and the same line. And, different styles are presented for different groups.

As shown in fig. 8, in the time sequence analysis configuration interface presented to the user, the manner of selecting the target upstream edge may include multiple manners, for example, grouping according to first-in first-out, grouping according to first-in last-out, grouping according to custom attributes, grouping according to fund staining, and the like, and grouping is performed in different manners according to the selection operation of the user. In first-in-first-out grouping, an edge tends to be considered as the same group as an upstream edge that occurs earliest, and the quantity attribute value margin of the upstream edge is sufficient to be assigned to the quantity attribute value of the current edge. And searching the upstream edges which occur next to the earliest until the quantity attribute value of the upstream edge which occurs earliest is distributed to be used as the same group. And by analogy, if the upstream edge which meets the condition does not exist, the upstream edges are independent into one group. The downstream edge looks for the same set of edges in the same way. In a first-in-last-out grouping, for an edge, it tends to be considered as the same group as the most recently occurring one of the upstream edges, and the margin of the number attribute value of the upstream edge is sufficient to be assigned the number attribute value of the current edge. And searching the next occurring upstream edge as the same group until the number attribute value of the recently occurring upstream edge is distributed. And by analogy, if the upstream edge which meets the condition does not exist, the upstream edges are independent into one group. The downstream edge looks for the same set of edges in the same way. In grouping by custom attributes, edges with equal attribute values are grouped into one group according to the edge attribute name specified by the user. In the fund-based staining grouping, the grouping is performed according to different sources, purposes and the like of the fund.

In one or more embodiments of the present specification, as shown in fig. 8, an interactive timeline (hereinafter referred to as timeline) is shown below the graph data, and is used for a user to interactively filter the data on the graph according to the time attribute. The time range and the single time point can be switched by a selector below the time axis. The user can drag and drop on the time axis to select time, and can click a play/pause button of the time axis to automatically play and pause.

Specifically, the user can perform the time-specifying operation through the time axis. The time designation operation refers to a user's filtering operation for a time point or a time range within the time axis. After the user selects the time point or the time range, an edge that matches the time point or the time range may be selected from the edges in the graph data (for example, whether the time attribute of the edge includes, is included, or crosses the time range or the time point, if so, it is considered to match the time attribute), and then the edge is prominently displayed, so that the user can pay attention to the edge.

Fig. 10 and 11 are schematic diagrams illustrating a filtering of time ranges and time points on an interaction time axis in an application scenario, where in fig. 10, a user selects a time range 10:07:49-11:59:19, six edges are total in accordance with the time range, and other edges are hidden, so that the six edges are prominently displayed. While in fig. 11 the 16:06:39 point in time is chosen, only four edges coinciding with this point in time are hidden from the others.

In one or more embodiments of the present specification, fig. 12 and fig. 13 are schematic diagrams of a list analysis interface and a chart analysis interface of an analysis result interface in an application scenario provided in one or more embodiments of the present specification. Also shown below the timing analysis configuration interface will be an analysis results interface, which can be presented in two ways: the table and the chart correspond to a result list and a statistical chart in the figure respectively.

After the analysis result is presented, if the user performs a statistical assignment operation therein, in the list analysis interface, the user may use a certain column as a basis for sorting, enter a row therein with a mouse, and present a corresponding element on the graph data in a highlighted manner (for example, highlighted presentation). In the chart analysis interface, the column data (e.g., the line graph of "value of quantity attribute" shown in fig. 13) is shown in the form of a graph (e.g., line graph, bar graph).

There is also a download icon in the upper right corner of the analysis result, and the user clicks the icon to download the current analysis result list, which may be a CSV file, a JSON file, or a data file in another adaptable format. The upper right corner of the time sequence analysis configuration interface is provided with a reset button, and a user clicks the reset button to clear the influence of the time sequence analysis tool on the canvas.

