Project processing method and device based on strategy tree

Abstract

The embodiments of the present specification provide a project processing method and device based on a strategy tree, wherein a project processing method based on a strategy tree includes: in the process of obtaining a target claim prompt word for project claim processing, obtaining a claim prompt word for a security project as the first node for strategy tree calculation, inputting a claim event and a claim prompt word into a pre-trained model, so as to perform claim processing on the claim event according to the claim rules in the claim prompt word to obtain a claim result, updating the claim prompt word according to the claim result to obtain an updated claim prompt word, and using the updated claim prompt word as the second node for strategy tree calculation, and after the strategy tree calculation is completed, determining the target claim prompt word according to the path parameter values ​​of each strategy tree path in the obtained strategy tree to perform project claim processing.

Description

Technical Field

[0001] This document relates to the field of data processing technology, and in particular to a project processing method and device based on a strategy tree. Background Art

[0002] With the rapid development of big language model technology, many tasks can be completed automatically through big language models. In the field of claims settlement of insurance projects, using big language models to process claims incidents is an important application direction. However, with the expansion of insurance scenarios, insurance agencies have launched diversified products to meet the needs of consumers. Each insurance project has its own unique claims rules. These differences have brought complexity to the claims settlement work of insurance projects. In this case, how to use big language models to generate more accurate claims results has become a key focus in big model application scenarios. Summary of the invention

[0003] One or more embodiments of the present specification provide a project processing method based on a strategy tree, including: obtaining a claim prompt word for a security project as the first node for performing a strategy tree calculation. Inputting a claim event and the claim prompt word into a pre-trained model to perform a claim processing on the claim event according to the claim rules in the claim prompt word to obtain a claim result. According to the claim result, the claim prompt word is updated to obtain an updated claim prompt word, and the updated claim prompt word is used as the second node for performing a strategy tree calculation. After the strategy tree calculation is completed, the target claim prompt word is determined based on the path parameter values ​​of each strategy tree path in the obtained strategy tree to perform project claim processing.

[0004] One or more embodiments of the present specification provide a project processing device based on a strategy tree, including: a first node determination module, configured to obtain a claim prompt word of a security project as the first node for performing strategy tree calculation. A claim processing module, configured to input a claim event and the claim prompt word into a pre-trained model, so as to perform claim processing on the claim event according to the claim rules in the claim prompt word to obtain a claim result. A second node determination module, configured to update the claim prompt word according to the claim result to obtain an updated claim prompt word, and use the updated claim prompt word as the second node for performing strategy tree calculation. After the strategy tree calculation is completed, the prompt word determination module is run, and the prompt word determination module is configured to determine the target claim prompt word according to the path parameter values ​​of each strategy tree path in the obtained strategy tree, so as to perform project claim processing.

[0005] One or more embodiments of the present specification provide a project processing device based on a strategy tree, including: a processor; and a memory configured to store computer executable instructions, wherein the computer executable instructions, when executed, cause the processor to: obtain a claim prompt word for a security project as the first node for performing a strategy tree calculation. Input a claim event and the claim prompt word into a pre-trained model, so as to perform a claim processing on the claim event according to the claim rules in the claim prompt word to obtain a claim result. According to the claim result, the claim prompt word is updated to obtain an updated claim prompt word, and the updated claim prompt word is used as the second node for performing a strategy tree calculation. After the strategy tree calculation is completed, the target claim prompt word is determined based on the path parameter values ​​of each strategy tree path in the obtained strategy tree to perform project claim processing.

[0006] One or more embodiments of the present specification provide a computer-readable storage medium for storing computer-executable instructions, which implement the following process when executed: Obtain a claim prompt word for the insured project as the first node for calculating the strategy tree. Input the claim event and the claim prompt word into a pre-trained model to perform claim processing on the claim event according to the claim rules in the claim prompt word to obtain a claim result. Update the claim prompt word according to the claim result to obtain an updated claim prompt word, and use the updated claim prompt word as the second node for calculating the strategy tree. After the strategy tree calculation is completed, determine the target claim prompt word based on the path parameter values ​​of each strategy tree path in the obtained strategy tree to perform project claim processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate one or more embodiments of the present specification or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present specification. For users of ordinary skills in the art, other drawings can be obtained based on these drawings without creative labor.

[0008] Figure 1 A schematic diagram of an implementation environment of a project processing method based on a strategy tree provided for one or more embodiments of this specification;

[0009] Figure 2 A process flow chart of a project processing method based on a strategy tree provided for one or more embodiments of this specification;

[0010] Figure 3 A schematic diagram of a claim verification prompt word provided for one or more embodiments of this specification;

[0011] Figure 4A schematic diagram of a claims settlement event provided for one or more embodiments of this specification;

[0012] Figure 5 A schematic diagram of a comparison failure reason provided for one or more embodiments of this specification;

[0013] Figure 6 A schematic diagram of a strategy tree provided for one or more embodiments of this specification;

[0014] Figure 7 A process flow chart of a project processing method based on a policy tree applied to a health insurance project claim settlement scenario provided by one or more embodiments of this specification;

[0015] Figure 8 A schematic diagram of an embodiment of a project processing device based on a strategy tree provided by one or more embodiments of this specification;

[0016] Fig. 9 A schematic diagram of the structure of a project processing device based on a strategy tree provided in one or more embodiments of this specification. DETAILED DESCRIPTION

[0017] In order to enable users in the technical field to better understand the technical solutions in one or more embodiments of this specification, the technical solutions in one or more embodiments of this specification will be clearly and completely described below in conjunction with the drawings in one or more embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on one or more embodiments of this specification, all other embodiments obtained by ordinary technical users in this field without creative work should fall within the scope of protection of this document.

