Criminal investigation knowledge graph construction method, system and equipment based on multi-agent collaboration

Abstract

The invention relates to the technical field of knowledge maps, in particular to a criminal investigation knowledge map construction method, system and device based on multi-agent collaboration, and aims to improve the quality of a knowledge map. The criminal investigation knowledge graph construction system based on multi-agent collaboration comprises a planning agent, an analysis agent and a verification agent. Wherein the planning agent is used for generating a first retrieval parameter and a planning prompt word according to a user demand, and obtaining first task data from the data platform according to the first retrieval parameter; the analysis agent is used for generating a structured expression of the target knowledge graph according to the planning prompt word and initializing the target knowledge graph according to the first task data; the verification agent is used for checking whether elements in the target knowledge graph have conflicts or not, and if yes, the element with the highest authenticity probability is selected. According to the method, the efficiency of constructing the knowledge graph is improved, and the quality of the knowledge graph is improved.

Description

Technical Field

[0001] The present invention relates to the field of knowledge graph technology, and specifically to a method, system and device for constructing a criminal investigation knowledge graph based on multi-agent collaboration. Background Art

[0002] As a key intelligent technology, building a knowledge graph for criminal investigation is gradually becoming the core force to improve investigation efficiency. The knowledge graph is a structured semantic knowledge base that stores and displays complex relationships between entities in a graphical way, providing a comprehensive and dynamically updated intelligence analysis platform for the field of criminal investigation.

[0003] In terms of the technology of automatically constructing criminal investigation knowledge graphs, there are currently three main models: the Universal Information Extraction (UIE) model, the Large Language Model (LLM) model, and the Single Agent model. However, the quality of the knowledge graphs constructed by these existing technologies when applied to the field of criminal investigation is still unsatisfactory. Summary of the invention

[0004] In order to solve the above problems in the prior art, the present invention proposes a method, system and device for constructing a criminal investigation knowledge graph based on multi-agent collaboration, which improves the quality of the constructed knowledge graph.

[0005] In a first aspect of the present invention, a criminal investigation knowledge graph construction system based on multi-agent collaboration is proposed, the system comprising:

[0006] A planning agent, used to generate a first search parameter and a planning prompt word according to user requirements, and obtain first task data from a data platform according to the first search parameter;

[0007] An analysis agent, used to generate a structured expression of a target knowledge graph according to the planning prompt words, and initialize the target knowledge graph according to the first task data;

[0008] The verification agent is used to check whether there is a conflict between the elements in the target knowledge graph. If so, the element with the highest probability of authenticity is selected.

[0009] Preferably, the system further comprises:

[0010] The line expansion agent is used to retrieve whether there is an expandable entity on the leaf node in the target knowledge graph, and if so, generate a line expansion demand and feed it back to the planning agent;

[0011] The planning agent is further used to generate a second search parameter and a line expansion prompt word according to the line expansion demand, and obtain second task data from the data platform according to the second search parameter;

[0012] The analysis agent is further used to expand the target knowledge graph according to the line expansion prompt word and the second task data;

[0013] The verification agent is also used to check whether there are conflicts between the elements in the expanded target knowledge graph. If so, the element with the highest probability of authenticity is selected.

[0014] Preferably, the system further comprises:

[0015] The first training module is used to train the planning agent, the analysis agent, the verification agent and the line expansion agent.

[0016] Preferably, the first training module includes:

[0017] A pre-training unit, used to perform pre-training on the planning agent, the analysis agent, the verification agent and the line expansion agent respectively;

[0018] The collaborative training unit is used to use the MADDPG algorithm to collaboratively train the planning agent, the analysis agent, the verification agent and the line expansion agent after preheating training until the quality score r1 of the generated knowledge graph reaches a maximum value.