Compared with the traditional second method, the method not only provides the functions of screening, playing and the like of the time axis, but also can automatically calculate and separate out more time sequence related information; the multidimensional joint analysis can be carried out by combining the quantity attributes; reasonable visual mapping can be performed according to the time attribute and the quantity attribute so as to generate a more visual result; and a plurality of result display modes can be provided, and more visual angles are provided for the user.

Fig. 14 is a flowchart illustrating a method for processing graph data based on timing information in an application scenario, according to one or more embodiments of the present disclosure; after the user starts to inquire data, the time attribute names and the number attribute names are configured in the time sequence analysis interface, or automatically recommended attribute names can be adopted, so that time sequence analysis is carried out. And the data are displayed in the modes of mapping edge thickness according to quantity attributes, sorting from 0 to N according to time attributes, a time axis, a result list, a statistical chart and the like.

Based on the same idea, one or more embodiments of the present specification further provide apparatuses and devices corresponding to the above-described method, as shown in fig. 15 to 16.

Fig. 15 is a schematic structural diagram of a graph schema-based graph data processing apparatus according to one or more embodiments of the present specification, where the graph data includes a plurality of nodes and edges between the nodes, and the apparatus includes:

a front-end display module 1502 which displays the graph data on a configuration interface for configuring a graph mode at the front end, wherein the graph mode is used for designating a sub-graph structure;

the editing module 1504 receives the editing operation of adding nodes and edges by the user through the configuration interface;

a first generation module 1506 that generates a target graph pattern in response to the editing operation;

a matching module 1508, which performs matching in the graph data according to the target graph pattern to obtain a sub-graph in the graph data that matches the target graph pattern;

a saliency display module 1510, configured to perform saliency display on an area where the sub-graph is located in the graph data displayed at the front end, so that the user can pay attention to the sub-graph.

Optionally, the front end display module 1502 obtains timing information of nodes and edges in the graph data respectively;

and correspondingly displaying nodes in the graph data in a hierarchical mode and displaying edges in the graph data in a sequential mode at the front end according to the hierarchical layout information and the sequence number information, wherein the hierarchical layers and the sequence numbers indicate corresponding time sequences.

Optionally, the editing module 1504, entering a mode editor provided by the configuration interface;

receiving, in the schema editor, a first editing operation of a user to add a node, the first editing operation comprising specifying a type and/or attribute condition for the added node, and,

after the new nodes are determined, receiving a second editing operation of adding edges between the new nodes and the user, wherein the second editing operation comprises specifying type and/or attribute conditions for the added edges.

Optionally, the method further comprises:

a clone pattern module 1512 that determines whether the user has selected a create pattern or a clone pattern;

if the user selects the clone mode, displaying one or more existing graph modes in the clone mode, and receiving the selection operation of the user on the existing graph modes so as to edit and obtain the target graph mode according to the graph mode selected by the user.

Fig. 16 is a schematic structural diagram of an apparatus for processing graph data based on timing information according to one or more embodiments of the present specification, where the graph data includes a plurality of nodes and edges between the nodes, and the apparatus includes:

an obtaining module 1602, obtaining timing information of edges in the graph data;

a second generating module 1604, configured to generate hierarchical layout information of each node and sequence number information of each edge according to the timing information;

the hierarchical display module 1606 correspondingly hierarchically displays the nodes in the graph data and sequentially displays the edges in the graph data at the front end according to the hierarchical layout information and the sequence number information, where the sequence number indicates a corresponding time sequence.

Optionally, the timing information includes a value of a time attribute, and a value of a quantity attribute related to the time attribute;

the obtaining module 1602, displaying a timing analysis configuration interface at a front end;

receiving selection operation of time attributes and quantity attributes through the time sequence analysis configuration interface;

and acquiring the values of the time attribute and the numerical attribute of the edge in the graph data according to the selection operation.

Optionally, the hierarchy presentation module 1606 presents the edges in the graph data according to the value of the quantity attribute and the corresponding thickness degree.

