Subgraph generation method and device based on artificial intelligence, computer equipment and medium

Abstract

The invention belongs to the field of artificial intelligence and the field of financial science and technology, and relates to an artificial intelligence-based subgraph generation method, which comprises the following steps of: obtaining text data input by a user; preprocessing the text data to obtain specified text data; constructing a corresponding word vector based on the specified text data; constructing a corresponding target map based on the word vector; performing optimization processing on a preset original ant colony algorithm based on a preset attribute heuristic factor and a genetic algorithm to obtain a corresponding target ant colony algorithm; performing sub-graph mining processing on the target graph based on the target ant colony algorithm to obtain a corresponding target sub-graph; and carrying out pushing processing on the user based on the target sub-graph. The invention further provides a subgraph generation device based on artificial intelligence, computer equipment and a storage medium. In addition, the target sub-graph can be stored in the block chain. According to the method, the processing efficiency and the processing accuracy of sub-graph mining can be effectively improved based on the use of the target ant colony algorithm.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence development technology and the field of financial technology, and in particular to artificial intelligence-based subgraph generation methods, devices, computer equipment and storage media. Background Art

[0002] In the context of the rapid development of big data and artificial intelligence, the processing and analysis of text data has become a core task in the financial industry. With the explosive growth of data volume, how to quickly and accurately extract valuable information from massive text data has become the focus of researchers and technology developers. Text data mining technology, especially the processing of complex data structures, such as subgraph mining using graph models, has become a research hotspot in the financial industry.

[0003] At present, subgraph mining for text data mainly adopts methods such as LB (Label-Based) and CBA (Candidate-Based Aggregation). These methods can realize text information mining and subgraph retrieval to a certain extent, but they have obvious limitations. Specifically, the LB method generates subgraphs through label matching. Although it is intuitive and easy to understand, its performance is seriously affected when processing large-scale text data, because the diversity and complexity of labels will significantly increase the computational cost, resulting in slower running speed. In addition, due to the possible ambiguity and overlap of labels, this method is also insufficient in terms of accuracy.

[0004] On the other hand, the CBA method realizes subgraph mining by generating and aggregating candidate subgraphs. When generating candidate subgraphs, this method needs to consider multiple possible combinations, which dramatically increases the number of candidate subgraphs and significantly increases the computational complexity. At the same time, since the CBA method relies on multiple iterations and aggregation operations, it has a high dependence on data and is prone to noise and errors, resulting in less than ideal accuracy of the final result.

[0005] Therefore, the existing subgraph mining of text data using LB and CBA methods has technical problems of low processing efficiency and low accuracy when facing large-scale and complex text data. Summary of the invention

[0006] The purpose of the embodiments of the present application is to propose a subgraph generation method, device, computer equipment and storage medium based on artificial intelligence to solve the technical problems of low processing efficiency and low accuracy in the existing subgraph mining of text data using LB and CBA methods.

[0007] In order to solve the above technical problems, the embodiment of the present application provides a subgraph generation method based on artificial intelligence, which adopts the following technical solution:

[0008] Get the text data to be processed input by the user;

[0009] Preprocessing the text data to obtain corresponding specified text data;

[0010] Construct a corresponding word vector based on the specified text data;

[0011] Construct a corresponding target graph based on the word vector;

[0012] Based on the preset attribute heuristic factors and genetic algorithm, the preset original ant colony algorithm is optimized to obtain the corresponding target ant colony algorithm;

[0013] Performing subgraph mining processing on the target graph based on the target ant colony algorithm to obtain a corresponding target subgraph;

[0014] Push processing is performed on the user based on the target subgraph.

[0015] Furthermore, the step of optimizing the preset original ant colony algorithm based on the preset attribute heuristic factor and the genetic algorithm to obtain the corresponding target ant colony algorithm specifically includes:

[0016] Generating the attribute inspiration factor based on a preset Bayesian formula;

[0017] Adding the attribute heuristic factor to the probability formula in the original ant colony algorithm to obtain a corresponding first ant colony algorithm;

[0018] Obtaining a crossover operator in the genetic algorithm;

[0019] Adjusting the volatility coefficient in the first ant colony algorithm based on the crossover operator to obtain an adjusted second ant colony algorithm;

[0020] The target ant colony algorithm is constructed based on the second ant colony algorithm.

[0021] Furthermore, the step of performing subgraph mining processing on the target graph based on the target ant colony algorithm to obtain the corresponding target subgraph specifically includes:

[0022] Performing subgraph mining processing on the target graph based on the target ant colony algorithm to obtain all corresponding iteration records;

[0023] For each of the iteration records, extracting a saved optimal subgraph and a quality assessment score corresponding to the optimal subgraph from the iteration record;

[0024] Numerically comparing the quality evaluation scores of all the optimal subgraphs, and selecting the global optimal subgraph with the highest quality evaluation score;

[0025] The global optimal subgraph is used as the target subgraph.

[0026] Furthermore, the step of performing push processing on the user based on the target subgraph specifically includes:

[0027] Performing visualization processing on the target sub-graph to obtain a corresponding sub-graph image;

[0028] Acquire attribute information of a target node in the target subgraph, and acquire a node connection relationship corresponding to the target node;

[0029] Generate target description information corresponding to the target subgraph based on the attribute information and the node connection relationship;

[0030] The sub-image and the target description information are pushed to the user.

[0031] Furthermore, the step of constructing a corresponding target graph based on the word vector specifically includes:

[0032] Obtain preset graph strategies and graph optimization strategies;

[0033] Processing the word vector based on the graph strategy to obtain corresponding first graph data;

[0034] Optimizing the first graph data based on the graph optimization strategy to obtain corresponding second graph data;

[0035] The second graph data is used as the target graph.

[0036] Furthermore, the step of constructing a corresponding word vector based on the specified text data specifically includes:

[0037] Calling preset natural language processing tools;

[0038] Performing vector conversion processing on the specified text data based on the natural language processing tool to obtain corresponding conversion data;

[0039] The converted data is used as the word vector.

[0040] Furthermore, the step of preprocessing the text data to obtain corresponding specified text data specifically includes:

[0041] Performing denoising processing on the text data to obtain corresponding first text data;

[0042] Performing word segmentation processing on the first text data to obtain corresponding second text data;

[0043] Performing stop word filtering processing on the second text data to obtain corresponding third text data;

[0044] The third text data is used as the designated text data.