[0018] The project processing method based on the strategy tree provided in one or more embodiments of this embodiment can be applied to the implementation environment of the project processing based on the strategy tree. Figure 1 The implementation environment at least includes: a pre-trained model 101 for generating the claim result of the claim event, a server 102 for generating the claim prompt words, and in addition, the implementation environment may also include a claim model 103 for performing project claim processing based on the claim prompt words;

[0019] In this implementation environment, in the process of generating claims prompt words that can be used for the claims model 103, the server 102 obtains the claims prompt words of the insurance project as the first node for strategy tree calculation, inputs the claims event and the claims prompt words into the pre-trained model, and processes the claims event according to the claims rules in the claims prompt words to obtain the claims result, updates the claims prompt words according to the claims result to obtain updated claims prompt words, and uses the updated claims prompt words as the second node for strategy tree calculation. After the strategy tree calculation is completed, the target claims prompt words are determined according to the path parameter values ​​of each strategy tree path in the obtained strategy tree to perform project claims processing, thereby improving the accuracy of claims processing for different insurance projects based on the claims model 103.

[0020] One or more embodiments of a project processing method based on a strategy tree provided in this specification are as follows:

[0021] Reference Figure 2 The project processing method based on the strategy tree provided in this embodiment specifically includes steps S202 to S208.

[0022] Step S202, obtaining the claim verification prompt word of the insured item as the first node for calculating the strategy tree.

[0023] The claims verification in this embodiment refers to the claims review conducted by the insurance agency on the insurance program in which the insured person participates. For example, after the insured person of the medical insurance program becomes ill or receives medical treatment, he or she submits a claim application for the medical insurance program to the insurance agency. The insurance agency conducts a claims review on the insured person with reference to the claims rules of the medical insurance program to clarify whether the insured person can be compensated according to the compensation amount in the claims rules.

[0024] In actual applications, different claims processing for different insurance items usually refer to different claims rules. For example, in the process of claims processing for health insurance item A, the claims rules A are referred to, and in the process of claims processing for health insurance item B, the claims rules B are referred to. Therefore, in order to simplify the claims processing process and avoid confusion between different insurance items, the claims processing process of different insurance items can be distinguished with the help of claims models. In this embodiment, in the process of project claims processing through the claims model, in order to improve the processing efficiency and accuracy of the claims model, the target claims prompt words are obtained based on the strategy tree calculation, so that in the process of obtaining the claims results of the target claims event, the target claims prompt words are input into the claims model to obtain more accurate and targeted claims results. It should be noted that the claims model can be a large model with a Transformer neural network or other neural network as the architecture, such as a unified claims model.

[0025] In order to improve the processing efficiency and accuracy of the claims settlement model, prompts are usually used to guide the big model to perform specific tasks or generate specific types of answers or outputs. Specifically, the model can be guided to perform specific tasks or generate specific types of answers or outputs by inputting prompts into the big model. In this embodiment, claims settlement prompts are obtained by combining prompts with various types of protection items in the protection scenario, and the claims settlement prompts are input into the big model. The big model can identify the claims settlement prompts to determine the tasks to be performed, that is, perform claims settlement processing based on the claims settlement rules of the protection items. Specifically, the big model can distinguish the claims settlement rules for different types of protection items, and perform claims settlement processing with reference to the corresponding claims settlement rules.

[0026] Specifically, in the process of combining prompt words with guarantee items, in order to obtain prompt words that are more closely combined with guarantee items, this embodiment introduces strategy tree calculation, and obtains prompt words that are closely combined with guarantee items through calculation of the algorithm, so that the large model can obtain more accurate claims results in the process of project claims processing based on the prompt words. The strategy tree calculation may be a Monte Carlo tree search (MCTS), which regards the process of prompt word generation as a strategy planning problem, and obtains prompt words that are closely integrated with the guarantee project (i.e., target claims settlement prompt words) through four processes: node selection, node expansion, simulation, and back propagation. Node selection refers to node selection based on node parameters for each layer of the Monte Carlo tree. In this embodiment, the process of determining the first node is the process of node selection; node expansion refers to downward expansion of nodes. In this embodiment, the process of node expansion is the process of determining the second node based on the first node; simulation refers to calculating the node parameters of each node until a node whose node parameters are greater than a node parameter threshold is found, or refers to continuously performing node expansion until the node level of the obtained terminal node is equal to the preset node level, and back propagation refers to calculating the path parameters of a path passing through a certain node to determine the target claims settlement prompt word.

[0027] The claim prompt words refer to the prompt words used to guide the big model to conduct claim review, that is, the claim review prompt words. Specifically, since each insurance item has its own claim rules, the claim prompt words here are prompts for the claim rules that the big model needs to refer to. After clarifying the claim prompt words, the big model can determine the claim task based on the input claim prompt words and process the claim according to the corresponding claim rules. The claim prompt words may include one or more of the claim items, claim rules, claim rule summaries, and claim experience; for example, in a health insurance scenario, the claim prompt words may be as follows: Figure 3As shown in, it may include compensation items: "Pure outpatient determination in health insurance items", compensation rule summary: "Inpatient materials exist ->> compensation", and compensation experience: "If it is an outpatient surgery, the compensation requirement must be for treatment rather than just examination". Optionally, the compensation prompt includes a compensation prompt or a compensation task prompt.

[0028] During specific implementation, a target claim prompt word that can be used to guide the claim model to output a more accurate claim result is obtained through strategy tree calculation. Specifically, in the process of strategy tree calculation, it is first necessary to obtain the claim prompt word of the security project as the first node for strategy tree calculation. In this embodiment, the current claim prompt word of the security project can be obtained as the first node for strategy tree calculation, that is, step S202 can also be replaced by: obtaining the current claim prompt word of the security project as the first node for strategy tree calculation, and correspondingly, the first node can be the current node; wherein the current claim prompt word can be the updated claim prompt word corresponding to the second node; in addition, with the help of the claims experience of the claims personnel of the security agency, the preset claim prompt word of the security project can be obtained by manual writing as the first node for strategy tree calculation, that is, step S202 can also be replaced by: obtaining the preset claim prompt word of the security project as the first node for strategy tree calculation, in this case, the first node can be called the root node.

