Artificial intelligence-based request processing method, device, computer equipment, and medium

Abstract

This application belongs to the field of artificial intelligence technology and relates to an artificial intelligence-based request processing method, device, computer equipment and storage medium, including: extracting the consultation content from the received user-triggered consultation request, and performing intent recognition on the consultation content to obtain the user's intention; if the user's intention is not a department that meets the specified conditions, performing department prediction processing on the consultation content, the user's age and gender based on the referral model to obtain a predicted department list result; if the current receiving department model and the predicted department list result do not meet the preset matching conditions, filtering out the target predicted department with the highest prediction value from the predicted department list result; calling the target department model corresponding to the target predicted department, and performing consultation processing on the user. In addition, user intention can be stored in the blockchain. Through this application, the best and most appropriate department model can be intelligently selected based on the user's actual personal situation for accurate consultation, thereby improving the accuracy of the consultation.

Description

Technical Field

[0001] The present application relates to the fields of artificial intelligence development technology and digital medicine, and in particular to artificial intelligence-based request processing methods, devices, computer equipment, and storage media. Background Art

[0002] With the rapid development of digital healthcare, intelligent medical consultation robots are playing an increasingly important role in modern healthcare. These robots primarily collect disease symptoms for real doctors, serving as an auxiliary tool for preliminary diagnosis. During a patient consultation, the medical robot first engages in a conversation with the patient, gathering basic symptoms through this interaction. However, each patient presents with unique symptoms during each consultation, posing a significant challenge for intelligent medical consultation robots: how to provide more precise consultation services for each patient.

[0003] Currently, intelligent medical consultation robots face technical limitations. Traditionally, the medical system relies on a single, generalist robot. While these robots can cover a wide range of medical fields when addressing various medical conditions, they are unable to accurately select the optimal and most appropriate department model based on the patient's individual circumstances. This can lead to the robot failing to accurately answer patients' questions and potentially misleading them through incorrect responses, compromising the accuracy and effectiveness of the consultation. Summary of the Invention

[0004] The purpose of the embodiments of the present application is to propose an artificial intelligence-based request processing method, device, computer equipment and storage medium to solve the technical problem that the existing medical consultation robot is unable to select the best and most appropriate department model for accurate consultation based on the patient's actual personal situation, and the consultation accuracy is low.

[0005] In order to solve the above technical problems, the present application provides an artificial intelligence-based request processing method, which adopts the following technical solutions:

[0006] When receiving a user-triggered consultation request, extracting consultation content from the consultation request; wherein the consultation request carries the consultation content;

[0007] Performing intent recognition on the medical inquiry content to obtain the corresponding user intent, and determining whether the user intent is for a department that meets specified conditions;

[0008] If the user's intention is not a department that meets the specified conditions, then obtain the user's age and gender;

[0009] Performing department prediction processing on the medical inquiry content, the age, and the gender based on a preset referral model to obtain a corresponding predicted department list result;

[0010] Determine whether the current receiving department model and the predicted department list result meet the preset matching conditions;

[0011] If the matching condition is not met, the target prediction department with the highest prediction value is selected from the prediction department list results;

[0012] A target department model corresponding to the target predicted department is called, and a medical consultation process is performed on the user based on the target department model.

[0013] Furthermore, before the step of performing department prediction processing on the medical inquiry content, the age, and the gender based on the preset referral model to obtain a corresponding predicted department list result, the method further includes:

[0014] Obtain pre-collected medical consultation text data;

[0015] Annotating the medical consultation text data to obtain corresponding initial sample data;

[0016] Preprocessing the initial sample data to obtain corresponding target sample data;

[0017] Calling a pre-trained language model, and performing model structure optimization processing on the pre-trained language model to obtain a corresponding first specified model;

[0018] Based on a preset construction strategy, the first specified model is trained and evaluated using the target sample data to obtain a second specified model that meets the preset construction requirements;

[0019] The second designated model is used as the referral model.

[0020] Furthermore, the step of performing model structure optimization processing on the pre-trained language model to obtain the corresponding first specified model specifically includes:

[0021] Get all preset departments;

[0022] Constructing a binary classification prediction head corresponding to each of the departments;

[0023] Connecting the output of the pre-trained language model to the input of each of the two-class prediction heads to obtain a corresponding first language model;

[0024] Applying a preset target activation function to the output of each of the two-class prediction heads to obtain a corresponding second language model;

[0025] The second language model is used as the first designated model.

[0026] Furthermore, the step of performing intent recognition on the medical inquiry content to obtain the corresponding user intent, and determining whether the user intent is for a department that meets specified conditions, specifically includes:

[0027] Call the preset intent recognition model;

[0028] Performing intent recognition on the medical inquiry content based on the intent recognition model to obtain a corresponding intent recognition label, and using the intent recognition label as the user intention;

[0029] Get a predefined set of intent tags;

[0030] If the user intention is included in the intention tag set, it is determined that the user intention is a department that meets the specified conditions;

[0031] If the user intention is not included in the intention tag set, it is determined that the user intention is not a department that meets the specified conditions.

[0032] Furthermore, based on the preset referral model, department prediction processing is performed on the consultation content, the age, and the gender to obtain the corresponding predicted department list result:

[0033] Performing department prediction processing on the consultation content, the age, and the gender based on the referral model to obtain a corresponding initial department prediction result;

[0034] Get the preset department threshold;

[0035] Filtering out the designated department prediction results that are greater than the department threshold from all the initial department prediction results;

[0036] Call the preset sorting algorithm;

[0037] Sorting the prediction results of the designated department in descending order based on the sorting algorithm to obtain corresponding sorting results;

[0038] The sorting result is used as the predicted department list result.