[0019] Preferably, the pre-training unit comprises:

[0020] The first subunit is configured to: generate a knowledge graph based on the planning agent and the analysis agent while the analysis agent is fixed, and train the planning agent using the RLHF process and the PPO algorithm until the quality score r2 of the generated knowledge graph reaches a maximum value;

[0021] The second subunit is configured to: generate a knowledge graph based on the planning agent and the analysis agent while the planning agent is fixed, and train the analysis agent using the RLHF process and the PPO algorithm until the quality score r3 of the generated knowledge graph reaches a maximum value;

[0022] The third subunit is configured to: train the verification agent based on the graph training set with conflicting elements until the recognition rate of the conflicting elements reaches a maximum value;

[0023] The fourth subunit is configured to: train the line expansion agent based on a graph training set in which expandable entities exist, until the recognition rate of the expandable entities reaches a maximum value.

[0024] Preferably, the system further comprises:

[0025] The reward model is used to calculate the quality score r1, the quality score r2 or the quality score r3 of the knowledge graph according to the following formula during the training process:

[0026] r G =w P P G +w R R G +w r r s

[0027] in,

[0028] r G represents the quality score r1, the quality score r2 or the quality score r3;

[0029] P G represents the precision rate; R G represents the recall rate; r s represents subjective rating; w P 、w R and w r Both represent weights, which are calculated by the multivariate gradient regression model.

[0030] Preferably, the system further comprises:

[0031] The subjective scoring model is used to calculate the subjective score;

[0032] A second training module, used for training the subjective scoring model;

[0033] The second training module includes:

[0034] An initialization unit, used to set random parameters for the planning agent, the analysis agent, the verification agent and the line expansion agent respectively;

[0035] A graph construction unit, used to construct K graph samples using the planning agent, the analysis agent, the verification agent and the line expansion agent; K is a preset value;

[0036] A sample generation unit is used to sort the K graph samples according to the scores of the case handlers, and generate a training samples;

[0037] A training unit, used for training the subjective scoring model based on the training samples and the ranking loss function;

[0038] in,

[0039] Each of the training samples is a triple (x, y1, y2), where x represents the input data for generating the knowledge graph, and y1 and y2 represent graphs that receive higher and lower scores from case handlers, respectively;

[0040] The sorting loss function is:

[0041]

[0042] D stands for A data set consisting of the training samples;

[0043] represents the expectation of the training samples from the data set D;

[0044] r θ (x,y1) represents the prediction score of the model when input (x,y1);

[0045] r θ (x,y2) represents the prediction score of the model when input (x,y2);

[0046] σ represents the sigmoid function.

[0047] Preferably, the system further comprises:

[0048] An authenticity assessment model, used to assess the authenticity probability of the graph elements;

[0049] A third training module is used to train the authenticity assessment model;

[0050] The third training module includes:

[0051] A sample generation unit, which is used to collect authoritative open source knowledge graph data as positive samples, and randomly replace a certain attribute of the positive samples with a false description to generate negative samples;

[0052] A training set construction unit, used to label the authenticity probabilities of the positive samples and the negative samples, and then construct a training set;

[0053] The training unit is used to select a graph element z from the training set, and record all element vectors connected within N hops of the element as Represent the authority of the data source of this element as a weight vector and will As the input of the authenticity assessment model, back propagation method is adopted for training.

[0054] The second aspect of the present invention provides a method for constructing a criminal investigation knowledge graph based on multi-agent collaboration, the method comprising:

[0055] The planning agent generates a first search parameter and a planning prompt word according to the user's needs, and obtains the first task data from the data platform according to the first search parameter;

[0056] The analysis agent generates a structured expression of a target knowledge graph according to the planning prompt words, and initializes the target knowledge graph according to the first task data;

[0057] The verification agent checks whether there are conflicts between the elements in the target knowledge graph, and if so, selects the element with the highest probability of authenticity.

[0058] According to a third aspect of the present invention, a computer-readable storage device is provided, storing a computer program that can be loaded by a processor and execute the method described above.

[0059] The present invention has the following beneficial effects:

[0060] The criminal investigation knowledge graph construction system based on multi-agent collaboration proposed in the present invention uses planning agents, analysis agents and verification agents to undertake planning, construction and verification functions respectively, and can automatically verify the authenticity when data conflicts. This not only improves the efficiency of knowledge graph construction, but also improves the quality of the target knowledge graph.