[0045] In order to solve the above technical problems, the embodiment of the present application further provides a sub-graph generation device based on artificial intelligence, which adopts the following technical solution:

[0046] An acquisition module is used to acquire text data to be processed input by a user;

[0047] A preprocessing module, used for preprocessing the text data to obtain corresponding specified text data;

[0048] A first construction module, used to construct a corresponding word vector based on the specified text data;

[0049] A second construction module is used to construct a corresponding target graph based on the word vector;

[0050] The optimization module is used to optimize the preset original ant colony algorithm based on the preset attribute heuristic factor and the genetic algorithm to obtain the corresponding target ant colony algorithm;

[0051] A mining module, used for performing subgraph mining processing on the target graph based on the target ant colony algorithm to obtain a corresponding target subgraph;

[0052] A push module is used to push the user based on the target subgraph.

[0053] In order to solve the above technical problems, the embodiment of the present application further provides a computer device, which adopts the following technical solution:

[0054] Get the text data to be processed input by the user;

[0055] Preprocessing the text data to obtain corresponding specified text data;

[0056] Construct a corresponding word vector based on the specified text data;

[0057] Construct a corresponding target graph based on the word vector;

[0058] Based on the preset attribute heuristic factors and genetic algorithm, the preset original ant colony algorithm is optimized to obtain the corresponding target ant colony algorithm;

[0059] Performing subgraph mining processing on the target graph based on the target ant colony algorithm to obtain a corresponding target subgraph;

[0060] Push processing is performed on the user based on the target subgraph.

[0061] In order to solve the above technical problems, the embodiment of the present application further provides a computer-readable storage medium, which adopts the following technical solution:

[0062] Get the text data to be processed input by the user;

[0063] Preprocessing the text data to obtain corresponding specified text data;

[0064] Construct a corresponding word vector based on the specified text data;

[0065] Construct a corresponding target graph based on the word vector;

[0066] Based on the preset attribute heuristic factors and genetic algorithm, the preset original ant colony algorithm is optimized to obtain the corresponding target ant colony algorithm;

[0067] Performing subgraph mining processing on the target graph based on the target ant colony algorithm to obtain a corresponding target subgraph;

[0068] Push processing is performed on the user based on the target subgraph.

[0069] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0070] The present application first obtains the text data to be processed input by the user; then preprocesses the text data to obtain the corresponding specified text data; then constructs the corresponding word vector based on the specified text data; and constructs the corresponding target graph based on the word vector; subsequently optimizes the preset original ant colony algorithm based on the preset attribute heuristic factor and genetic algorithm to obtain the corresponding target ant colony algorithm; further performs subgraph mining on the target graph based on the target ant colony algorithm to obtain the corresponding target subgraph; and finally pushes the user based on the target subgraph. The present application obtains the specified text data by preprocessing the text data input by the user, then constructs the corresponding target graph based on the specified text data, then optimizes the original ant colony algorithm based on the attribute heuristic factor and genetic algorithm to obtain the target ant colony algorithm, and then performs subgraph mining on the target graph based on the target ant colony algorithm, which can quickly and accurately complete the subgraph mining processing of the text data, effectively improves the processing efficiency and processing accuracy of the subgraph mining, and effectively guarantees the accuracy of the generated target subgraph. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] In order to more clearly illustrate the scheme in the present application, a brief introduction is given below to the drawings required for use in the description of the embodiments of the present application. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0072] Figure 1 is an exemplary system architecture diagram to which the present application may be applied;

[0073] Figure 2 A flowchart of an embodiment of a subgraph generation method based on artificial intelligence according to the present application;

[0074] Figure 3 is a structural schematic diagram of an embodiment of a sub-graph generating device based on artificial intelligence according to the present application;

[0075] Figure 4 It is a structural diagram of an embodiment of a computer device according to the present application. DETAILED DESCRIPTION

[0076] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of the present application; the terms used in the specification of the application herein are only for the purpose of describing specific embodiments and are not intended to limit the present application; the terms "including" and "having" and any variations thereof in the specification and claims of the present application and the above-mentioned drawings are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of the present application or the above-mentioned drawings are used to distinguish different objects, not to describe a specific order.

[0077] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0078] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.

[0079] like Figure 1As shown, the system architecture 100 may include a terminal device 101, a network 102 and a server 103. The terminal device 101 may be a laptop 1011, a tablet computer 1012 or a mobile phone 1013. The network 102 is used to provide a medium for a communication link between the terminal device 101 and the server 103. The network 102 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0080] The user can use the terminal device 101 to interact with the server 103 through the network 102 to receive or send messages, etc. Various communication client applications can be installed on the terminal device 101, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0081] The terminal device 101 can be any electronic device with a display screen and supporting web browsing. In addition to the laptop computer 1011, tablet computer 1012 or mobile phone 1013, the terminal device 101 can also be an e-book reader, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 (Moving Picture Experts Group Audio Layer IV), a laptop computer, a desktop computer, etc.

[0082] The server 103 may be a server that provides various services, such as a background server that provides support for a web page displayed on the terminal device 101 .

[0083] It should be noted that the artificial intelligence-based sub-graph generation method provided in the embodiment of the present application is generally executed by a server / terminal device, and accordingly, the artificial intelligence-based sub-graph generation device is generally set in the server / terminal device.

[0084] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0085] Continue to refer Figure 2, showing a flowchart of an embodiment of the sub-graph generation method based on artificial intelligence according to the present application. According to different requirements, the order of the steps in the flowchart can be changed, and some steps can be omitted. The sub-graph generation method based on artificial intelligence provided in the embodiment of the present application can be applied to any scenario where image generation is required, and the sub-graph generation method based on artificial intelligence can be applied to products in these scenarios, for example, image generation in the financial insurance field. The sub-graph generation method based on artificial intelligence includes the following steps:

[0086] Step S201, obtaining text data to be processed input by the user.