[0039] Furthermore, after the step of determining whether the user's intention is to enter a department that meets the specified conditions, the method further includes:

[0040] If the user intends to search for a department that meets the specified conditions, entity recognition is performed on the inquiry content based on a preset named entity recognition model to obtain the corresponding department type and department entity;

[0041] Performing standardization on the department entity to obtain a corresponding target department entity;

[0042] Call the preset department support list;

[0043] Performing a matching query on the department support list based on the target department entity, and determining whether the target department entity is included in the department support list;

[0044] If so, call the designated department model corresponding to the department type to conduct a medical consultation process on the user.

[0045] Furthermore, after the step of determining whether the target department entity is included in the department support list, the method further includes:

[0046] If the target department entity is not included in the department support list, the processing of the consultation request is restricted;

[0047] Get preset reminder information;

[0048] Push the reminder information to the user.

[0049] In order to solve the above technical problems, the embodiment of the present application further provides an artificial intelligence-based request processing device, which adopts the following technical solution:

[0050] an extraction module, configured to extract the medical inquiry content from the medical inquiry request when receiving a medical inquiry request triggered by a user; wherein the medical inquiry request carries the medical inquiry content;

[0051] A first judgment module is used to perform intention recognition on the medical inquiry content to obtain the corresponding user intention, and to determine whether the user intention is a department that meets the specified conditions;

[0052] A first acquisition module is used to acquire the user's age and gender if the user's intention is not a department that meets the specified conditions;

[0053] A prediction module, configured to perform department prediction processing on the medical inquiry content, the age, and the gender based on a preset referral model, and obtain a corresponding predicted department list result;

[0054] The second judgment module is used to judge whether the current receiving department model and the predicted department list result meet the preset matching conditions;

[0055] A screening module, configured to screen out a target prediction department with the highest prediction value from the prediction department list results if the matching condition is not met;

[0056] The first processing module is used to call the target department model corresponding to the target predicted department, and perform a medical consultation process on the user based on the target department model.

[0057] In order to solve the above technical problems, the embodiment of the present application further provides a computer device, which adopts the following technical solution:

[0058] When receiving a user-triggered consultation request, extracting consultation content from the consultation request; wherein the consultation request carries the consultation content;

[0059] Performing intent recognition on the medical inquiry content to obtain the corresponding user intent, and determining whether the user intent is for a department that meets specified conditions;

[0060] If the user's intention is not a department that meets the specified conditions, then obtain the user's age and gender;

[0061] Performing department prediction processing on the medical inquiry content, the age, and the gender based on a preset referral model to obtain a corresponding predicted department list result;

[0062] Determine whether the current receiving department model and the predicted department list result meet the preset matching conditions;

[0063] If the matching condition is not met, the target prediction department with the highest prediction value is selected from the prediction department list results;

[0064] A target department model corresponding to the target predicted department is called, and a medical consultation process is performed on the user based on the target department model.

[0065] In order to solve the above technical problems, the embodiment of the present application further provides a computer-readable storage medium, which adopts the following technical solution:

[0066] When receiving a user-triggered consultation request, extracting consultation content from the consultation request; wherein the consultation request carries the consultation content;

[0067] Performing intent recognition on the medical inquiry content to obtain the corresponding user intent, and determining whether the user intent is for a department that meets specified conditions;

[0068] If the user's intention is not a department that meets the specified conditions, then obtain the user's age and gender;

[0069] Performing department prediction processing on the medical inquiry content, the age, and the gender based on a preset referral model to obtain a corresponding predicted department list result;

[0070] Determine whether the current receiving department model and the predicted department list result meet the preset matching conditions;

[0071] If the matching condition is not met, the target prediction department with the highest prediction value is selected from the prediction department list results;

[0072] A target department model corresponding to the target predicted department is called, and a medical consultation process is performed on the user based on the target department model.

[0073] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0074] When the present application receives a user-triggered consultation request, the consultation content is extracted from the consultation request; wherein, the consultation request carries the consultation content; then, the consultation content is subjected to intent recognition to obtain the corresponding user intent, and it is determined whether the user intent is a department that meets the specified conditions; if the user intent is not a department that meets the specified conditions, the user's age and gender are obtained; then, based on a preset referral model, department prediction processing is performed on the consultation content, the age and the gender to obtain a corresponding predicted department list result; subsequently, it is determined whether the current receiving department model and the predicted department list result meet the preset matching conditions; if they do not meet the matching conditions, the target predicted department with the highest prediction value is screened out from the predicted department list result; finally, the target department model corresponding to the target predicted department is called, and the user is subjected to consultation processing based on the target department model. When this application receives a user-triggered consultation request containing consultation content, it can intelligently and accurately select the best and most appropriate department model based on the user's actual personal situation for precise consultation by combining intention recognition, referral model and department model for matching and referral steps, thereby effectively improving the accuracy of the consultation. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] In order to more clearly illustrate the solutions in this application, a brief introduction will be given below to the drawings required for use in the description of the embodiments of this application. Obviously, the drawings described below are some embodiments of this application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0076] Figure 1 is an exemplary system architecture diagram to which the present application may be applied;