[0029] In addition, when the first node is the current node, after obtaining the claim verification prompt word of the insurance project as the first node step for performing the strategy tree calculation, it also includes: detecting whether the node parameter of the first node is greater than the node parameter threshold, if not, executing step S204; if so, determining that the strategy tree calculation is completed; or, after obtaining the claim verification prompt word of the insurance project as the first node step for performing the strategy tree calculation, it also includes: detecting whether the node level of the first node is equal to the preset node level, if not, executing step S204; if so, determining that the strategy tree calculation is completed.

[0030] Step S204, inputting the claim verification event and the claim verification prompt word into the pre-trained model, so as to perform claim verification processing on the claim verification event according to the claim verification rules in the claim verification prompt word to obtain a claim verification result.

[0031] After obtaining the claim prompt words of the insured item as the first node for strategy tree calculation, in this step, when the node parameter of the first node is less than or equal to the node parameter threshold, or when the node level of the first node is less than the preset node level, node expansion is performed based on the first node, and specifically the claim event and the claim prompt words are input into the pre-trained model. The claim event is processed according to the claim rules in the claim prompt words through the pre-trained model to obtain the claim result, so as to determine the second node according to the claim result.

[0032] The claim event refers to an event in which the insured person undergoes claim review based on medical materials (such as medical records) after receiving medical treatment. Therefore, the claim event can also be referred to as a medical treatment event. In this embodiment, the claim event can also be one or more claim events with preset claim results, i.e., preset claim events. The claim event includes medical records information.

[0033] The pre-trained model refers to a natural language model obtained by pre-training. The pre-trained model can adopt a neural network architecture containing a large number of parameters, or the pre-trained model can adopt a pre-trained large language model (LargeLanguage Model, LLM), such as chat GPT (chat Generative Pre-trained Transformer, generative pre-trained transformation model) and other open source large language models.

[0034] During the specific execution process, when it is detected that the node parameter of the first node is less than or equal to the node parameter threshold, or when it is detected that the node level of the first node is less than the preset node level, it indicates that the strategy tree has not been fully expanded and the node expansion needs to continue. Specifically, the claim event and the claim prompt word corresponding to the first node can be input into the pre-trained model, and the pre-trained model can be used to perform semantic recognition on the claim event to obtain the event features of the claim event, and detect whether the event features hit the claim rules in the claim prompt word. If so, it is determined that the claim result of the claim event is passed, indicating that compensation can be made for the claim event. If not, it is determined that the claim result of the claim event is failed, indicating that the claim event needs to be rejected.

[0035] For example, you can Figure 3 The claims reminder words and Figure 4 Each preset claim event shown is input into a pre-trained model, and semantic recognition is performed on each preset claim event through the pre-trained model to obtain the event features of each preset claim event, and detect whether the event features of each preset claim event hit the claims rules of each preset claim event. If each preset claim event contains a preset claim event that does not hit the claims rules, the claims result of the preset claim event that does not hit is claim failure. If each preset claim event contains a preset claim event that hits the claims rules, the claims result of the preset claim event that hits is claim approval.

[0036] Step S206: updating the claim verification prompt word according to the claim verification result to obtain an updated claim verification prompt word, and using the updated claim verification prompt word as the second node for calculating the strategy tree.

[0037] After the claim event and the claim prompt word are input into the pre-trained model to process the claim event according to the claim rules in the claim prompt word to obtain the claim result, in this step, the claim prompt word is updated according to the claim result to obtain an updated claim prompt word, and the updated claim prompt word is used as the second node for calculating the strategy tree. The second node includes the extended node obtained by node expansion of the first node.

[0038] During the specific implementation process, after obtaining the claim verification result output by the pre-trained model, the claim verification prompt word can be updated according to the claim verification result. In an optional implementation provided by this embodiment, the claim verification prompt word is updated according to the claim verification result in the following manner to obtain an updated claim verification prompt word:

[0039] Comparing the claim settlement result with the preset claim settlement result of the claim settlement event;

[0040] If the comparison result is a comparison failure, the task prompt word and the compensation result are input into the analysis model to analyze the cause of failure and obtain the reason for the comparison failure;

[0041] The claim verification prompt word is updated based on the reason for the comparison failure to obtain the updated claim verification prompt word.

[0042] Specifically, after obtaining the claims result of the claims event, the claims result is compared with the preset claims result of the claims event. If the comparison result is a comparison failure, it indicates that the claims result of the claims event is inconsistent with the preset claims result. The task prompt words and the claims result can be input into the analysis model to analyze the reasons for the comparison failure and obtain the reasons for the comparison failure. After obtaining the reasons for the comparison failure, a prompt word update strategy can be determined based on the reasons for the comparison failure, and the claims prompt words can be updated according to the prompt word update strategy to obtain updated claims prompt words. If the comparison result is a comparison success, it indicates that the claims result of the claims event is consistent with the preset claims result and no processing is required.

[0043] The task prompt word may be a text that guides the analysis model to perform the cause analysis task. For example, the task prompt word may be "analyze the reasons why the claim result of the claim settlement event is inconsistent with the preset claim result". By inputting this task prompt word and the claim result of the claim settlement event into the analysis model for comparison failure reason analysis, the comparison failure reason such as "the comparison result of the claim settlement event is a comparison failure because the pre-trained model did not accurately identify whether the medical record involves relevant treatment methods for major diseases or rare diseases" may be obtained;

[0044] like Figure 5As shown, after obtaining the reason for the failed comparison, the prompt word update strategy is determined based on the reason for the failed comparison: "1. Major diseases or rare diseases need to be defined in more detail, such as providing more disease examples and disease names. 2. Highlight the characteristics of major disease treatment", and the prompt word update strategy is added to the claim prompt words to obtain updated claim prompt words: claim items: "Pure outpatient judgment in health insurance items", claim rule summary: "There are hospitalization materials ->> payment", and claim experience: "1. If it is an outpatient surgery, the compensation requirement must be for treatment rather than just examination; 2. Examples of major diseases or rare diseases: malignant tumors, leukemia; 3. The characteristic of major disease treatment is the use of chemotherapy and radiotherapy."