[0087] In this embodiment, the electronic device (eg Figure 1 The server / terminal device shown in the figure) can obtain text data through a wired connection or a wireless connection. It should be noted that the above-mentioned wireless connection method may include but is not limited to 3G / 4G / 5G connection, WiFi connection, Bluetooth connection, WiMAX connection, Zigbee connection, UWB (ultra wideband) connection, and other wireless connection methods currently known or to be developed in the future. The executor of the present application may be a sub-graph generation system, or simply referred to as a system. In the business scenario of finance and insurance, the above-mentioned text data may include data in text form related to transaction data, payment data, business data and other content summarized in the financial field.

[0088] Step S202: pre-process the text data to obtain corresponding specified text data.

[0089] In this embodiment, the above-mentioned specific implementation process of preprocessing the text data to obtain the corresponding specified text data will be further described in detail in subsequent specific embodiments of the present application, and will not be elaborated on here.

[0090] Step S203, constructing a corresponding word vector based on the specified text data.

[0091] In this embodiment, the specific implementation process of constructing the corresponding word vector based on the specified text data will be further described in detail in subsequent specific embodiments of the present application and will not be elaborated on here.

[0092] Step S204: construct a corresponding target graph based on the word vector.

[0093] In this embodiment, the specific implementation process of constructing the corresponding target graph based on the word vector will be further described in detail in subsequent specific embodiments of the present application and will not be elaborated here.

[0094] Step S205 , optimizing the preset original ant colony algorithm based on the preset attribute heuristic factor and the genetic algorithm to obtain a corresponding target ant colony algorithm.

[0095] In this embodiment, the above-mentioned specific implementation process of optimizing the preset original ant colony algorithm based on the preset attribute inspiration factor and the genetic algorithm to obtain the corresponding target ant colony algorithm will be further described in detail in the subsequent specific embodiments of this application, and will not be elaborated in detail here.

[0096] Step S206: performing subgraph mining processing on the target graph based on the target ant colony algorithm to obtain a corresponding target subgraph.

[0097] In this embodiment, the above-mentioned specific implementation process of performing subgraph mining on the target graph based on the target ant colony algorithm to obtain the corresponding target subgraph will be further described in detail in the subsequent specific embodiments of this application, and will not be elaborated on here. Among them, in the process of using the target ant colony algorithm to perform subgraph mining on the target graph, after each iteration, the optimal rule (subgraph) can be formed according to the path crawled by the ant colony, and the samples covered by the selected optimal rule will be deleted from the sample set, and the next iteration will be entered. This process ends when the remaining samples in the training set are less than the initially set threshold, that is, less than a fixed value, and the subgraph mining process is completed.

[0098] Step S207: Push processing is performed on the user based on the target subgraph.

[0099] In this embodiment, the specific implementation process of the push processing to the user based on the target subgraph will be described in further detail in subsequent specific embodiments of the present application, and will not be elaborated on here.

[0100] The present application first obtains the text data to be processed input by the user; then preprocesses the text data to obtain the corresponding specified text data; then constructs the corresponding word vector based on the specified text data; and constructs the corresponding target graph based on the word vector; subsequently optimizes the preset original ant colony algorithm based on the preset attribute heuristic factor and genetic algorithm to obtain the corresponding target ant colony algorithm; further performs subgraph mining on the target graph based on the target ant colony algorithm to obtain the corresponding target subgraph; and finally pushes the user based on the target subgraph. The present application obtains the specified text data by preprocessing the text data input by the user, then constructs the corresponding target graph based on the specified text data, then optimizes the original ant colony algorithm based on the attribute heuristic factor and genetic algorithm to obtain the target ant colony algorithm, and then performs subgraph mining on the target graph based on the target ant colony algorithm, which can quickly and accurately complete the subgraph mining processing of the text data, effectively improves the processing efficiency and processing accuracy of the subgraph mining, and effectively guarantees the accuracy of the generated target subgraph.

[0101] In some optional implementations, step S205 includes the following steps:

[0102] The attribute heuristic factor is generated based on a preset Bayesian formula.

[0103] In this embodiment, the corresponding attribute heuristic factor may be generated according to the class conditional probability in the above Bayesian formula.

[0104] The attribute heuristic factor is added to the probability formula in the original ant colony algorithm to obtain the corresponding first ant colony algorithm.

[0105] In this embodiment, a big data graph G of the retrieval object is prepared in advance, and it is assumed that the big data graph G has g sub-categories. Secondly, the ant colony is divided into g categories and the rules of each category are mined respectively. The antecedent (rule) initially carried by each ant is empty, and all feasible terms are regarded as nodes. The nodes are selected by the probability selection formula to be added to the path (rule), and the process stops when no more can be added to form a rule. In the target ant colony algorithm, when the qth ant mines the rule of the pth category, the probability of selecting the lth value of the kth attribute is according to the following formula:

[0106]

[0107] Among them, τ pkl (t) is the pheromone concentration between categories p and q at time t in the ant colony algorithm, η pkl(t) is the heuristic information in the ant colony algorithm, ∝ is the pheromone factor, β is the heuristic function factor, γ is the attribute heuristic function factor, and allowedp,q is the set of cities that the ants are allowed to visit next. The above target ant colony algorithm is obtained by adding ξ to the probability formula of the traditional ant colony algorithm (original ant colony algorithm). pk , that is, the attribute heuristic factor, is used to measure the importance of attribute k to the pth class. Specifically, by introducing the class conditional probability in the Bayesian formula, that is:

[0108] Z p (a) = ξ p0 +ξ p1 a1+...+ξ ph a h

[0109] Z p (a) is the attribute of an instance object of class p (a=(a1,...a h )) score, coefficient ξ pk Reflects the importance of attribute k to class p. Among them:

[0110]

[0111] Among them, a j is the mean of attribute j of class p, s 2 kj is the covariance of class p attributes k, j. Assume that in each class of samples, each attribute obeys Gaussian distribution, that is, class p attribute k obeys:

[0112]

[0113] Where x is a random variable, μ is the mean, σ is the standard deviation, exp is the base of the natural logarithm, and π is the circumference of a circle. By adding ξ to the probability formula of the ant colony algorithm pk Attribute heuristic factors can effectively improve algorithm performance.