[0077] Figure 2 A flowchart of an embodiment of an artificial intelligence-based request processing method according to the present application;

[0078] Figure 3 is a structural diagram of an embodiment of an artificial intelligence-based request processing device according to the present application;

[0079] Figure 4It is a structural diagram of an embodiment of a computer device according to the present application. DETAILED DESCRIPTION

[0080] Unless otherwise defined, all technical and scientific terms used herein have the same meanings as commonly understood by those skilled in the art to which this application belongs. The terms used in the specification of the application are for the purpose of describing specific embodiments only and are not intended to limit this application. The terms "including" and "having" and any variations thereof in the specification and claims of this application and the above-mentioned drawings are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of this application or the above-mentioned drawings are used to distinguish different objects, not to describe a specific order.

[0081] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute an independent or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0082] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.

[0083] like Figure 1 As shown, system architecture 100 may include a terminal device 101, a network 102, and a server 103. Terminal device 101 may be a laptop computer 1011, a tablet computer 1012, or a mobile phone 1013. Network 102 is a medium for providing a communication link between terminal device 101 and server 103. Network 102 may include various connection types, such as wired or wireless communication links or fiber optic cables.

[0084] The user can use the terminal device 101 to interact with the server 103 via the network 102 to receive or send messages, etc. Various communication client applications can be installed on the terminal device 101, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0085] The terminal device 101 can be various electronic devices with a display screen and supporting web browsing. In addition to the laptop computer 1011, tablet computer 1012 or mobile phone 1013, the terminal device 101 can also be an e-book reader, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 (Moving Picture Experts Group Audio Layer IV) player, a laptop computer and a desktop computer, etc.

[0086] The server 103 may be a server that provides various services, such as a background server that provides support for web pages displayed on the terminal device 101 .

[0087] It should be noted that the artificial intelligence-based request processing method provided in the embodiment of the present application is generally executed by a server / terminal device, and accordingly, the artificial intelligence-based request processing device is generally set in the server / terminal device.

[0088] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is merely illustrative. Any number of terminal devices, networks and servers may be provided as required.

[0089] Continue to refer Figure 2 , shows a flowchart of an embodiment of the request processing method based on artificial intelligence according to the present application. According to different needs, the order of the steps in the flowchart can be changed, and some steps can be omitted. The request processing method based on artificial intelligence provided by the embodiment of the present application can be applied to any scenario that requires request processing, and the request processing method based on artificial intelligence can be applied to products in these scenarios, for example, request processing in the field of digital medicine. The request processing method based on artificial intelligence includes the following steps:

[0090] Step S201: When a user-triggered consultation request is received, the consultation content is extracted from the consultation request; wherein the consultation request carries the consultation content.

[0091] In this embodiment, the request processing method based on artificial intelligence is executed on the electronic device (e.g. Figure 1The server / terminal device shown in the figure) can obtain the consultation content through a wired connection or a wireless connection. It should be pointed out that the above-mentioned wireless connection method may include but is not limited to 3G / 4G / 5G connection, WiFi connection, Bluetooth connection, WiMAX connection, Zigbee connection, UWB (ultra wideband) connection, and other wireless connection methods currently known or to be developed in the future. The executive subject of this application is specifically a consultation system, or a consultation platform, and can be referred to as a system for short. Specifically, a user-friendly interface is provided in advance, and users can enter their consultation content and trigger corresponding consultation requests. Among them, the input of the consultation content can be a text input box, voice input or other forms. In addition, after receiving the consultation request, the system can parse the information of the consultation request to obtain the corresponding consultation content.

[0092] Step S202: perform intent recognition on the medical inquiry content to obtain the corresponding user intent, and determine whether the user intent is a department that meets the specified conditions.

[0093] In this embodiment, the above-mentioned specific implementation process of identifying the intention of the medical inquiry content to obtain the corresponding user intention and determining whether the user intention is a department that meets the specified conditions will be described in further detail in subsequent specific embodiments of the present application and will not be elaborated on here.

[0094] Step S203: If the user's intention is not a department that meets the specified conditions, the user's age and gender are obtained.

[0095] In this embodiment, the medical consultation request may also include the user's age and gender. The medical consultation request may be parsed to obtain the desired user's age and gender. Alternatively, the user's age and gender may be retrieved by searching for the user's name.

[0096] Step S204: performing department prediction processing on the medical inquiry content, the age, and the gender based on a preset referral model to obtain a corresponding predicted department list result.

[0097] In this embodiment, the above-mentioned specific implementation process of performing department prediction processing on the consultation content, the age and the gender based on the preset referral model to obtain the corresponding predicted department list results will be further described in detail in the subsequent specific embodiments of this application and will not be elaborated on here.

[0098] Step S205: determine whether the current receiving department model and the predicted department list result meet the preset matching conditions.

[0099] In this embodiment, the current reception department model for the user's consultation can be determined, and the consultation scope of the reception department model can be obtained, and then it can be judged whether the consultation scope of the reception department model includes any corresponding predicted department in the above-mentioned predicted department list results. If it does, it is determined that the current reception department model and the predicted department list results meet the matching conditions; otherwise, it is determined that the current reception department model and the predicted department list results do not meet the matching conditions.

[0100] Among them, if the department list predicted by the current receiving department model meets the preset matching conditions, it means that the consultation scope of the current receiving department model includes the predicted department, and the receiving department model will continue to be used for consultation and no referral will be made.