Optionally, the edge in the graph data is a directed edge;

the hierarchy presentation module 1606 determines a start node and a destination node among nodes in the graph data according to a direction of an edge in the graph data;

correspondingly dividing the node hierarchy according to a directed path formed by directed edges between the starting node and the end node;

and displaying the nodes in the graph data at the front end according to the node hierarchy.

the hierarchy presentation module 1606, according to the directed path, determines a plurality of upstream edges of the specified edge in the directed path;

selecting a target upstream side from the plurality of upstream sides according to the current scene and the value of the time attribute;

judging whether the value of the quantity attribute of the target upstream edge is enough to be distributed as the value of the corresponding quantity attribute of the specified edge or not according to the value of the quantity attribute;

if so, dividing the target upstream edge and the specified edge into the same group, displaying the edges in the same group according to the same style, and displaying the edges in different groups according to different styles.

Optionally, the method further comprises:

an analysis configuration presentation module 1608 for presenting a timing analysis configuration interface at the front end;

receiving selection operation of a side grouping mode through the time sequence analysis configuration interface;

the hierarchy presentation module 1606 selects, according to the selection operation, the upstream edge with the earliest time as the target upstream edge, or selects the upstream edge with the latest time as the target upstream edge.

Optionally, the hierarchy presentation module 1606 presents an interaction timeline;

receiving a time designation operation through the interaction timeline;

and screening out corresponding edges from the edges in the graph data for showing in a significant manner according to the time points or the time ranges specified by the time specification operation so as to facilitate the attention of the user.

Optionally, the hierarchical display module 1606 displays an analysis result interface in the time sequence analysis configuration interface, where the analysis result interface includes a list analysis interface and/or a graph analysis interface;

receiving a statistical specified operation through the analysis result interface;

sorting the column data corresponding to the statistic designation operation in the list analysis interface according to the statistic designation operation, and displaying the row data corresponding to the statistic designation operation in the graph data in a significance manner,

in the chart analysis interface, the column data corresponding to the statistical specified operation is prominently displayed in a graph form.

One or more embodiments of the present specification provide a graph data processing apparatus based on a graph schema, the graph data including a plurality of nodes and edges between the nodes, the apparatus including:

at least one processor; and the number of the first and second groups,

a memory communicatively coupled to the at least one processor; wherein, the first and the second end of the pipe are connected with each other,

generating a target graph mode in response to the editing operation;

One or more embodiments of the present specification provide a graph data processing apparatus based on timing information, the graph data including a plurality of nodes and edges between the nodes, the apparatus including:

at least one processor; and the number of the first and second groups,

acquiring time sequence information of edges in the graph data;

Based on the same idea, one or more embodiments of the present specification further provide a non-volatile computer storage medium corresponding to the above method, and storing computer-executable instructions configured to:

generating a target graph mode in response to the editing operation;

acquiring time sequence information of edges in the graph data;

The above description is merely one or more embodiments of the present disclosure and is not intended to limit the present disclosure. Various modifications and alterations to one or more embodiments of the present description will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement or the like made within the spirit and principle of one or more embodiments of the present specification should be included in the scope of the claims of the present specification.

Claims

1. A method of processing graph data, the graph data comprising a plurality of nodes and edges between the nodes, the method comprising:

displaying the graph data on a configuration interface for configuring a graph mode at the front end, wherein the graph mode is used for designating a sub-graph structure;

generating a target graph mode in response to the editing operation;

matching in the graph data according to the target graph mode to obtain a subgraph matched with the target graph mode in the graph data;

2. The method according to claim 1, wherein the displaying the graph data at a front end specifically comprises:

acquiring time sequence information of nodes and edges in the graph data respectively;

3. The method according to claim 1, wherein the receiving, through the configuration interface, an editing operation of a user for adding a node and an edge specifically includes:

entering a mode editor provided by the configuration interface;

receiving, in the schema editor, a first editing operation of a user to add a node, the first editing operation comprising specifying a type and/or property condition for the added node, and,

4. The method of claim 3, prior to receiving a first edit operation by a user to add a node, the method further comprising:

determining whether the user has selected a create mode or a clone mode;

if the user selects the clone mode, displaying one or more existing graph modes in the clone mode, and receiving the selection operation of the user on the existing graph modes so as to edit the target graph mode according to the graph mode selected by the user.