[0114] In the traditional ant colony algorithm, the order of attribute selection is fixed. However, for multi-label problems, the roles of attributes in different labels are different. Using a fixed attribute selection order will lead to reduced algorithm performance and weaker transfer learning capabilities. This embodiment considers that different attributes contribute differently in category differentiation. This embodiment adds attribute inspiration factors to the traditional ant colony probability selection formula. For different classes, by analyzing the information content of different attributes in each class (obtained through the Bayesian discrimination coefficient), the attributes are ranked from high to low according to the size of the information content, and the node selection order is provided to the ants. Compared with the traditional indiscriminate treatment of each attribute, the improvement can effectively improve the accuracy and efficiency of rule formation;

[0115] Obtain a crossover operator in the genetic algorithm.

[0116] In this embodiment, the volatility coefficient in the ant colony algorithm is also called the pheromone volatility factor.

[0117] The volatility coefficient in the first ant colony algorithm is adjusted based on the crossover operator to obtain an adjusted second ant colony algorithm.

[0118] In this embodiment, when all ants complete a path selection, multiple rules will be formed, from which the best rule is selected to update the pheromone. The update formula is as follows:

[0119] Initially, the pheromone concentration of the p-type attribute k with a value of l is equal to its heuristic factor value, which is equal to the ratio of the number of cases covered by the p-type attribute k with a value of l to the number of all cases in the sample, that is: After that, Among them, ρ is the volatility coefficient, which indicates that past information will evaporate over time; S and C are the support and credibility of the added rules respectively. The higher the support and credibility, the higher the quality of the rules, which has a positive feedback effect on the subsequent ants' selection of the node. In the traditional ant colony algorithm (i.e., the original ant colony algorithm), ρ remains unchanged. The selection of ρ is very important. When ρ is too large, it is easy to cause local optimality, and when ρ is too small, it will lead to too slow convergence. Therefore, this embodiment uses the crossover operator in the genetic algorithm to improve this so that it can be dynamically adjusted. Specifically, Among them, ρ min Set to 0.6, ρ max Set according to the maximum number of iterations, g is the current number of iterations, and G is the maximum number of iterations in the evolution process:

[0120]

[0121] In the traditional ant colony algorithm, the pheromone concentration is fixed. However, if the volatility factor is set too high, it will easily lead to a local optimum, and if it is set too low, it will lead to a slow convergence speed. Therefore, it is necessary to dynamically adjust the volatility factor. This embodiment introduces a crossover operator to dynamically adjust the volatility concentration. At the beginning of the iteration, the volatility factor value is high, and the model can iterate quickly; in the later stage of the iteration, the volatility factor value becomes smaller, and the iteration speed slows down to avoid the local optimum, taking into account the balance between model effect and operation efficiency.

[0122] The target ant colony algorithm is constructed based on the second ant colony algorithm.

[0123] The present application generates the attribute heuristic factor based on a preset Bayesian formula; then adds the attribute heuristic factor to the probability formula in the original ant colony algorithm to obtain the corresponding first ant colony algorithm; then obtains the crossover operator in the genetic algorithm; subsequently adjusts the volatility coefficient in the first ant colony algorithm based on the crossover operator to obtain the adjusted second ant colony algorithm; finally, the target ant colony algorithm is constructed based on the second ant colony algorithm. The present application generates the attribute heuristic factor based on a preset Bayesian formula, adds the attribute heuristic factor to the probability formula in the original ant colony algorithm to obtain the corresponding first ant colony algorithm, and adjusts the volatility coefficient in the first ant colony algorithm based on the crossover operator in the genetic algorithm, so as to realize the intelligent optimization of the original ant colony algorithm. Compared with the traditional ant colony algorithm, the accuracy and efficiency of rule formation can be improved, and the balance between model effect and operation efficiency can be taken into account. The subsequent subgraph mining process of the target graph using the optimized target ant colony algorithm can effectively improve the efficiency and accuracy of subgraph retrieval.

[0124] In some optional implementations of this embodiment, step S206 includes the following steps:

[0125] Based on the target ant colony algorithm, subgraph mining is performed on the target graph to obtain all corresponding iteration records.

[0126] In this embodiment, the target graph is subjected to subgraph mining by using the target ant colony algorithm, that is, after the target graph is subjected to solution space construction and iterative optimization processing of the ant colony search process by using the target ant colony algorithm, the optimal subgraph in each iteration (i.e., the subgraph with the highest quality evaluation function score) is recorded and a corresponding iteration record is generated.

[0127] Specifically, the above-mentioned construction of the solution space includes: defining the solution space as the set of all possible subgraphs of the graph. Then designing a subgraph quality evaluation function, taking into account factors such as the average similarity between nodes in the subgraph, the size of the subgraph, and the closeness of the connection between nodes. The above-mentioned ant colony search process includes: each ant starts from a random node, selects the next node using a probabilistic transfer rule based on pheromones and heuristic information (such as the similarity between nodes), and constructs a subgraph. Ants release pheromones according to the quality of the subgraph during movement. After completing a round of search, the pheromone matrix is ​​updated, taking into account pheromone evaporation and pheromones released by ants. The above-mentioned iterative optimization process includes: repeating the ant colony search process until a preset number of iterations is reached or a convergence condition is met (such as the change in the pheromone matrix is ​​less than a certain threshold). And record the optimal subgraph in each iteration (that is, the subgraph with the highest score in the quality evaluation function).

[0128] For each of the iteration records, a saved optimal subgraph and a quality assessment score corresponding to the optimal subgraph are extracted from the iteration record.

[0129] In this embodiment, in all iteration records of the target ant colony algorithm, each step will record the optimal subgraph generated by the current iteration. After the target ant colony algorithm ends, the global optimal subgraph can be extracted through all iteration records. Specifically, all iteration records can be traversed. For each iteration record, the optimal subgraph saved in the iteration record and its corresponding quality evaluation score are extracted.

[0130] The quality evaluation scores of all the optimal subgraphs are numerically compared to screen out the global optimal subgraph with the highest quality evaluation score.

[0131] In this embodiment, the quality evaluation scores of all optimal subgraphs may be compared, and the subgraph with the highest quality evaluation score may be found as the global optimal subgraph.

[0132] The global optimal subgraph is used as the target subgraph.

[0133] In this embodiment, the above-mentioned global optimal subgraph and its related information (such as node set, edge set, classification label, etc.) can also be saved for subsequent use.