[0061] Using the expansion agent to detect expandable entities and feed back to the planning agent provides an effective way to dynamically expand the target knowledge graph, making the target knowledge graph information more and more complete and providing a good decision-making basis for criminal investigation personnel.

[0062] In the first training module of the present invention, when training multiple agents, each agent is first preheated and then trained collaboratively, which reduces the training cost and further improves the quality of the knowledge graph. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 It is a schematic diagram of the main components of the first embodiment of the criminal investigation knowledge graph construction system based on multi-agent collaboration of the present invention;

[0064] Figure 2 is a schematic diagram of a target knowledge graph generated in an embodiment of the present invention;

[0065] Figure 3 It is a schematic diagram of the main components of Example 2 of the criminal investigation knowledge graph construction system based on multi-agent collaboration of the present invention. DETAILED DESCRIPTION

[0066] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0067] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0068] It should be noted that, in the description of the present invention, the terms "first" and "second" are only for the convenience of description, and do not indicate or imply the relative importance of the devices, elements or parameters, and therefore cannot be understood as limiting the present invention. In addition, the term "and / or" in the present invention is only a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article, unless otherwise specified, generally indicates that the associated objects before and after are in an "or" relationship.

[0069] Building a criminal investigation knowledge graph mainly relies on several key elements:

[0070] (1) Entity Recognition: In the construction of knowledge graphs, entities refer to independent objects that can be clearly identified and distinguished, including people, cases, objects, scenes, files, etc. Each entity has its own unique attributes, such as the name of a person, the number of a case, the description of an object, etc. These attributes provide accurate identification for the entity.

[0071] (2) Relationship modeling: The relationships between entities constitute the skeleton of the knowledge graph, which reveals the mutual influence and connection between entities. For example, the relationship between people and cases may include roles such as suspects, victims, and witnesses; the relationship between cases and scenes may involve key information such as the location and time of the crime.

[0072] (3) Attribute extraction: Each entity is equipped with a series of attributes that provide a detailed description of the entity. Taking the case entity as an example, it may contain key details such as the specific time and place of the case, the suspects involved, and the victim information.

[0073] (4) Data fusion: Knowledge graphs integrate data from different sources, such as transcript data, police information data, on-site investigations, social media, police records, etc., in order to reduce information redundancy and improve data consistency and accuracy.

[0074] (5) Dynamic update: Given that criminal intelligence investigation is an ever-changing process, the knowledge graph must be able to update in real time to reflect the latest intelligence and data. This includes the entry of new cases, the discovery of new evidence, the identification of new suspects, etc., to ensure that the intelligence analysis platform is always up to date.

[0075] (6) Quality assessment: A strict quality assessment system is the guarantee of the credibility of the knowledge graph. It requires multi-faceted assessment to ensure the accuracy and authenticity of graph elements and their actual contribution to case investigation.

[0076] Multi-Agent Deep Deterministic Policy Gradient (MADDPG) is a reinforcement learning algorithm used in multi-agent environments. It is an extension of the Deep Deterministic Policy Gradient (DDPG) algorithm in multi-agent scenarios. MADDPG is mainly used to solve cooperation and competition problems in multi-agent environments, especially when the interactions between agents may be very complex.

[0077] The MADDPG algorithm is based on the Actor-Critic structure, where each agent has its own Actor and Critic network. The Actor is used to output action strategies, while the Critic is used to evaluate the value of actions. The uniqueness of this algorithm is that the Critic network is global, that is, it not only depends on the state and action of a single agent, but also uses the state and action of all agents.

[0078] In an embodiment of the present invention, a planning agent, an analysis agent, a verification agent and a line expansion agent are used to construct and update a criminal investigation knowledge graph. Each agent is a large language model-based agent (LLM-BasedAgent), and each agent includes an Actor network and a Critic network.

[0079] Figure 1 Schematic diagram of the main components of the first embodiment of the criminal investigation knowledge graph construction system based on multi-agent collaboration of the present invention. Figure 1 As shown, the system 100 of this embodiment includes: a planning agent 101, an analysis agent 102 and a verification agent 103.

[0080] The planning agent 101 is used to generate a first search parameter and a planning prompt word according to user needs, and obtain the first task data from the data platform according to the first search parameter.