[0101] Step S206: If the matching condition is not met, the target prediction department with the highest prediction value is selected from the prediction department list results.

[0102] In this embodiment, the target prediction department refers to the first-ranked prediction department obtained after sorting the prediction department list results in descending order. The target prediction department is also the prediction department with the highest prediction score among all prediction department list results.

[0103] Step S207: calling a target department model corresponding to the target predicted department, and performing a medical consultation process on the user based on the target department model.

[0104] In this embodiment, the target department model corresponding to the above-mentioned target prediction department can be called, and then the user information, consultation content and referral department information can be passed to the target department model, thereby accurately achieving automatic referral for the user.

[0105] When the present application receives a user-triggered consultation request, the consultation content is extracted from the consultation request; wherein, the consultation request carries the consultation content; then, the consultation content is subjected to intent recognition to obtain the corresponding user intent, and it is determined whether the user intent is a department that meets the specified conditions; if the user intent is not a department that meets the specified conditions, the user's age and gender are obtained; then, based on a preset referral model, department prediction processing is performed on the consultation content, the age and the gender to obtain a corresponding predicted department list result; subsequently, it is determined whether the current receiving department model and the predicted department list result meet the preset matching conditions; if they do not meet the matching conditions, the target predicted department with the highest prediction value is screened out from the predicted department list result; finally, the target department model corresponding to the target predicted department is called, and the user is subjected to consultation processing based on the target department model. When this application receives a user-triggered consultation request containing consultation content, it can intelligently and accurately select the best and most appropriate department model based on the user's actual personal situation for precise consultation by combining intention recognition, referral model and department model for matching and referral steps, thereby effectively improving the accuracy of the consultation.

[0106] In some optional implementations, before step S204, the electronic device may further perform the following steps:

[0107] Obtain pre-collected medical consultation text data.

[0108] In this embodiment, a large amount of medical consultation data in text form, namely the above-mentioned medical consultation text data, is collected in advance. The medical consultation text data includes the patient's age, gender, medical consultation content, and corresponding department label.

[0109] The medical inquiry text data is annotated to obtain corresponding initial sample data.

[0110] In this embodiment, the above-mentioned medical consultation text data can be annotated using a machine annotation method based on the collected department labels, thereby obtaining corresponding initial sample data.

[0111] The initial sample data is preprocessed to obtain corresponding target sample data.

[0112] In this embodiment, the corresponding target sample data is obtained by preprocessing the initial sample data, such as word segmentation, removal of stop words, construction of a vocabulary, and conversion of the text into an input format acceptable to the pre-trained language model.

[0113] A pre-trained language model is called, and a model structure optimization process is performed on the pre-trained language model to obtain a corresponding first specified model.

[0114] In this embodiment, the above-mentioned pre-trained language model can specifically adopt a pre-trained BERT model, such as BERT-base or BERT-large. Among them, a model framework based on BERT+ multi-head output can be adopted. The purpose of such a design is to achieve one-time prediction and output multiple results. If the consultation platform supports a total of N independent consultation departments, N binary classification prediction heads are added after the BERT model when training the referral model. Each binary classification prediction head independently predicts whether the input belongs to the category. Through N independent binary classification modules, the use of multi-head output can more accurately judge the current consultation content and the scope of the consultation department.

[0115] Based on a preset construction strategy, the first designated model is trained and evaluated using the target sample data to obtain a second designated model that meets the preset construction requirements.

[0116] In this embodiment, the above-mentioned process of training and evaluating the first specified model using the target sample data includes: first, using the binary cross-entropy loss function as the loss function of each prediction head. Since there are multiple prediction heads, the total loss is the average or weighted sum of the losses of all prediction heads. Then, an optimizer, such as Adam, SGD, etc., is selected to update the weights of the model. Thereafter, the target sample data is divided into training data, test data, and validation data, and the first specified model is iterated multiple times on the training data. In each iteration, the data is randomly shuffled and divided into small batches. For each small batch, forward propagation calculates the prediction results and loss, backpropagation calculates the gradient, and the optimizer is used to update the model weights. Subsequently, the test data is used to evaluate the performance of the model, and indicators such as accuracy, recall rate, F1 score, etc. are calculated for each department. And the overall performance, such as the macro-average or micro-average F1 score, is calculated. Further, based on the evaluation results, the model parameters, such as the learning rate, batch size, number of training rounds, etc., are adjusted. For example, different model architectures or preprocessing steps can be tried to further improve performance; regularization techniques (such as L2 regularization and dropout) can be used to prevent overfitting; and hyperparameter tuning can be performed to find the optimal model configuration. Finally, the optimized model performance is verified on validation data to ensure that the model performs well on unseen data. The trained model weights and configuration information are then saved. The trained model (referral model) is then deployed to a production environment for department prediction based on actual consultation data.

[0117] The second designated model is used as the referral model.

[0118] This application obtains pre-collected medical inquiry text data; then annotates the medical inquiry text data to obtain corresponding initial sample data; then pre-processes the initial sample data to obtain corresponding target sample data; subsequently calls a pre-trained language model, and optimizes the model structure of the pre-trained language model to obtain a corresponding first designated model; further based on a preset construction strategy, uses the target sample data to train and evaluate the first designated model to obtain a second designated model that meets the preset construction requirements; finally, uses the second designated model as a referral model. This application constructs target sample data based on pre-collected medical inquiry text data, then optimizes the model structure of the pre-trained language model to obtain a first designated model, and then based on a construction strategy, uses the target sample data to train and evaluate the first designated model, thereby achieving rapid and accurate construction of a referral model that meets the construction requirements, improving the construction efficiency of the referral model, and ensuring the construction effect of the referral model obtained.