[0134] The present application performs subgraph mining on the target graph based on the target ant colony algorithm to obtain all corresponding iteration records; then for each iteration record, extracts the saved optimal subgraph and the quality assessment score corresponding to the optimal subgraph from the iteration record; then numerically compares the quality assessment scores of all the optimal subgraphs to screen out the global optimal subgraph with the highest quality assessment score; and subsequently uses the global optimal subgraph as the target subgraph. The present application performs subgraph mining on the target graph based on the use of the target ant colony algorithm, and extracts the saved optimal subgraph and the quality assessment score corresponding to the optimal subgraph from all the generated iteration records, and then numerically compares the quality assessment scores of all the optimal subgraphs to screen out the global optimal subgraph with the highest quality assessment score as the final target subgraph, which effectively improves the generation efficiency of the target subgraph and ensures the data accuracy of the obtained target subgraph.

[0135] In some optional implementations, step S207 includes the following steps:

[0136] The target sub-graph is visualized to obtain a corresponding sub-graph image.

[0137] In this embodiment, the target subgraph can be visualized by using a graphics library to more intuitively display the structure and characteristics of the optimal subgraph. Among them, the above-mentioned graphics library can adopt a library with visualization functions. Specifically, by creating a new graph object, and adding nodes and edges to the graph according to the node and edge information of the target subgraph. Then set the style of the nodes and edges (such as color, size, label, etc.) to better display the structure and characteristics of the subgraph. Subsequently, use the drawing function of the imported graphics library to draw to generate an image of the above-mentioned target subgraph, that is, the above-mentioned subgraph image.

[0138] Acquire the attribute information of the target node in the target subgraph, and acquire the node connection relationship corresponding to the target node.

[0139] In this embodiment, the basis for text classification can be effectively explained by analyzing the nodes (ie, words) in the target subgraph and the connection relationship between the nodes.

[0140] Target description information corresponding to the target subgraph is generated based on the attribute information and the node connection relationship.

[0141] In this embodiment, by checking the target nodes in the target subgraph, these nodes represent the key words in the text. Then analyze the connection relationship between the target nodes, and the closely connected nodes may belong to the same theme or category. Afterwards, according to the attribute information and connection relationship of the target nodes, infer the main content and classification basis of the text to obtain the above-mentioned target description information. Among them, the above-mentioned target description information includes information such as keywords, classification labels, connection relationships, etc. of the nodes in the target subgraph. In addition, the frequency of occurrence, importance, etc. of each target node in the target subgraph can be further counted to support the interpretation of the classification results, thereby realizing the enrichment of the target description information.

[0142] The sub-image and the target description information are pushed to the user.

[0143] In this embodiment, the information push processing for the user can be completed by outputting the sub-image and the target description information to a user interface.

[0144] The present application obtains the corresponding subgraph image by visualizing the target subgraph; then obtains the attribute information of the target node in the target subgraph, and obtains the node connection relationship corresponding to the target node; then generates the target description information corresponding to the target subgraph based on the attribute information and the node connection relationship; and subsequently pushes the subgraph image and the target description information to the user. The present application performs subgraph mining on the target graph based on the target ant colony algorithm, and after obtaining the corresponding target subgraph, it also intelligently provides a detailed explanation for the target subgraph based on the attribute information of the target node in the target subgraph and the node connection relationship corresponding to the target node, and visualizes the target subgraph to obtain a subgraph image, so that by subsequently pushing the subgraph image and the target description information to the user, it can help the user understand and evaluate the classification effect of the input text data, improve the intelligence of data push, and help improve the user experience.

[0145] In some optional implementations, step S204 includes the following steps:

[0146] Get the preset graph strategy and graph optimization strategy.

[0147] In this embodiment, the strategy content of the above-mentioned graph strategy includes node definition processing, edge construction processing and weight setting processing. Specifically, the node definition processing includes: creating an empty graph structure. Then traverse all word vectors. For each word vector, it is regarded as a new node in the graph and a unique identifier (such as an index, hash value or custom ID) is used to mark the node. The node is subsequently added to the graph, and the mapping between the node ID and the word vector may be stored in a dictionary for subsequent use. The edge construction processing includes: traversing all word vector combinations. For each pair of word vectors, the similarity between them is calculated using the cosine similarity formula. Cosine similarity can be calculated by using the dot function and the norm function. Then set a similarity threshold, and only when the similarity is higher than the threshold, build an edge for this pair of word vectors in the graph. Subsequently, the similarity value is used as the weight of the edge and the edge is added to the graph. The weight setting processing includes: while building the edge, the calculated similarity value is directly used as the weight of the edge. Among them, the add_edge method can be used to add weighted edges by specifying the weight parameter.

[0148] The strategy content of the above-mentioned graph optimization strategy includes edge filtering processing and graph representation processing. Specifically, edge filtering processing includes: traversing all edges in the graph. Check the weight of each edge. If the weight is lower than a preset filtering threshold (this threshold may be different from the threshold used when the edge is constructed, which is used to further reduce the complexity of the graph), delete the edge from the graph. Among them, the remove_edge method can be used to delete the edge. The graph representation processing includes: selecting a suitable representation method according to the sparsity or density of the graph. If the graph is sparse (that is, the number of edges is much less than the square of the number of nodes), it may be more efficient to use the adjacency list representation. Among them, the graph is stored in the form of an adjacency list by default, so usually no additional work is required. If the graph is dense, or an adjacency matrix representation is required for some reason, the graph can be converted into an adjacency matrix. The to_numpy_array method can be used to obtain the adjacency matrix representation of the graph.

[0149] The word vector is processed based on the graph strategy to obtain corresponding first graph data.

[0150] In this embodiment, the word vector may be processed based on the strategy content included in the above-mentioned graph strategy to obtain the corresponding first graph data.

[0151] The first graph data is optimized based on the graph optimization strategy to obtain corresponding second graph data.

[0152] In this embodiment, the first graph data may be optimized according to the strategy content included in the above graph optimization strategy, so as to obtain the corresponding second graph data.

[0153] The second graph data is used as the target graph.