[0081] Among them, the planning prompt words include the steps and precautions for building the knowledge graph, etc., which are used to guide the analysis agent 102 to build the knowledge graph; the data platform stores various types of data that may be used in handling cases, such as the smart criminal investigation platform; the first task data is the data required to create the target knowledge graph according to user needs, and the first task data may include: case data, personnel data, transcript data, and police situation data. In some embodiments, the planning agent 101 will also search for an interface for obtaining the first task data according to the interface document of the data platform 300, and then obtain the first task data required for building the knowledge graph from the data platform according to the interface and the first search parameter.

[0082] The analysis agent 102 is used to generate a structured expression of the target knowledge graph 400 according to the planning prompt words, and initialize the target knowledge graph 400 according to the first task data.

[0083] Among them, the step of initializing the target knowledge graph according to the first task data is specifically: extracting entities, relationships and attributes from the first task data, and then filling them into the target knowledge graph.

[0084] The verification agent 103 is used to check whether there is any conflict between the elements in the target knowledge graph 400. If so, the element with the highest probability of authenticity is selected.

[0085] Specifically, the verification agent 103 checks whether the entity has attributes with the same or similar meanings, checks whether these attributes cause conflicts, sorts out all conflict points and the data sources of the conflict points, calculates the authenticity probability of the conflict points, and updates the map.

[0086] For example, the user's requirement is: to build an organizational relationship knowledge graph of Company B based on the transcript information of the arrested suspects in the fraud case in City A.

[0087] The first search parameters generated by the planning agent 101 include: information retrieval parameters of organizations involved in the case (company ID), information retrieval parameters of persons involved in the case (company ID and case ID), case information retrieval parameters (case ID), information retrieval parameters of persons involved in the case, information retrieval parameters of transcripts of persons involved in the case, etc.

[0088] The prompt planning words generated by the planning agent 101 include: (1) Entity recognition: Identify entities such as companies, individuals, cases, etc. from the first task data and assign a unique identifier to each entity. (2) Relationship extraction: Determine the relationship between entities, such as the relationship between employees and companies, the relationship between persons involved in a case and cases, etc. (3) Attribute assignment: Assign corresponding attributes to each entity, such as company name, employee position, case type, etc. (4) Graph construction: Use a graph database or other suitable tools to construct a knowledge graph based on entities and relationships. (5) Graph optimization: Optimize the graph, including merging duplicate entities, adjusting relationship directions and types, etc., to improve the readability and usability of the graph. (6) Graph verification: Verify the accuracy and completeness of the graph to ensure that all key information is correctly represented.

[0089] The target knowledge graph constructed and initialized by the analysis agent 102 according to the planning prompt words is as follows Figure 2 As shown. The figure shows multiple entities and the relationships between them by way of example, but does not show the attributes. In this example, the verification agent 103 found that there was a conflict in the attribute information of the victims, that is, the number of victims stated in Zhang San's transcript was 60, which was inconsistent with the number of 130 victims recorded in the case investigation data. After calculating the authenticity probability of these two attribute values ​​respectively, it is believed that the value of 130 is more authentic, so the number of victims in the target knowledge graph is set to 130.

[0090] Figure 3 Schematic diagram of the main components of the second embodiment of the criminal investigation knowledge graph construction system based on multi-agent collaboration of the present invention. Figure 3 As shown, the system of this embodiment includes: a planning agent 201, an analysis agent 202, a verification agent 203 and a line expansion agent 204.

[0091] The planning agent 201 is used to generate a first search parameter and a planning prompt word according to user requirements, and obtain the first task data from the data platform according to the first search parameter, and is also used to generate a second search parameter and a line expansion prompt word according to the line expansion requirements sent by the line expansion agent 204, and obtain the second task data from the data platform 300 according to the second search parameter. Among them, the line expansion prompt word includes the steps and precautions for expanding the knowledge graph, etc., which are used to guide the analysis agent 202 to expand the target knowledge graph.

[0092] The analysis agent 202 is used to generate a structured expression of the target knowledge graph 400 according to the planning prompt words, initialize the target knowledge graph 400 according to the first task data, and expand the target knowledge graph 400 according to the expansion prompt words and the second task data.