[0119] In some optional implementations of this embodiment, performing model structure optimization processing on the pre-trained language model to obtain the corresponding first specified model includes the following steps:

[0120] Get all preset departments.

[0121] In this embodiment, the above-mentioned departments refer to the departments pre-set in the outpatient platform.

[0122] Construct a binary classification prediction head corresponding to each of the departments.

[0123] In this embodiment, a binary classification prediction head is initialized for each department. Specifically, the weight matrix and bias vector are randomly initialized. Each binary classification prediction head independently predicts whether the input belongs to that category.

[0124] The output of the pre-trained language model is connected to the input of each of the two-class prediction heads to obtain a corresponding first language model.

[0125] In this embodiment, the corresponding first language model can be obtained by connecting the output of the pre-trained language model after transformation by the fully connected layer to the input of each binary classification prediction head.

[0126] A preset target activation function is applied to the output of each of the two-class prediction heads to obtain a corresponding second language model.

[0127] In this embodiment, the target activation function may specifically be a Sigmoid activation function. The Sigmoid activation function may be applied to the output of each binary prediction head in the first language model to convert the output into a probability value between 0 and 1, thereby obtaining the corresponding second language model.

[0128] The second language model is used as the first designated model.

[0129] This application obtains all preset departments; then constructs a binary classification prediction head corresponding to each of the departments; then connects the output of the pre-trained language model to the input of each of the binary classification prediction heads to obtain the corresponding first language model; subsequently applies a preset target activation function to the output of each of the binary classification prediction heads to obtain the corresponding second language model; finally, uses the second language model as the first designated model. This application constructs a binary classification prediction head corresponding to each of the departments; then connects the output of the pre-trained language model to the input of each of the binary classification prediction heads to obtain the corresponding first language model; and then applies a preset target activation function to the output of each of the binary classification prediction heads, thereby achieving rapid and accurate completion of the model structure optimization processing of the pre-trained language model and obtaining a first designated model with multi-head output function, thereby improving the prediction effect of the obtained first designated model.

[0130] In some optional implementations, step S202 includes the following steps:

[0131] Call the preset intent recognition model.

[0132] In this example, a large amount of text data from patient consultations is collected in advance and labeled with intent (e.g., specifying a department, describing symptoms, etc.). Machine learning or deep learning algorithms (e.g., support vector machines, naive Bayesian, LSTM, etc.) are then used to train an intent recognition model capable of identifying intent. The trained intent recognition model can recognize multiple intents, including but not limited to specifying a department, describing symptoms, and asking for treatment recommendations.

[0133] The intention of the medical inquiry is recognized based on the intention recognition model to obtain a corresponding intention recognition label, and the intention recognition label is used as the user intention.

[0134] In this embodiment, by inputting the medical inquiry content into the intention recognition model, the intention recognition model will output a predicted intention label, that is, obtain the user intention of the user.

[0135] Gets a set of predefined intent labels.

[0136] In this embodiment, the above-mentioned intention tag set may be a set of tags corresponding to designated departments that is pre-built according to actual business needs.

[0137] If the user intention is included in the intention tag set, it is determined that the user intention is a department that meets the specified conditions.

[0138] In this embodiment, the predicted user intention is usually compared with a predefined intention label set. If the user intention is included in the intention label set, it is determined that the user intention is a department that meets the specified conditions.

[0139] If the user intention is not included in the intention tag set, it is determined that the user intention is not a department that meets the specified conditions.

[0140] In this embodiment, the predicted user intention is usually compared with a predefined intention label set. If the user intention is not included in the intention label set, it is determined that the user intention is not for a department that meets the specified conditions.

[0141] This application calls a preset intention recognition model; then, based on the intention recognition model, the medical inquiry content is subjected to intention recognition to obtain a corresponding intention recognition label, and the intention recognition label is used as the user intention; subsequently, a predefined intention label set is obtained; if the user intention is included in the intention label set, the user intention is determined to be a department that meets the specified conditions; and if the user intention is not included in the intention label set, the user intention is determined not to be a department that meets the specified conditions. This application can intelligently and accurately perform intention recognition on medical inquiry content based on the use of an intention recognition model to obtain the corresponding user intention, and then, by comparing the user intention with the predefined intention label set, it can quickly and accurately detect whether the user intention is a department that meets the specified conditions based on the comparison result, thereby improving the processing efficiency of user intention detection and the accuracy of the generated detection results.

[0142] In some optional implementations, step S204 includes the following steps:

[0143] Based on the referral model, department prediction processing is performed on the medical inquiry content, the age and the gender to obtain a corresponding initial department prediction result.

[0144] In this embodiment, the referral model is a trained model that can predict possible departments based on the patient's age, gender, and consultation content. The referral model is used to perform department prediction on the consultation content, age, and gender, thereby outputting multiple predicted departments, i.e., the initial department prediction results.

[0145] Get the preset department threshold.

[0146] In this embodiment, there is no specific limitation on the value of the above-mentioned department threshold, which can be set according to actual business needs.