[0154] The present application obtains a preset graphing strategy and a graph optimization strategy; then processes the word vector based on the graphing strategy to obtain corresponding first graph data; then optimizes the first graph data based on the graph optimization strategy to obtain corresponding second graph data; and subsequently uses the second graph data as the target graph. The present application processes the word vector based on the use of a graphing strategy to obtain first graph data, and then optimizes the first graph data based on a graph optimization strategy to obtain second graph data as the corresponding target graph, thereby combining the use of a graphing strategy and a graph optimization strategy to realize fast and accurate construction of the target graph corresponding to the word vector, effectively improving the efficiency of target graph construction and ensuring the data accuracy of the obtained target graph.

[0155] In some optional implementations of this embodiment, step S203 includes the following steps:

[0156] Invoke the preset natural language processing tool.

[0157] In this embodiment, the above-mentioned natural language processing tool can specifically adopt a tool with the function of converting words in natural language into vector representation, so that the computer can understand natural language text. The natural language processing tool learns the co-occurrence pattern of words in a large amount of text and represents each word as a vector in a high-dimensional space. These vectors capture the semantic and grammatical relationship between words. This representation method makes similar words close in position in the vector space, thereby realizing semantic analysis and processing of the text.

[0158] Based on the natural language processing tool, vector conversion processing is performed on the specified text data to obtain corresponding conversion data.

[0159] In this embodiment, a natural language processing tool may be pre-trained, and then the pre-processed specified text data may be subjected to vector conversion processing to convert each word contained in the specified text data into a vector in a high-dimensional space to obtain corresponding converted data.

[0160] The converted data is used as the word vector.

[0161] This application calls a preset natural language processing tool; then performs vector conversion processing on the specified text data based on the natural language processing tool to obtain corresponding conversion data; and subsequently uses the conversion data as the word vector. This application performs vector conversion processing on the specified text data based on the use of a natural language processing tool, which can quickly and accurately construct a word vector corresponding to the specified text data, effectively improving the efficiency and intelligence of word vector construction.

[0162] In some optional implementations of this embodiment, step S202 includes the following steps:

[0163] The text data is subjected to denoising processing to obtain corresponding first text data.

[0164] In this embodiment, the denoising process may include using regular expressions or other string processing methods to remove HTML tags, special symbols, etc.

[0165] The first text data is processed by word segmentation to obtain corresponding second text data.

[0166] In this embodiment, the first text data may be segmented using a common segmentation tool, wherein an exact mode, a full mode or a search engine mode may be selected.

[0167] Stop word filtering is performed on the second text data to obtain corresponding third text data.

[0168] In this embodiment, the corresponding third text data can be obtained by loading a pre-built stop word list and filtering out the word segmentation results, that is, the stop words in the second text data according to the use of the stop word list.

[0169] The third text data is used as the designated text data.

[0170] The present application performs denoising processing on the text data to obtain the corresponding first text data; then performs word segmentation processing on the first text data to obtain the corresponding second text data; then performs stop word filtering processing on the second text data to obtain the corresponding third text data; and subsequently uses the third text data as the designated text data. The present application performs denoising processing, word segmentation processing, and stop word filtering processing on the text data, thereby realizing intelligent and accurate preprocessing of the text data, effectively ensuring the accuracy of the generated designated text data, and thus facilitating the data accuracy of the target subgraph generated subsequently.

[0171] In some optional implementations, the user information obtained is subject to the user's consent and complies with relevant laws and policies.

[0172] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0173] It should be emphasized that in order to further ensure the privacy and security of the above-mentioned target subgraph, the above-mentioned target subgraph can also be stored in a node of a blockchain.

[0174] The blockchain referred to in this application is a new application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, encryption algorithm, etc. Blockchain is essentially a decentralized database, a string of data blocks generated by cryptographic methods. Each data block contains a batch of network transaction information, which is used to verify the validity of its information (anti-counterfeiting) and generate the next block. Blockchain can include the underlying blockchain platform, platform product service layer, and application service layer.

[0175] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0176] AI basic technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. AI software technologies mainly include computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0177] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through computer-readable instructions, and the computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the aforementioned storage medium can be a non-volatile storage medium such as a disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0178] It should be understood that, although the steps in the flowchart of the accompanying drawings are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

[0179] Further references Figure 3 , as a response to the above Figure 2 The present application provides an embodiment of a sub-graph generation device based on artificial intelligence, and the device embodiment is similar to Figure 2 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0180] like Figure 3 As shown, the artificial intelligence-based subgraph generation device 300 described in this embodiment includes: an acquisition module 301, a preprocessing module 302, a first construction module 303, a second construction module 304, an optimization module 305, a mining module 306 and a push module 307. Among them:

[0181] The acquisition module 301 is used to acquire the text data to be processed input by the user;

[0182] A preprocessing module 302 is used to preprocess the text data to obtain corresponding specified text data;

[0183] A first construction module 303 is used to construct a corresponding word vector based on the specified text data;

[0184] A second construction module 304 is used to construct a corresponding target graph based on the word vector;

[0185] The optimization module 305 is used to optimize the preset original ant colony algorithm based on the preset attribute heuristic factor and the genetic algorithm to obtain the corresponding target ant colony algorithm;

[0186] A mining module 306 is used to perform subgraph mining processing on the target graph based on the target ant colony algorithm to obtain a corresponding target subgraph;

[0187] The push module 307 is used to push the information to the user based on the target subgraph.

[0188] In this embodiment, the operations performed by the above modules or units respectively correspond one-to-one to the steps of the artificial intelligence-based sub-graph generation method in the aforementioned implementation mode, and will not be repeated here.

[0189] In some optional implementations of this embodiment, the optimization module 305 includes:

[0190] A first generating submodule, used for generating the attribute inspiration factor based on a preset Bayesian formula;

[0191] An adding submodule, used for adding the attribute heuristic factor into the probability formula in the original ant colony algorithm to obtain a corresponding first ant colony algorithm;

[0192] A first acquisition submodule, used to acquire a crossover operator in the genetic algorithm;

[0193] An adjustment submodule, configured to adjust the volatility coefficient in the first ant colony algorithm based on the crossover operator to obtain an adjusted second ant colony algorithm;

[0194] A construction submodule is used to construct the target ant colony algorithm based on the second ant colony algorithm.

[0195] In this embodiment, the operations performed by the above modules or units respectively correspond one-to-one to the steps of the artificial intelligence-based sub-graph generation method in the aforementioned implementation mode, and will not be repeated here.