[0093] The verification agent 203 is used to check whether there is a conflict between the elements in the target knowledge graph. If so, the element with the highest probability of authenticity is selected.

[0094] The line expansion agent 204 is used to retrieve whether there are expandable entities on the leaf nodes in the target knowledge graph. If so, it generates a line expansion demand and feeds it back to the planning agent 201.

[0095] For example, the line expansion agent 204 detects Figure 2 The leaf node in the target knowledge graph shown in the figure has an attribute of the ID card information of "Zhang San". Then, the family members, criminal records and other information related to the ID card information can be obtained through the data platform. At this time, the entity "Zhang San" is considered to be expandable, and the expansion demand is generated and fed back to the planning agent 201. The planning agent 201 generates the second search parameter and the expansion prompt word according to the expansion demand, and obtains the second task data (Zhang San's family members, criminal records, etc.) according to the second search parameter. The analysis agent 202 expands the target knowledge graph according to the expansion prompt word, thereby providing more detailed information for the case handlers.

[0096] In an optional embodiment, the knowledge graph construction system of the present invention may also include: a first training module, a reward model, a subjective scoring model, a second training module, an authenticity evaluation model and a third training module.

[0097] In this embodiment, the first training module is used to train the planning agent, the analysis agent, the verification agent and the line expansion agent. The first training module may include: a pre-training unit and a collaborative training unit.

[0098] The pre-training unit is used to perform pre-training on the planning agent, the analysis agent, the verification agent and the line expansion agent respectively. The pre-training unit may include: a first sub-unit, a second sub-unit, a third sub-unit and a fourth sub-unit.

[0099] Among them, the first subunit is configured as follows: in the case of a fixed analysis agent, a knowledge graph is generated based on the planning agent and the analysis agent, and the planning agent is trained using the RLHF process and the PPO algorithm until the quality score r2 of the generated knowledge graph reaches a maximum value; the second subunit is configured as follows: in the case of a fixed planning agent, a knowledge graph is generated based on the planning agent and the analysis agent, and the analysis agent is trained using the RLHF process and the PPO algorithm until the quality score r3 of the generated knowledge graph reaches a maximum value; the third subunit is configured as follows: based on the graph training set with conflicting elements, the verification agent is trained until the recognition rate of the conflicting elements reaches a maximum value; the fourth subunit is configured as follows: based on the graph training set with extensible entities, the line expansion agent is trained until the recognition rate of the extensible entities reaches a maximum value.

[0100] The collaborative training unit is used to use the MADDPG algorithm to collaboratively train the planning agent, analysis agent, verification agent and line expansion agent after preheating training until the quality score r1 of the generated knowledge graph reaches the maximum value.

[0101] In this embodiment, the reward model is used to calculate the quality score r1, quality score r2 or quality score r3 of the knowledge graph during the training process according to the method shown in the following formula (1):

[0102] r G = w P P G +w R R G +w r r s (1)

[0103] Among them, r G represents the quality score r1, quality score r2 or quality score r3; P G represents the precision rate; R G represents the recall rate; r s represents subjective rating; w P 、w R and w r Both represent weights, which are calculated by the multivariate gradient regression model.

[0104] Precision is the ratio of the intersection of the knowledge graph and the baseline (i.e., the reference standard or known correct knowledge graph) to the knowledge graph. G The higher it is, the more accurate the information in the constructed knowledge graph is (consistent with the baseline). Recall is the ratio of the intersection of the knowledge graph and the baseline to the baseline. G The higher it is, the more the constructed knowledge graph successfully covers more information in the baseline.

[0105] In this embodiment, the subjective scoring model adopts the ResNet network structure to calculate the subjective score r in formula (1) s .

[0106] In this embodiment, the second training module is used to train the subjective scoring model, and the second training module includes: an initialization unit, a graph construction unit, a sample generation unit and a training unit.