[0147] Filter out the designated department prediction results that are greater than the department threshold from all the initial department prediction results.

[0148] In this embodiment, the initial department prediction results can be compared with the department thresholds respectively, so as to screen out the designated department prediction results that are greater than the department thresholds.

[0149] Call the preset sorting algorithm.

[0150] In this embodiment, there is no specific limitation on the selection of the above sorting algorithm. For example, a comparison sorting algorithm such as bubble sort, selection sort, etc. can be used, or a more efficient sorting algorithm such as merge sort, quick sort, etc. can also be used.

[0151] The prediction results of the designated department are sorted in descending order based on the sorting algorithm to obtain corresponding sorting results.

[0152] In this embodiment, the prediction results of the designated departments can be sorted in descending order according to the sorting processing steps of the above-mentioned sorting algorithm to obtain corresponding sorting results.

[0153] The sorting result is used as the predicted department list result.

[0154] This application performs department prediction processing on the medical inquiry content, the age and the gender based on the referral model to obtain the corresponding initial department prediction result; then obtains a preset department threshold; then filters out the designated department prediction results that are greater than the department threshold from all the initial department prediction results; subsequently calls a preset sorting algorithm; and sorts the designated department prediction results in descending order based on the sorting algorithm to obtain the corresponding sorting result; finally uses the sorting result as the predicted department list result. This application performs department prediction processing on the medical inquiry content, age and gender based on the use of a referral model to obtain the initial department prediction result, then filters out the designated department prediction results that are greater than the preset department threshold from all the obtained initial department prediction results, and then sorts the designated department prediction results in descending order based on the use of the sorting algorithm, thereby achieving the rapid and accurate generation of the required predicted department list result, improving the accuracy of the generated predicted department list result.

[0155] In some optional implementations of this embodiment, after step S202, the electronic device may further perform the following steps:

[0156] If the user intends to search for a department that meets the specified conditions, entity recognition is performed on the consultation content based on a preset named entity recognition model to obtain the corresponding department type and department entity.

[0157] In this embodiment, the named entity recognition model can be a NER model. It can be trained using a dataset of labeled department names to obtain a NER model that accurately identifies department types and department entities. The trained NER model is then deployed in the medical consultation system to process input text in real time.

[0158] The department entity is standardized to obtain a corresponding target department entity.

[0159] In this embodiment, standardized rules for processing department entities are pre-established. These rules include standardized processing of department name spellings, abbreviations, synonyms, and other aspects. By invoking these processing rules, the aforementioned department entities can be standardized to obtain the corresponding target department entity. Furthermore, the standardized rules can be regularly updated based on the actual conditions of the hospital or platform to ensure their accuracy and applicability.

[0160] Call the preset department support list.

[0161] In this embodiment, a department support list is maintained within the system, and the department support list includes the names of all departments that the platform supports for appointments or consultations.

[0162] A matching query is performed on the department support list based on the target department entity to determine whether the target department entity is included in the department support list.

[0163] In this embodiment, a matching query is performed between the target department entity and the department support list to determine whether the target department entity exists in the department support list, that is, to detect whether the target department entity is supported in the department support list.

[0164] If so, call the designated department model corresponding to the department type to conduct a medical consultation process on the user.

[0165] In this embodiment, if the result of the department support query is supported, that is, the target department entity is included in the department support list, the designated department model corresponding to the above department type is automatically called to process the user's consultation. Among them, for each supported department, a corresponding department model is pre-developed. These department models can be used to process the consultation content related to the department based on technologies such as deep learning and machine learning. The developed department models are subsequently deployed to the system, and it is ensured that they can be seamlessly integrated with the NER model.

[0166] If this application detects that the user's intention is a department that meets the specified conditions, the inquiry content is subject to entity recognition based on the preset named entity recognition model to obtain the corresponding department type and department entity; the department entity is then standardized to obtain the corresponding target department entity; the preset department support list is then called; subsequently, a matching query is performed on the department support list based on the target department entity to determine whether the target department entity is included in the department support list; if so, the specified department model corresponding to the department type is called to perform inquiry processing on the user. When this application detects that the user's intention is to go to a department that meets the specified conditions, it will perform entity recognition on the inquiry content based on the use of the named entity recognition model to obtain the department type and department entity, and standardize the department entity to obtain the target department entity. It will then intelligently perform a matching query on the preset department support list based on the obtained target department entity. If it is detected that the target department entity is included in the department support list, the specified department model corresponding to the department type will be automatically called to process the user's inquiry, thereby improving the accuracy and adaptability of the inquiry processing, which is conducive to improving the user experience.

[0167] In some optional implementations of this embodiment, after the step of determining whether the target department entity is included in the department support list, the electronic device may further perform the following steps:

[0168] If the target department entity is not included in the department support list, processing of the medical consultation request is restricted.

[0169] In this embodiment, if it is detected that the target department entity is not included in the above-mentioned department support list, it indicates that the consultation platform does not support the department corresponding to the target department entity, and then the system will perform a jump process (jump out of the automatic consultation stage), that is, limit the subsequent processing of the above-mentioned consultation request.

[0170] Get the preset reminder information.

[0171] In this embodiment, the system will automatically generate a corresponding pop-up reason corresponding to the situation where the platform does not support the department, that is, the above-mentioned reminder message. For example, the content of the reminder message may include: "Dear user, the department you specified is temporarily unsupported by the platform."