[0045] In the specific execution process, after the updated claim verification prompt word is used as the second node, the node parameter of the second node needs to be calculated until a node whose node parameter is greater than the node parameter threshold is found, thereby determining that the strategy tree calculation is completed. In an optional implementation provided by this embodiment, after the updated claim verification prompt word is used as the second node for strategy tree calculation, the following operations are performed:

[0046] Calculating a node parameter value of the second node, and detecting whether the node parameter value is greater than a node parameter value threshold;

[0047] If so, it is determined that the strategy tree calculation is completed, and the strategy tree is obtained; optionally, if the first node is a root node, the strategy tree includes the first node and the second node, and if the first node is not a root node, the strategy tree also includes an upper node whose node level is before the first node;

[0048] If not, it indicates that the second node cannot be used as the end node of the strategy tree, and further node expansion needs to be performed based on the second node until the end node of the strategy tree is found.

[0049] Specifically, after the claim prompt word is updated according to the claim result to obtain an updated claim prompt word, and the updated claim prompt word is used as the second node for strategy tree calculation, the node parameter value of the second node is calculated, and it is detected whether the node parameter value is greater than the node parameter value threshold. If so, it is determined that the strategy tree calculation is completed to obtain the strategy tree; if not, the node is expanded based on the second node. In an optional implementation provided by this embodiment, the node is expanded based on the second node in the following manner: the claim event and the intermediate claim prompt word are input into the pre-trained model for claim processing and the prompt word is updated according to the claim result to determine the remaining nodes; the node parameter values ​​of the remaining nodes are calculated, and when the node parameter values ​​of the remaining nodes are greater than the parameter value threshold, it is determined that the strategy tree calculation is completed to obtain the strategy tree.

[0050] The intermediate claim verification prompt word is the claim verification prompt word corresponding to the intermediate node whose node parameter value is updated to be less than or equal to the node parameter value threshold. The remaining nodes may be a node set including an extended node obtained by performing node expansion based on the second node. When the node parameter of the second node is less than or equal to the node parameter threshold, node expansion needs to be performed based on the second node until an end node whose node parameter is greater than the node parameter threshold is found.

[0051] It should be noted that the above-mentioned process of obtaining the second node by performing node expansion based on the first node can be used as a reference, and any node expansion process can refer to the above-mentioned process of obtaining the second node by performing node expansion based on the first node. For example, when the second node is obtained and the node parameter of the second node is less than the node parameter threshold or the node level of the second node is less than the preset node level, the updated claim prompt word corresponding to the second node is used as the first node, and the claim event and the updated claim prompt word corresponding to the first node are input into the pre-trained model to process the claim event according to the claim rules in the updated claim prompt word to obtain the claim result; the updated claim prompt word is updated for the second time according to the claim result to obtain the secondary updated claim prompt word, and the secondary updated claim prompt word is used as the second node for strategy tree calculation.

[0052] Furthermore, in an optional implementation manner provided by this embodiment, the node parameter value of the second node is calculated in the following manner, including:

[0053] Inputting the updated claim verification prompt word corresponding to the second node and the claim verification event into the pre-trained model for claim verification processing to obtain an updated claim verification result of the claim verification event;

[0054] Based on the preset claim result of the claim event, the updated claim result and the number of claim events, the accuracy of the updated claim prompt word is calculated, and the accuracy is used as the node parameter value of the second node.

[0055] For example, the updated claim prompt corresponding to the second node is: claim item: "Pure outpatient judgment in health insurance items", claim rule summary: "There are hospitalization materials ->> compensation", and claim experience: "1. If it is an outpatient surgery, the compensation requirement must be for treatment rather than just examination; 2. Examples of major diseases or rare diseases: malignant tumors, leukemia; 3. The characteristics of major disease treatment are the use of chemotherapy and radiotherapy". By inputting the updated claim prompt and 3 preset claim events into the training model for claim processing, the updated claim results of the 3 preset claim events are compensation, rejection, and compensation respectively; if the preset claim results of the third preset claim event are rejection, rejection, and compensation respectively, the correct rate of the updated claim prompt can be calculated as: correct result / number of results = 2 / 3. It should be noted that the node parameter value of any node in the strategy tree can be calculated with reference to the above calculation method.

[0056] In addition, after the updated claim verification prompt word is used as the second node, it is also possible to determine whether the strategy tree calculation is completed by detecting whether the node level of the second node is equal to the preset node level. In an optional implementation provided by this embodiment, after the updated claim verification prompt word is used as the second node for strategy tree calculation, the following operations are performed:

[0057] Detecting whether the node level of the second node is equal to a preset node level;

[0058] If yes, it is determined that the strategy tree calculation is completed, and the strategy tree is obtained;

[0059] If not, continue to expand the nodes until the preset node level is reached. Specifically, the updated claims prompt word corresponding to the second node can be updated multiple times to determine the remaining nodes.

[0060] Specifically, after the claim verification prompt word is updated according to the claim verification result to obtain the updated claim verification prompt word, and the updated claim verification prompt word is used as the second node for strategy tree calculation, it is detected whether the node level of the second node in the current strategy tree is equal to the preset node level. If so, it is determined that the calculation is completed and the strategy tree is obtained. If not, the node expansion continues until the preset node level is reached.

[0061] In an optional implementation manner provided in this embodiment, node expansion is performed in the following manner:

[0062] Input the claim verification event and the obtained updated claim verification prompt words into the pre-trained model for claim verification processing, and update the prompt words according to the claim verification result to determine the remaining nodes;

[0063] It is detected whether the node levels of the remaining nodes are equal to the preset node level, and when the node levels of the remaining nodes are equal to the preset node level, it is determined that the strategy tree calculation is completed, and the strategy tree is obtained.

[0064] Specifically, the updated claims prompt words corresponding to the second node may be updated multiple times to determine the remaining nodes, by inputting the claims event and the obtained updated claims prompt words into the pre-trained model for claims processing, and updating the prompt words according to the claims results to determine the remaining nodes, and detecting whether the node level of the remaining nodes is equal to the preset node level. When the node level of the remaining nodes is equal to the preset node level, it is determined that the strategy tree calculation is completed to obtain the strategy tree; optionally, when the first node is the root node, the strategy tree includes the first node and the second node. In addition, if the first node is not the root node, the strategy tree includes the first node, the second node, and an upper node whose node level is before the first node.