[0107] The initialization unit is used to set random parameters for the planning agent, analysis agent, verification agent and line expansion agent respectively; the graph construction unit is used to use the planning agent, analysis agent, verification agent and line expansion agent to construct K graph samples, K is a preset value; the sample generation unit is used to sort the K graph samples according to the scores of the case handlers, and generate training samples; the training unit is used to train the subjective rating model based on the training samples and the ranking loss function.

[0108] Each training sample is a triple (x, y1, y2), where x represents the input data for generating the knowledge graph, and y and y2 represent the graphs that receive higher and lower scores from case handlers, respectively. The ranking loss function is shown in the following formula (2):

[0109]

[0110] Where D represents A dataset consisting of training samples; represents the expectation of the training samples from the dataset D; r θ (x,y1) represents the predicted score of the model when input (x,y1); r θ (x,y2) represents the prediction score of the model when input (x,y2); σ represents the sigmoid function.

[0111] In this embodiment, the authenticity assessment model adopts a ResNet network structure to assess the authenticity probability of graph elements.

[0112] In this embodiment, the third training module is used to train the authenticity assessment model; the third training module includes: a sample generation unit, a training set construction unit and a training unit.

[0113] The sample generation unit is used to collect authoritative open source knowledge graph data as positive samples, and randomly replace a certain attribute of the positive sample with a false description to generate a negative sample; the training set construction unit is used to annotate the authenticity probability of the positive and negative samples, and then construct the training set; the training unit is used to select the graph element z from the training set, and record all the element vectors connected within N hops of the element as Represent the authority of the data source of this element as a weight vector and will As the input of the authenticity assessment model, it is trained using the back-propagation method.

[0114] Based on the above system embodiment, the present invention also provides an embodiment of a method for constructing a criminal investigation knowledge graph based on multi-agent collaboration. The method of this embodiment mainly includes steps S10-S30:

[0115] Step S10: The planning agent generates a first search parameter and a planning prompt word according to user needs, and obtains first task data from the data platform according to the first search parameter.

[0116] Step S20: The analysis agent generates a structured expression of the target knowledge graph according to the planning prompt words, and initializes the target knowledge graph according to the first task data.

[0117] Step S30: The verification agent checks whether there is any conflict between the elements in the target knowledge graph. If so, the element with the highest probability of authenticity is selected.

[0118] Furthermore, the present invention also provides an embodiment of a computer-readable storage device, in which the storage device of this embodiment stores a computer program that can be loaded by a processor and execute the method described above.

[0119] The computer-readable storage device may include: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0120] Those skilled in the art should be able to appreciate that the method steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of electronic hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in electronic hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0121] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. A criminal investigation knowledge graph construction system based on multi-agent collaboration, characterized in that: The system comprises: A planning agent, used to generate a first search parameter and a planning prompt word according to user requirements, and obtain first task data from a data platform according to the first search parameter; An analysis agent, used to generate a structured expression of a target knowledge graph according to the planning prompt words, and initialize the target knowledge graph according to the first task data; The verification agent is used to check whether there is a conflict between the elements in the target knowledge graph. If so, the element with the highest probability of authenticity is selected.

2. The criminal investigation knowledge graph construction system based on multi-agent collaboration according to claim 1 is characterized in that: The system further comprises: The line expansion agent is used to retrieve whether there is an expandable entity on the leaf node in the target knowledge graph, and if so, generate a line expansion demand and feed it back to the planning agent; The planning agent is further used to generate a second search parameter and a line expansion prompt word according to the line expansion demand, and obtain second task data from the data platform according to the second search parameter; The analysis agent is further used to expand the target knowledge graph according to the line expansion prompt word and the second task data; The verification agent is also used to check whether there are conflicts between the elements in the expanded target knowledge graph. If so, the element with the highest probability of authenticity is selected.

3. The criminal investigation knowledge graph construction system based on multi-agent collaboration according to claim 2 is characterized in that: The system further comprises: The first training module is used to train the planning agent, the analysis agent, the verification agent and the line expansion agent.

4. The criminal investigation knowledge graph construction system based on multi-agent collaboration according to claim 3 is characterized in that: The first training module includes: A pre-training unit, used to perform pre-training on the planning agent, the analysis agent, the verification agent and the line expansion agent respectively; The collaborative training unit is used to use the MADDPG algorithm to collaboratively train the planning agent, the analysis agent, the verification agent and the line expansion agent after preheating training until the quality score r1 of the generated knowledge graph reaches a maximum value.