[0196] In some optional implementations of this embodiment, the mining module 306 includes:

[0197] A mining submodule, used for performing subgraph mining processing on the target graph based on the target ant colony algorithm to obtain all corresponding iteration records;

[0198] An extraction submodule, configured to extract, for each of the iteration records, a stored optimal subgraph and a quality assessment score corresponding to the optimal subgraph from the iteration record;

[0199] A screening submodule, used for numerically comparing the quality evaluation scores of all the optimal subgraphs, and screening out the global optimal subgraph with the highest quality evaluation score;

[0200] The first determination submodule is used to use the global optimal subgraph as the target subgraph.

[0201] In this embodiment, the operations performed by the above modules or units respectively correspond one-to-one to the steps of the artificial intelligence-based sub-graph generation method in the aforementioned implementation mode, and will not be repeated here.

[0202] In some optional implementations of this embodiment, the push module 307 includes:

[0203] A first processing submodule is used to perform visualization processing on the target sub-image to obtain a corresponding sub-image image;

[0204] A second acquisition submodule is used to acquire attribute information of a target node in the target subgraph, and acquire a node connection relationship corresponding to the target node;

[0205] A second generating submodule, configured to generate target description information corresponding to the target subgraph based on the attribute information and the node connection relationship;

[0206] The push submodule is used to push the sub-image and the target description information to the user.

[0207] In this embodiment, the operations performed by the above modules or units respectively correspond one-to-one to the steps of the artificial intelligence-based sub-graph generation method in the aforementioned implementation mode, and will not be repeated here.

[0208] In some optional implementations of this embodiment, the second building module 304 includes:

[0209] The third acquisition submodule is used to acquire the preset graph strategy and graph optimization strategy;

[0210] A second processing submodule, configured to process the word vector based on the graph strategy to obtain corresponding first graph data;

[0211] an optimization submodule, configured to optimize the first graph data based on the graph optimization strategy to obtain corresponding second graph data;

[0212] The second determining submodule is used to use the second graph data as the target graph.

[0213] In this embodiment, the operations performed by the above modules or units respectively correspond one-to-one to the steps of the artificial intelligence-based sub-graph generation method in the aforementioned implementation mode, and will not be repeated here.

[0214] In some optional implementations of this embodiment, the first building module 303 includes:

[0215] The calling submodule is used to call the preset natural language processing tool;

[0216] A conversion submodule, used for performing vector conversion processing on the specified text data based on the natural language processing tool to obtain corresponding conversion data;

[0217] The third determination submodule is used to use the conversion data as the word vector.

[0218] In this embodiment, the operations performed by the above modules or units respectively correspond one-to-one to the steps of the artificial intelligence-based sub-graph generation method in the aforementioned implementation mode, and will not be repeated here.

[0219] In some optional implementations of this embodiment, the preprocessing module 302 includes:

[0220] A denoising submodule, used for performing denoising processing on the text data to obtain corresponding first text data;

[0221] A word segmentation submodule, used to perform word segmentation processing on the first text data to obtain corresponding second text data;

[0222] A filtering submodule, configured to perform stop word filtering processing on the second text data to obtain corresponding third text data;

[0223] The fourth determining submodule is configured to use the third text data as the designated text data.

[0224] In this embodiment, the operations performed by the above modules or units respectively correspond one-to-one to the steps of the artificial intelligence-based sub-graph generation method in the aforementioned implementation mode, and will not be repeated here.

[0225] To solve the above technical problems, the present application also provides a computer device. Figure 4 , Figure 4 This is a basic structural block diagram of the computer device in this embodiment.

[0226] The computer device 4 includes a memory 41, a processor 42, and a network interface 43 that are interconnected through a system bus. It should be noted that the figure only shows a computer device 4 with components 41-43, but it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented instead. Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application specific integrated circuits (Application Specific Integrated Circuit, ASIC), programmable gate arrays (Field-Programmable Gate Array, FPGA), digital processors (Digital Signal Processor, DSP), embedded devices, etc.

[0227] The computer device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The computer device may interact with a user through a keyboard, a mouse, a remote controller, a touch pad, or a voice control device.

[0228] The memory 41 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc. In some embodiments, the memory 41 can be an internal storage unit of the computer device 4, such as a hard disk or memory of the computer device 4. In other embodiments, the memory 41 can also be an external storage device of the computer device 4, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (FlashCard), etc. equipped on the computer device 4. Of course, the memory 41 can also include both the internal storage unit of the computer device 4 and its external storage device. In this embodiment, the memory 41 is generally used to store the operating system and various application software installed on the computer device 4, such as computer-readable instructions of the sub-graph generation method based on artificial intelligence, etc. In addition, the memory 41 can also be used to temporarily store various types of data that have been output or are to be output.

[0229] The processor 42 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips in some embodiments. The processor 42 is generally used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is used to run computer-readable instructions stored in the memory 41 or process data, such as computer-readable instructions for running the subgraph generation method based on artificial intelligence.

[0230] The network interface 43 may include a wireless network interface or a wired network interface. The network interface 43 is generally used to establish a communication connection between the computer device 4 and other electronic devices.

[0231] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0232] The present application first obtains the text data to be processed input by the user; then preprocesses the text data to obtain the corresponding specified text data; then constructs the corresponding word vector based on the specified text data; and constructs the corresponding target graph based on the word vector; subsequently optimizes the preset original ant colony algorithm based on the preset attribute heuristic factor and genetic algorithm to obtain the corresponding target ant colony algorithm; further performs subgraph mining on the target graph based on the target ant colony algorithm to obtain the corresponding target subgraph; and finally pushes the user based on the target subgraph. The present application obtains the specified text data by preprocessing the text data input by the user, then constructs the corresponding target graph based on the specified text data, then optimizes the original ant colony algorithm based on the attribute heuristic factor and genetic algorithm to obtain the target ant colony algorithm, and then performs subgraph mining on the target graph based on the target ant colony algorithm, which can quickly and accurately complete the subgraph mining processing of the text data, effectively improves the processing efficiency and processing accuracy of the subgraph mining, and effectively guarantees the accuracy of the generated target subgraph.