[0065] The following reference Figure 6 , the process of generating the target claim prompt word for any insurance project through strategy tree calculation is explained: the preset claim prompt word determined manually is used as the root node for strategy tree calculation , if the node parameter of the root node is less than or equal to the node parameter value threshold, or the node level of the root node is less than the preset node level, the next round of node expansion is performed, that is, the root node The corresponding preset claim prompt words and claim events are input into the pre-trained model for claim processing to obtain the claim result, and the preset claim prompt words are updated according to the claim result to obtain the updated claim prompt word 1, and the updated claim prompt word 1 is used as the intermediate node , and further calculate the intermediate nodes Node parameters Is it greater than the node parameter threshold, until a leaf node with a node parameter greater than the node parameter threshold appears (in the figure ), determine that the strategy tree calculation is completed, or detect whether the node level of the intermediate node is equal to the preset node level, until a leaf node with a node level equal to the preset node level appears (in the figure, the fourth level ), confirm that the strategy tree calculation is completed, and obtain the root node , intermediate node and leaf nodes A strategy tree, where - - - is strategy tree path 1, where - - - The strategy tree path 2 is a strategy tree path 2. The determination process of the strategy tree path 2 can refer to the above determination process of the strategy tree path 1, and this embodiment will not be repeated here.

[0066] Among them, the process of calculating the node parameters of the root node may include: inputting the preset claim prompt word and claim event corresponding to the root node into the pre-trained model for claim processing, obtaining the claim result of the claim event, calculating the accuracy of the preset claim prompt word based on the preset claim result of the claim event, the claim result and the number of claim events, and using the accuracy as the node parameter value of the root node. It should be noted that the node parameters of any node can be obtained through a calculation process similar to the above.

[0067] Step S208, after the strategy tree calculation is completed, the target claim verification prompt word is determined according to the path parameter values ​​of each strategy tree path in the obtained strategy tree to perform project claim verification processing.

[0068] After the above-mentioned updating of the claim prompt word according to the claim result to obtain the updated claim prompt word and using the updated claim prompt word as the second node for the strategy tree calculation, in this step, after the strategy tree calculation is completed, the target claim prompt word is determined according to the path parameter value of each strategy tree path in the obtained strategy tree to perform project claim processing. The project claim processing refers to the claim processing for the insured project.

[0069] In the specific implementation, in the process of determining the target claim verification prompt word, the claim verification prompt word corresponding to the end node of the target strategy tree path whose path parameter value in each strategy tree path is greater than the path parameter value threshold can be used as the target claim verification prompt word. In an optional implementation provided by this embodiment, the target claim verification prompt word is determined according to the obtained path parameter value of each strategy tree path in the strategy tree in the following manner:

[0070] Calculating a path parameter value of each strategy tree path in at least one strategy tree path of the strategy tree;

[0071] determining a target strategy tree path in the at least one strategy tree path based on the path parameter value;

[0072] The claim verification prompt word corresponding to the end node of the target strategy tree path is used as the target claim verification prompt word.

[0073] Specifically, the strategy tree path whose path parameter value is higher than the parameter value threshold can be used as the target strategy tree path, and the claim verification prompt word corresponding to the end node of the target strategy tree path can be used as the target claim verification prompt word. Further, in the process of calculating the path parameter value of each strategy tree path in the strategy tree, in an optional implementation provided by this embodiment, the path parameter value of each strategy tree path in the strategy tree is calculated in the following manner:

[0074] Determine at least one node included in each of the strategy tree paths;

[0075] The path parameter values ​​of the strategy tree paths are calculated based on the number of nodes and the node parameter value of the at least one node.

[0076] For example, to obtain Figure 6 After the strategy tree shown, - - - This strategy tree path is taken as an example, based on , , , The node parameters of the four nodes are 0, The average node parameter value is calculated as the path parameter value by using the number of nodes 4. If the path parameter value is greater than the path parameter value threshold, - - - The terminal node in the strategy tree path The corresponding claim verification prompt word is used as the target claim verification prompt word, where r can be the reward value, that is, the accuracy of the claim verification prompt word.

[0077] In the specific execution process, after the target claim verification prompt word is determined according to the path parameter values ​​of each strategy tree path in the obtained strategy tree, the target claim verification prompt word can be used as the standard claim verification prompt word for the guarantee item. In an optional implementation provided by this embodiment, after the operation of determining the target claim verification prompt word according to the path parameter values ​​of each strategy tree path in the obtained strategy tree is executed, it also includes:

[0078] Associating the target claim verification prompt word with the guarantee item;

[0079] The associated target claim prompt words are stored to perform project claim processing on the target claim event under the protection project based on the target claim prompt words.

[0080] Specifically, after determining the target claim prompt words, the target claim prompt words can be used as standard claim prompt words for the protection project. In this way, when processing the claims for the protection project, the user does not need to edit the prompt words by himself. Instead, the standard claim prompt words are used to input the claims model for project claims. Specifically, by associating the target claim prompt words with the protection project, the associated target claim prompt words are stored, so that when it is necessary to process the target claim event under the protection project, the target claim prompt words can be called according to the protection project to process the target claim event for project claims.

[0081] Furthermore, in the process of project compensation processing, in an optional implementation manner provided in this embodiment, project compensation processing is performed in the following manner:

[0082] Input the target claim prompt word and the target claim event into the claim model to perform project claim, and obtain the claim result of the target claim event;

[0083] Optionally, the claims model performs semantic recognition on the target claims prompt words, and performs project claims on the target claims event according to the recognition result to obtain the claims result.