5. The criminal investigation knowledge graph construction system based on multi-agent collaboration according to claim 4 is characterized in that: The pre-training unit comprises: The first subunit is configured to: generate a knowledge graph based on the planning agent and the analysis agent while the analysis agent is fixed, and train the planning agent using the RLHF process and the PPO algorithm until the quality score r2 of the generated knowledge graph reaches a maximum value; The second subunit is configured to: generate a knowledge graph based on the planning agent and the analysis agent while fixing the planning agent, and train the analysis agent using the RLHF process and the PPO algorithm until the quality score r3 of the generated knowledge graph reaches a maximum value; The third subunit is configured to: train the verification agent based on the graph training set with conflicting elements until the recognition rate of the conflicting elements reaches a maximum value; The fourth subunit is configured to: train the line expansion agent based on a graph training set in which expandable entities exist, until the recognition rate of the expandable entities reaches a maximum value.

6. The criminal investigation knowledge graph construction system based on multi-agent collaboration according to claim 5 is characterized in that: The system further comprises: The reward model is used to calculate the quality score r1, the quality score r2 or the quality score r3 of the knowledge graph according to the following formula during the training process: r G =w P P G +w R R G +w r r s in, r G represents the quality score r1, the quality score r2 or the quality score r3; P G represents the precision rate; R G represents the recall rate; r s represents subjective rating; w P 、w R and w r Both represent weights, which are calculated by the multivariate gradient regression model.

7. The criminal investigation knowledge graph construction system based on multi-agent collaboration according to claim 6 is characterized in that: The system further comprises: The subjective scoring model is used to calculate the subjective score; A second training module, used for training the subjective scoring model; The second training module includes: An initialization unit, used to set random parameters for the planning agent, the analysis agent, the verification agent and the line expansion agent respectively; A graph construction unit, used to construct K graph samples using the planning agent, the analysis agent, the verification agent and the line expansion agent; K is a preset value; A sample generation unit is used to sort the K graph samples according to the scores of the case handlers, and generate a training samples; A training unit, used for training the subjective scoring model based on the training samples and the ranking loss function; in, Each of the training samples is a triple (x, y1, y2), where x represents the input data for generating the knowledge graph, and y1 and y2 represent graphs that receive higher and lower scores from case handlers, respectively; The ranking loss function is: D stands for A data set consisting of the training samples; represents the expectation of the training samples from the data set D; r θ (x,y1) represents the prediction score of the model when input (x,y1); r θ (x,y2) represents the prediction score of the model when input (x,y2); σ represents the sigmoid function.

8. The criminal investigation knowledge graph construction system based on multi-agent collaboration according to claim 6 is characterized in that: The system further comprises: An authenticity assessment model, used to assess the authenticity probability of the graph elements; A third training module is used to train the authenticity assessment model; The third training module includes: A sample generation unit, which is used to collect authoritative open source knowledge graph data as positive samples, and randomly replace a certain attribute of the positive samples with a false description to generate negative samples; A training set construction unit, used to label the authenticity probabilities of the positive samples and the negative samples, and then construct a training set; The training unit is used to select a graph element z from the training set, and record all element vectors connected within N hops of the element as Represent the authority of the data source of this element as a weight vector and will As the input of the authenticity assessment model, back propagation method is adopted for training.

9. A method for constructing a criminal investigation knowledge graph based on multi-agent collaboration, characterized in that: The method comprises: The planning agent generates a first search parameter and a planning prompt word according to the user's needs, and obtains the first task data from the data platform according to the first search parameter; The analysis agent generates a structured expression of a target knowledge graph according to the planning prompt words, and initializes the target knowledge graph according to the first task data; The verification agent checks whether there are conflicts between the elements in the target knowledge graph, and if so, selects the element with the highest probability of authenticity.

10. A computer readable storage device, characterized in that: A computer program is stored which can be loaded by a processor and execute the method as claimed in claim 9.

Images