5. A method of processing graph data, the graph data comprising a plurality of nodes and edges between the nodes, the method comprising:

acquiring time sequence information of edges in the graph data;

and displaying the nodes in the graph data in a corresponding hierarchy mode and displaying the edges in the graph data in a sequence mode at the front end according to the hierarchy layout information and the sequence number information, wherein the sequence numbers indicate corresponding time sequences.

6. The method of claim 5, the timing information comprising a value of a time attribute, and a value of a quantity attribute associated with the time attribute;

the acquiring of the timing information of the edge in the graph data specifically includes:

displaying a time sequence analysis configuration interface at the front end;

7. The method according to claim 6, wherein the displaying nodes in the graph data in a hierarchical manner and the displaying edges in the graph data with sequence numbers at a front end correspondingly comprises:

and displaying the edges in the graph data according to the value of the quantity attribute and the corresponding thickness degree.

8. The method of claim 5, wherein the edges in the graph data are directed edges;

the displaying of the nodes in the graph data in a corresponding hierarchy at the front end specifically includes:

determining a starting point node and an end point node in the nodes in the graph data according to the direction of the edge in the graph data;

correspondingly dividing node levels according to a directed path formed by directed edges between the starting point node and the end point node;

9. The method of claim 8, the timing information comprising a value of a time attribute, and a value of a quantity attribute associated with the time attribute;

the correspondingly hierarchically showing nodes in the graph data at the front end and showing edges in the graph data with sequence numbers specifically includes:

determining a plurality of upstream edges of the appointed edges in the directed path according to the directed path;

10. The method of claim 9, before the serially exposing edges in the graph data, the method further comprising:

displaying a time sequence analysis configuration interface at the front end;

the selecting the target upstream edge specifically includes:

according to the selection operation, the upstream edges with the earliest time are all selected as target upstream edges, or the upstream edges with the latest time are all selected as target upstream edges.

11. The method of claim 5, wherein the displaying nodes in the graph data and the edges in the graph data with sequence numbers in a front end correspondingly and hierarchically comprises:

displaying an interactive time axis;

receiving a time designation operation through the interaction timeline;

12. The method of claim 6, wherein the displaying nodes in the graph data and the edges in the graph data with sequence numbers in a front end correspondingly and hierarchically comprises:

displaying an analysis result interface in the time sequence analysis configuration interface, wherein the analysis result interface comprises a list analysis interface and/or a chart analysis interface;

13. An apparatus for processing graph data, the graph data including a plurality of nodes and edges between the nodes, the apparatus comprising:

and the significant display module is used for performing significant display on the area where the sub-graph is located in the graph data displayed at the front end so as to facilitate the user to pay attention to.

14. An apparatus for processing graph data, the graph data including a plurality of nodes and edges between the nodes, the apparatus comprising:

the second generation module generates the hierarchical layout information of each node and the serial number information of each edge according to the time sequence information;

and the level display module correspondingly displays the nodes in the graph data in a level mode at the front end and displays the edges in the graph data in a sequence mode according to the level layout information and the sequence number information, and the sequence number indicates a corresponding time sequence.

15. A graph data processing apparatus, said graph data comprising a plurality of nodes and edges between the nodes, said apparatus comprising:

at least one processor; and the number of the first and second groups,

generating a target graph mode in response to the editing operation;

16. A graph data processing apparatus, said graph data comprising a plurality of nodes and edges between the nodes, said apparatus comprising:

at least one processor; and the number of the first and second groups,

acquiring time sequence information of edges in the graph data;