[0084] To summarize, the project processing method based on the strategy tree provided in this embodiment obtains the claim prompt words of the insured project as the first node for strategy tree calculation, inputs the claim event and the claim prompt words into the pre-trained model, and processes the claim event according to the claim rules in the claim prompt words to obtain the claim result, updates the claim prompt words according to the claim result to obtain the updated claim prompt words, and uses the updated claim prompt words as the second node for strategy tree calculation. After the strategy tree calculation is completed, the target claim prompt words are determined according to the path parameter values ​​of each strategy tree path in the obtained strategy tree to perform project claim processing, thereby improving the effectiveness of the claim prompt words, thereby improving the efficiency and success rate of claims.

[0085] The following takes the application of a project processing method based on a strategy tree provided in this embodiment in a health insurance project claim settlement scenario as an example. Figure 7 , the project processing method based on the strategy tree provided in this embodiment is further described, see Figure 7 , a project processing method based on a strategy tree applied to the claims settlement scenario of a health insurance project specifically includes the following steps.

[0086] Step S702, obtaining the claim prompt word of the health insurance item as the first node for calculating the strategy tree.

[0087] Optionally, the claim verification prompt word may be a preset claim verification prompt word set manually.

[0088] Step S704: when the node parameter of the first node is less than or equal to the node parameter threshold, the claim verification event and the claim verification prompt word are input into the pre-training model.

[0089] Optionally, step S704 may also be replaced by inputting the claims verification event and the claims verification prompt word into the pre-trained model when the node level of the first node is less than the preset node level.

[0090] Step S706, comparing the claim result with the preset claim result of the claim event.

[0091] Step S708: If the comparison result is a comparison failure, the task prompt word and the claim verification result are input into the analysis model to analyze the reason for the comparison failure and obtain the reason for the comparison failure.

[0092] Step S710, updating the claim verification prompt word based on the reason for the comparison failure, obtaining an updated claim verification prompt word, and using the updated claim verification prompt word as the second node for calculating the strategy tree.

[0093] Step S712, calculate the node parameter value of the second node, and detect whether the node parameter value is greater than the node parameter value threshold. If so, determine that the strategy tree calculation is completed and obtain the strategy tree.

[0094] Optionally, detect whether the node parameter value is greater than the node parameter value threshold. If so, determine that the strategy tree calculation is completed, and obtain the strategy tree. It can also be replaced by: when the node level of the second node is equal to the preset node level, determine that the strategy tree calculation is completed, and obtain the strategy tree.

[0095] Step S714, determining each node included in each strategy tree path in the strategy tree.

[0096] Step S716, calculating the path parameter value based on the node parameter value of each node and the number of nodes.

[0097] Step S718, taking the claims verification prompt word corresponding to the end node of the target strategy tree path whose path parameter value in each strategy tree path is greater than the path parameter value threshold as the target claims verification prompt word.

[0098] Step S720, associating the target claim verification prompt word with the health insurance item, and storing the associated target claim verification prompt word.

[0099] Step S722, input the target claim prompt word and the target claim event into the claim model to perform project claim and obtain the claim result of the target claim event.

[0100] An embodiment of a project processing device based on a strategy tree provided in this specification is as follows:

[0101] In the above-mentioned embodiment, a project processing method based on a strategy tree is provided, and correspondingly, a project processing device based on a strategy tree is also provided, which will be described below with reference to the accompanying drawings.

[0102] Reference Figure 8 , which shows a schematic diagram of a project processing device based on a strategy tree provided in this embodiment.

[0103] Since the device embodiment corresponds to the method embodiment, the description is relatively simple, and the relevant parts can refer to the corresponding description of the method embodiment provided above. The device embodiment described below is only illustrative.

[0104] This embodiment provides a project processing device based on a strategy tree, including:

[0105] The first node determination module 802 is configured to obtain a claim verification prompt word of the insured item as the first node for calculating the strategy tree;

[0106] The claim processing module 804 is configured to input the claim event and the claim prompt word into the pre-trained model, so as to perform claim processing on the claim event according to the claim rules in the claim prompt word to obtain a claim result;

[0107] The second node determination module 806 is configured to update the claim verification prompt word according to the claim verification result to obtain an updated claim verification prompt word, and use the updated claim verification prompt word as the second node for calculating the strategy tree;

[0108] After the strategy tree calculation is completed, the prompt word determination module 808 is run, and the prompt word determination module 808 is configured to determine the target claim prompt word according to the path parameter value of each strategy tree path in the obtained strategy tree to perform project claim processing.

[0109] An embodiment of a project processing device based on a strategy tree provided in this specification is as follows:

[0110] Corresponding to the above-described project processing method based on a strategy tree, based on the same technical concept, one or more embodiments of this specification further provide a project processing device based on a strategy tree, wherein the project processing device based on a strategy tree is used to execute the above-provided project processing method based on a strategy tree. Fig. 9 A schematic diagram of the structure of a project processing device based on a strategy tree provided in one or more embodiments of this specification.

[0111] This embodiment provides a project processing device based on a strategy tree, including:

[0112] like Fig. 9As shown, the project processing device based on the policy tree may have relatively large differences due to different configurations or performances, and may include one or more processors 901 and memory 902, and the memory 902 may store one or more storage applications or data. Among them, the memory 902 may be a short-term storage or a persistent storage. The application stored in the memory 902 may include one or more modules (not shown in the figure), and each module may include a series of computer executable instructions in the project processing device based on the policy tree. Furthermore, the processor 901 may be configured to communicate with the memory 902, and execute a series of computer executable instructions in the memory 902 on the project processing device based on the policy tree. The project processing device based on the policy tree may also include one or more power supplies 903, one or more wired or wireless network interfaces 904, one or more input / output interfaces 905, one or more keyboards 906, etc.

[0113] In a specific embodiment, the project processing device based on the strategy tree includes a memory and one or more programs, wherein the one or more programs are stored in the memory, and the one or more programs may include one or more modules, and each module may include a series of computer executable instructions in the project processing device based on the strategy tree, and the one or more programs are configured to be executed by one or more processors, including the following computer executable instructions:

[0114] Obtain the claim verification prompt word of the insured item as the first node for strategy tree calculation;

[0115] Inputting the claim event and the claim prompt word into the pre-trained model, so as to process the claim event according to the claim rules in the claim prompt word to obtain a claim result;

[0116] The claim verification prompt word is updated according to the claim verification result to obtain an updated claim verification prompt word, and the updated claim verification prompt word is used as the second node for calculating the strategy tree;

[0117] After the strategy tree calculation is completed, the target claims prompt word is determined according to the path parameter values ​​of each strategy tree path in the strategy tree to carry out project claims processing.