[0233] The present application also provides another implementation, namely, providing a computer-readable storage medium, wherein the computer-readable storage medium stores computer-readable instructions, and the computer-readable instructions can be executed by at least one processor to enable the at least one processor to perform the steps of the artificial intelligence-based sub-graph generation method as described above.

[0234] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0235] The present application first obtains the text data to be processed input by the user; then preprocesses the text data to obtain the corresponding specified text data; then constructs the corresponding word vector based on the specified text data; and constructs the corresponding target graph based on the word vector; subsequently optimizes the preset original ant colony algorithm based on the preset attribute heuristic factor and genetic algorithm to obtain the corresponding target ant colony algorithm; further performs subgraph mining on the target graph based on the target ant colony algorithm to obtain the corresponding target subgraph; and finally pushes the user based on the target subgraph. The present application obtains the specified text data by preprocessing the text data input by the user, then constructs the corresponding target graph based on the specified text data, then optimizes the original ant colony algorithm based on the attribute heuristic factor and genetic algorithm to obtain the target ant colony algorithm, and then performs subgraph mining on the target graph based on the target ant colony algorithm, which can quickly and accurately complete the subgraph mining processing of the text data, effectively improves the processing efficiency and processing accuracy of the subgraph mining, and effectively guarantees the accuracy of the generated target subgraph.

[0236] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0237] Obviously, the embodiments described above are only some embodiments of the present application, rather than all embodiments. The preferred embodiments of the present application are given in the accompanying drawings, but they do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present application more thorough and comprehensive. Although the present application is described in detail with reference to the aforementioned embodiments, for those skilled in the art, it is still possible to modify the technical solutions recorded in the aforementioned specific implementation methods, or to perform equivalent replacement of some of the technical features therein. Any equivalent structure made using the contents of the specification and drawings of this application, directly or indirectly used in other related technical fields, is similarly within the scope of patent protection of this application.

Claims

1. A subgraph generation method based on artificial intelligence, characterized in that: The steps include: Get the text data to be processed input by the user; Preprocessing the text data to obtain corresponding specified text data; Construct a corresponding word vector based on the specified text data; Construct a corresponding target graph based on the word vector; Based on the preset attribute heuristic factors and genetic algorithm, the preset original ant colony algorithm is optimized to obtain the corresponding target ant colony algorithm; Performing subgraph mining processing on the target graph based on the target ant colony algorithm to obtain a corresponding target subgraph; Push processing is performed on the user based on the target subgraph.

2. The subgraph generation method based on artificial intelligence according to claim 1, characterized in that: The step of optimizing the preset original ant colony algorithm based on the preset attribute heuristic factor and the genetic algorithm to obtain the corresponding target ant colony algorithm specifically includes: Generating the attribute inspiration factor based on a preset Bayesian formula; Adding the attribute heuristic factor to the probability formula in the original ant colony algorithm to obtain a corresponding first ant colony algorithm; Obtaining a crossover operator in the genetic algorithm; Adjusting the volatility coefficient in the first ant colony algorithm based on the crossover operator to obtain an adjusted second ant colony algorithm; The target ant colony algorithm is constructed based on the second ant colony algorithm.

3. The subgraph generation method based on artificial intelligence according to claim 1, characterized in that: The step of performing subgraph mining processing on the target graph based on the target ant colony algorithm to obtain the corresponding target subgraph specifically includes: Performing subgraph mining processing on the target graph based on the target ant colony algorithm to obtain all corresponding iteration records; For each of the iteration records, extracting a saved optimal subgraph and a quality assessment score corresponding to the optimal subgraph from the iteration record; Numerically comparing the quality evaluation scores of all the optimal subgraphs, and selecting the global optimal subgraph with the highest quality evaluation score; The global optimal subgraph is used as the target subgraph.

4. The subgraph generation method based on artificial intelligence according to claim 1, characterized in that: The step of performing push processing on the user based on the target subgraph specifically includes: Performing visualization processing on the target sub-graph to obtain a corresponding sub-graph image; Acquire attribute information of a target node in the target subgraph, and acquire a node connection relationship corresponding to the target node; Generate target description information corresponding to the target subgraph based on the attribute information and the node connection relationship; The sub-image and the target description information are pushed to the user.

5. The subgraph generation method based on artificial intelligence according to claim 1, characterized in that: The step of constructing a corresponding target graph based on the word vector specifically includes: Obtain preset graph strategies and graph optimization strategies; Processing the word vector based on the graph strategy to obtain corresponding first graph data; Optimizing the first graph data based on the graph optimization strategy to obtain corresponding second graph data; The second graph data is used as the target graph.

6. The subgraph generation method based on artificial intelligence according to claim 1, characterized in that: The step of constructing a corresponding word vector based on the specified text data specifically includes: Calling preset natural language processing tools; Performing vector conversion processing on the specified text data based on the natural language processing tool to obtain corresponding conversion data; The converted data is used as the word vector.

7. The subgraph generation method based on artificial intelligence according to claim 1, characterized in that: The step of preprocessing the text data to obtain corresponding specified text data specifically includes: Performing denoising processing on the text data to obtain corresponding first text data; Performing word segmentation processing on the first text data to obtain corresponding second text data; Performing stop word filtering processing on the second text data to obtain corresponding third text data; The third text data is used as the designated text data.

8. A sub-graph generation device based on artificial intelligence, characterized in that: include: An acquisition module is used to acquire text data to be processed input by a user; A preprocessing module, used for preprocessing the text data to obtain corresponding specified text data; A first construction module, used to construct a corresponding word vector based on the specified text data; A second construction module is used to construct a corresponding target graph based on the word vector; The optimization module is used to optimize the preset original ant colony algorithm based on the preset attribute heuristic factor and the genetic algorithm to obtain the corresponding target ant colony algorithm; A mining module, used for performing subgraph mining processing on the target graph based on the target ant colony algorithm to obtain a corresponding target subgraph; A push module is used to push the user based on the target subgraph.

9. A computer device, characterized in that: It comprises a memory and a processor, wherein the memory stores computer-readable instructions, and when the processor executes the computer-readable instructions, the steps of the sub-graph generation method based on artificial intelligence as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by a processor, the steps of the sub-graph generation method based on artificial intelligence as described in any one of claims 1 to 7 are implemented.

Images