[0172] Push the reminder information to the user.

[0173] In this embodiment, the above-mentioned reminder information can be displayed to the user. This can be a simple message prompt or a dialog box containing more detailed information. In addition, while the feedback result is being provided, the user can be guided to select other supported departments or perform other operations, such as re-entering the consultation content, selecting a symptom description, etc.

[0174] If this application detects that the target department entity is not included in the department support list, it will limit the processing of the consultation request; then obtain the preset reminder information; and subsequently push the reminder information to the user. When this application detects that the target department entity is not included in the department support list, it will automatically limit the processing of the consultation request, thereby improving the intelligence of the processing of the consultation request and reducing the waste of useless work during the processing of the consultation request. In addition, the obtained reminder information will be intelligently pushed to the user so that the user can clearly understand the current reason for the jump, which is conducive to improving the user experience.

[0175] In some optional implementations, the user information obtained is obtained with the user's consent and complies with relevant laws and policies.

[0176] In addition, any software tools or components not provided by our company that appear in the embodiments of this application are merely examples and do not represent actual use.

[0177] Furthermore, existing medical consultation platforms often use fixed scripts to assign departments after consultations or use general practice models, failing to automatically implement department referrals based on the patient's specific circumstances. The proposed fully automated medical referral system (medical consultation system) based on medical consultations uses intent recognition to determine designated and predicted departments. For designated departments, named entity recognition is used to extract department entities, determine whether the medical consultation platform supports them, and select the final consultation method. To address the diverse input from different patients within the same department, this case standardizes department entities to ensure that the designated department output meets the platform's specifications. Regarding the predicted department, this case innovatively employs a multi-head prediction output framework, using multiple independent binary classification modules to determine department assignments, improving the accuracy of predicting that the same consultation belongs to multiple departments. When the predicted department's scope of inquiry does not overlap with the current department's scope of inquiry, the system automatically makes a referral based on the prediction results, invoking the corresponding consultation model for the predicted department. Automatic and accurate referrals based on patient conversations can provide patients with more accurate medical consultation services and improve user satisfaction.

[0178] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0179] It should be emphasized that in order to further ensure the privacy and security of the above referral model, the above referral model can also be stored in a node of a blockchain.

[0180] The blockchain referred to in this application is a new application model for computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanisms, and encryption algorithms. Blockchain is essentially a decentralized database, a series of data blocks generated using cryptographic methods. Each data block contains information about a batch of network transactions, which is used to verify the validity of this information (to prevent counterfeiting) and generate the next block. Blockchain can include the underlying blockchain platform, the platform product service layer, and the application service layer.

[0181] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Artificial Intelligence (AI) is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use knowledge to achieve optimal results.

[0182] Fundamental AI technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interaction systems, and mechatronics. AI software technologies primarily encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0183] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing related hardware via computer-readable instructions. The computer-readable instructions can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes in the above-described method embodiments. The aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0184] It should be understood that although the steps in the flowcharts of the accompanying drawings are shown in sequence as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some of the steps in the flowcharts of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.

[0185] Further references Figure 3 , as a response to the above Figure 2 The present application provides an embodiment of a request processing device based on artificial intelligence, which is similar to Figure 2 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0186] like Figure 3 As shown, the artificial intelligence-based request processing device 300 described in this embodiment includes: an extraction module 301, a first judgment module 302, a first acquisition module 303, a prediction module 304, a second judgment module 305, a screening module 306 and a first processing module 307.

[0187] in:

[0188] The extraction module 301 is configured to extract the medical inquiry content from the medical inquiry request when receiving the medical inquiry request triggered by the user; wherein the medical inquiry request carries the medical inquiry content;

[0189] The first judgment module 302 is used to perform intention recognition on the medical inquiry content to obtain the corresponding user intention, and to determine whether the user intention is a department that meets the specified conditions;

[0190] The first acquisition module 303 is configured to acquire the user's age and gender if the user's intention is not a department that meets the specified conditions;

[0191] The prediction module 304 is configured to perform department prediction processing on the medical inquiry content, the age, and the gender based on a preset referral model to obtain a corresponding predicted department list result;

[0192] The second judgment module 305 is used to judge whether the current receiving department model and the predicted department list result meet the preset matching conditions;

[0193] A screening module 306 is configured to screen out a target prediction department with the highest prediction value from the prediction department list result if the matching condition is not met;

[0194] The first processing module 307 is used to call the target department model corresponding to the target predicted department, and perform a medical consultation process on the user based on the target department model.

[0195] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based request processing method in the aforementioned embodiment, and are not repeated here.

[0196] In some optional implementations of this embodiment, the artificial intelligence-based request processing device further includes:

[0197] The second acquisition module is used to obtain pre-collected medical consultation text data;

[0198] A labeling module is used to label the medical consultation text data to obtain corresponding initial sample data;

[0199] A preprocessing module, configured to preprocess the initial sample data to obtain corresponding target sample data;

[0200] An optimization module, configured to call a pre-trained language model and perform model structure optimization processing on the pre-trained language model to obtain a corresponding first specified model;

[0201] A second processing module is configured to train and evaluate the first designated model using the target sample data based on a preset construction strategy to obtain a second designated model that meets the preset construction requirements;

[0202] A determination module is used to use the second designated model as a referral model.