[0118] An embodiment of a computer-readable storage medium provided in this specification is as follows:

[0119] Corresponding to the above-described project processing method based on a strategy tree, based on the same technical concept, one or more embodiments of this specification also provide a computer-readable storage medium.

[0120] The computer-readable storage medium provided in this embodiment is used to store computer-executable instructions, and the computer-executable instructions implement the following process when executed:

[0121] Obtain the claim verification prompt word of the insured item as the first node for strategy tree calculation;

[0122] Inputting the claim event and the claim prompt word into the pre-trained model, so as to process the claim event according to the claim rules in the claim prompt word to obtain a claim result;

[0123] The claim verification prompt word is updated according to the claim verification result to obtain an updated claim verification prompt word, and the updated claim verification prompt word is used as the second node for calculating the strategy tree;

[0124] After the strategy tree calculation is completed, the target claims prompt word is determined according to the path parameter values ​​of each strategy tree path in the strategy tree to carry out project claims processing.

[0125] It should be noted that an embodiment of a computer-readable storage medium in this specification and an embodiment of a project processing method based on a strategy tree in this specification are based on the same inventive concept, so the specific implementation of this embodiment can refer to the implementation of the aforementioned corresponding method, and the repeated parts will not be repeated.

[0126] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referenced to each other. Each embodiment focuses on the differences from other embodiments. For example, the device embodiment, equipment embodiment and computer-readable storage medium embodiment are similar to the method embodiment, so the description is relatively simple. To read the relevant contents in the device embodiment, equipment embodiment and computer-readable storage medium embodiment, please refer to the partial description of the method embodiment.

[0127] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0128] In the 1930s, it was very clear whether the improvement of a technology was hardware improvement (for example, improvement of the circuit structure of diodes, transistors, switches, etc.) or software improvement (improvement of the method flow). However, with the development of technology, many improvements of the method flow today can be regarded as direct improvements of the hardware circuit structure. Design users almost always obtain the corresponding hardware circuit structure by programming the improved method flow into the hardware circuit. Therefore, it cannot be said that the improvement of a method flow cannot be implemented with a hardware entity module. For example, a programmable logic device (PLD) (such as a field programmable gate array (FPGA)) is such an integrated circuit whose logical function is determined by the user's programming of the device. The design user can "integrate" a digital system on a PLD by programming it himself, without having to ask a chip manufacturer to design and make a dedicated integrated circuit chip. Moreover, nowadays, instead of manually making integrated circuit chips, this kind of programming is mostly implemented by "logic compiler" software, which is similar to the software compiler used when developing and writing programs. The original code before compilation must also be written in a specific programming language, which is called hardware description language (HDL). There is not only one HDL, but many kinds, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, RHDL (Ruby Hardware Description Language), etc. The most commonly used ones are VHDL (Very-High-Speed ​​Integrated Circuit Hardware Description Language) and Verilog. Technical users in this field should also be aware that only by slightly programming the method flow in the above-mentioned hardware description languages ​​and programming it into the integrated circuit, the hardware circuit that implements the logic method flow can be easily obtained.

[0129] The controller can be implemented in any suitable manner, for example, the controller can take the form of a microprocessor or processor and a computer-readable medium storing a computer-readable program code (such as software or firmware) executable by the (micro)processor, a logic gate, a switch, an application-specific integrated circuit (ASIC), a programmable logic controller, and an embedded microcontroller. Examples of controllers include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051F320. The memory controller can also be implemented as part of the control logic of the memory. Users of the art also know that in addition to implementing the controller in a purely computer-readable program code manner, the controller can be implemented in the form of a logic gate, a switch, an application-specific integrated circuit, a programmable logic controller, and an embedded microcontroller by logically programming the method steps. Therefore, such a controller can be considered as a hardware component, and the devices for implementing various functions included therein can also be regarded as structures within the hardware component. Or even, the devices for implementing various functions can be regarded as both software modules for implementing the method and structures within the hardware component.

[0130] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by commodities with certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0131] For the convenience of description, the above devices are described in terms of functions and are divided into various units. Of course, when implementing the embodiments of this specification, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0132] It should be understood by technical users in the art that one or more embodiments of this specification may be provided as a method, system or computer program commodity. Therefore, one or more embodiments of this specification may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, this specification may take the form of a computer program commodity implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0133] This specification is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of this specification. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0134] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0135] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0136] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0137] Memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0138] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined in this article, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0139] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0140] One or more embodiments of the present specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. One or more embodiments of the present specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0141] The above description is only an embodiment of this document and is not intended to limit this document. For users of this technical field, this document may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this document should be included in the scope of the claims of this document.

Claims

1. A project processing method based on a strategy tree, comprising: Obtain the claim verification prompt word of the insured item as the first node for strategy tree calculation; Inputting the claim event and the claim prompt word into the pre-trained model, so as to process the claim event according to the claim rules in the claim prompt word to obtain a claim result; Compare the claim result with the preset claim result of the claim event. If the comparison result is a comparison failure, input the task prompt word and the claim result into the analysis model to analyze the failure cause, update the claim prompt word based on the comparison failure cause obtained to obtain an updated claim prompt word, and use the updated claim prompt word as the second node for calculating the strategy tree; After the strategy tree calculation is completed, the target claims prompt words are determined according to the path parameter values ​​of each strategy tree path in the strategy tree to carry out project claims processing.

2. According to the project processing method based on the strategy tree of claim 1, the claim result is compared with the preset claim result of the claim event. If the comparison result is a comparison failure, the task prompt word and the claim result are input into the analysis model to analyze the cause of failure, and the claim prompt word is updated based on the obtained comparison failure cause to obtain an updated claim prompt word, and the updated claim prompt word is used as the second node step of the strategy tree calculation. After the strategy tree calculation is completed, the target claim prompt word is determined according to the path parameter value of each strategy tree path in the obtained strategy tree to perform the project claim processing step, and it also includes: Calculating a node parameter value of the second node, and detecting whether the node parameter value is greater than a node parameter value threshold; If so, it is determined that the strategy tree calculation is completed, and the strategy tree is obtained.

3. According to the project processing method based on the strategy tree in claim 2, the step of calculating the node parameter value of the second node comprises: Inputting the updated claim verification prompt word corresponding to the second node and the claim verification event into the pre-trained model for claim verification processing to obtain an updated claim verification result of the claim verification event; Based on the preset claim result of the claim event, the updated claim result and the number of claim events, the accuracy of the updated claim prompt word is calculated, and the accuracy is used as the node parameter value of the second node.

4. According to the project processing method based on the strategy tree in claim 2, if the result of the step of calculating the node parameter value of the second node and detecting whether the node parameter value is greater than the node parameter value threshold is no, perform the following operations: Input the claim settlement event and the intermediate claim settlement prompt words into the pre-trained model for claim settlement processing and update the prompt words according to the claim settlement result to determine the remaining nodes; Calculate the node parameter values ​​of the remaining nodes, determine that the strategy tree calculation is completed when the node parameter values ​​of the remaining nodes are greater than the parameter value threshold, and obtain the strategy tree.

5. According to the project processing method based on the strategy tree of claim 1, the claim result is compared with the preset claim result of the claim event. If the comparison result is a comparison failure, the task prompt word and the claim result are input into the analysis model to analyze the cause of failure, and the claim prompt word is updated based on the obtained comparison failure cause to obtain an updated claim prompt word, and the updated claim prompt word is used as the second node step of the strategy tree calculation. After the strategy tree calculation is completed, the target claim prompt word is determined according to the path parameter value of each strategy tree path in the obtained strategy tree to perform the project claim processing step, and it also includes: Detecting whether the node level of the second node is equal to a preset node level; If so, it is determined that the strategy tree calculation is completed, and the strategy tree is obtained.

6. According to the project processing method based on the strategy tree in claim 5, if the result of the step of detecting whether the node level of the second node is equal to the preset node level is no, the following operations are performed: Input the claim verification event and the obtained updated claim verification prompt word into the pre-trained model for claim verification processing, and update the prompt word according to the claim verification result to determine the remaining nodes; It is detected whether the node levels of the remaining nodes are equal to the preset node level, and when the node levels of the remaining nodes are equal to the preset node level, it is determined that the strategy tree calculation is completed, and the strategy tree is obtained.

7. According to the project processing method based on the strategy tree in claim 1, the step of determining the target claim verification prompt word according to the obtained path parameter value of each strategy tree path in the strategy tree comprises: Calculating a path parameter value of each strategy tree path in at least one strategy tree path of the strategy tree; determining a target strategy tree path in the at least one strategy tree path based on the path parameter value; The claim verification prompt word corresponding to the end node of the target strategy tree path is used as the target claim verification prompt word.

8. The project processing method based on the strategy tree according to claim 7, wherein the step of calculating the path parameter value of each strategy tree path in at least one strategy tree path of the strategy tree comprises: Determine at least one node included in each of the strategy tree paths; The path parameter values ​​of the strategy tree paths are calculated based on the number of nodes and the node parameter value of the at least one node.

9. According to the project processing method based on the strategy tree in claim 1, after the operation of determining the target claim verification prompt word according to the path parameter value of each strategy tree path in the obtained strategy tree is performed, it also includes: Associating the target claim verification prompt word with the guarantee item; The associated target claim prompt words are stored to perform project claim processing on the target claim event under the protection project based on the target claim prompt words.

10. According to the project processing method based on the strategy tree in claim 1, the project claim processing comprises: Input the target claim prompt and the target claim event under the insured project into the claim model to perform project claim settlement and obtain the claim result of the target claim event; The claims model performs semantic recognition on the target claims prompt words, and performs project claims on the target claims event according to the recognition result to obtain the claims result.

11. A project processing device based on a strategy tree, comprising: A first node determination module is configured to obtain a claim verification prompt word of the insured item as a first node for calculating the strategy tree; A claim processing module is configured to input the claim event and the claim prompt word into a pre-trained model, so as to perform claim processing on the claim event according to the claim rules in the claim prompt word to obtain a claim result; The second node determination module is configured to compare the claim result with the preset claim result of the claim event, and if the comparison result is a comparison failure, input the task prompt word and the claim result into the analysis model to analyze the failure cause, update the claim prompt word based on the obtained comparison failure cause to obtain an updated claim prompt word, and use the updated claim prompt word as the second node for calculating the strategy tree; After the strategy tree calculation is completed, the prompt word determination module is run, and the prompt word determination module is configured to determine the target claim prompt word according to the path parameter value of each strategy tree path in the obtained strategy tree to perform project claim processing.

12. A project processing device based on a strategy tree, comprising: processor; and a memory configured to store computer executable instructions that, when executed, cause the processor to: Obtain the claim verification prompt word of the insured item as the first node for strategy tree calculation; Inputting the claim event and the claim prompt word into the pre-trained model, so as to process the claim event according to the claim rules in the claim prompt word to obtain a claim result; Compare the claim result with the preset claim result of the claim event. If the comparison result is a comparison failure, input the task prompt word and the claim result into the analysis model to analyze the failure cause, update the claim prompt word based on the comparison failure cause obtained to obtain an updated claim prompt word, and use the updated claim prompt word as the second node for calculating the strategy tree; After the strategy tree calculation is completed, the target claims prompt words are determined according to the path parameter values ​​of each strategy tree path in the strategy tree to carry out project claims processing.

13. A computer-readable storage medium for storing computer-executable instructions, wherein the computer-executable instructions implement the steps of the method of claim 1 when executed.

Images