[0203] In some optional implementations of this embodiment, the optimization module includes:

[0204] The first acquisition submodule is used to acquire all preset departments;

[0205] A construction submodule is used to construct a binary classification prediction head corresponding to each of the departments;

[0206] A first processing submodule, configured to connect the output of the pre-trained language model to the input of each of the binary classification prediction heads to obtain a corresponding first language model;

[0207] A second processing submodule is configured to apply a preset target activation function to the output of each of the binary classification prediction heads to obtain a corresponding second language model;

[0208] The first determining submodule is configured to use the second language model as the first designated model.

[0209] In some optional implementations of this embodiment, the first determining module 302 includes:

[0210] The first calling submodule is used to call a preset intent recognition model;

[0211] an identification submodule, configured to perform intent recognition on the medical inquiry content based on the intent recognition model to obtain a corresponding intent recognition label, and use the intent recognition label as the user intention;

[0212] The second acquisition submodule is used to obtain a predefined intent tag set;

[0213] A first determination submodule, configured to determine that the user intention is a department that meets specified conditions if the user intention is included in the intention tag set;

[0214] The second determination submodule is configured to determine that the user intention is not a department that meets the specified conditions if the user intention is not included in the intention tag set.

[0215] In some optional implementations of this embodiment, the prediction module 304 includes:

[0216] A prediction submodule, configured to perform department prediction processing on the consultation content, the age, and the gender based on the referral model to obtain a corresponding initial department prediction result;

[0217] The third acquisition submodule is used to obtain the preset department threshold;

[0218] A screening submodule, configured to screen out, from all the initial department prediction results, a designated department prediction result that is greater than the department threshold;

[0219] The second calling submodule is used to call a preset sorting algorithm;

[0220] A sorting submodule, configured to sort the prediction results of the designated department in descending order based on the sorting algorithm to obtain corresponding sorting results;

[0221] The second determining submodule is configured to use the ranking result as the predicted department list result.

[0222] In some optional implementations of this embodiment, the artificial intelligence-based request processing device further includes:

[0223] An identification module is used to perform entity recognition on the consultation content based on a preset named entity recognition model to obtain the corresponding department type and department entity if the user intends to search for a department that meets the specified conditions;

[0224] A standardization module, used to perform standardization processing on the department entity to obtain a corresponding target department entity;

[0225] Calling module, used to call the preset department support list;

[0226] a matching module, configured to perform a matching query on the department support list based on the target department entity, and determine whether the target department entity is included in the department support list;

[0227] The third processing module is configured to, if yes, call a designated department model corresponding to the department type to perform a medical consultation process on the user.

[0228] In some optional implementations of this embodiment, the artificial intelligence-based request processing device further includes:

[0229] a response module, configured to restrict processing of the medical consultation request if the target department entity is not included in the department support list;

[0230] The third acquisition module is used to obtain preset reminder information;

[0231] A push module is used to push the reminder information to the user.

[0232] To solve the above technical problems, the present application also provides a computer device. Figure 4 , Figure 4 This is a basic structural block diagram of the computer device in this embodiment.

[0233] The computer device 4 includes a memory 41, a processor 42, and a network interface 43 that are interconnected through a system bus. It should be noted that the figure only shows a computer device 4 with components 41-43, but it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented instead. Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application-specific integrated circuits (ASICs), programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0234] The computer device may be a desktop computer, notebook computer, PDA, cloud server, etc. The computer device may interact with the user via a keyboard, mouse, remote control, touchpad, or voice control device.

[0235] The memory 41 includes at least one type of readable storage medium, including flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 41 can be an internal storage unit of the computer device 4, such as the hard disk or memory of the computer device 4. In other embodiments, the memory 41 can also be an external storage device of the computer device 4, such as a plug-in hard disk equipped on the computer device 4, a smart memory card (SMC), a secure digital (SD) card, a flash memory card, etc. Of course, the memory 41 can also include both the internal storage unit of the computer device 4 and its external storage device. In this embodiment, the memory 41 is generally used to store the operating system and various application software installed on the computer device 4, such as computer-readable instructions of the request processing method based on artificial intelligence. In addition, the memory 41 can also be used to temporarily store various types of data that have been output or are to be output.

[0236] In some embodiments, the processor 42 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip. The processor 42 is generally used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is used to execute computer-readable instructions stored in the memory 41 or process data, such as computer-readable instructions for executing the artificial intelligence-based request processing method.

[0237] The network interface 43 may include a wireless network interface or a wired network interface. The network interface 43 is generally used to establish a communication connection between the computer device 4 and other electronic devices.

[0238] The present application also provides another embodiment, namely, providing a computer-readable storage medium, which stores computer-readable instructions, and the computer-readable instructions can be executed by at least one processor to enable the at least one processor to perform the steps of the above-mentioned artificial intelligence-based request processing method.

[0239] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0240] Obviously, the embodiments described above are only some of the embodiments of the present application, rather than all of the embodiments. The preferred embodiments of the present application are given in the accompanying drawings, but they do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present application more thorough and comprehensive. Although the present application has been described in detail with reference to the aforementioned embodiments, for those skilled in the art, it is still possible to modify the technical solutions described in the aforementioned specific embodiments, or to make equivalent replacements for some of the technical features therein. Any equivalent structure made using the contents of the present application specification and the accompanying drawings, directly or indirectly used in other related technical fields, is also within the scope of patent protection of the present application.

Claims

1. A request processing method based on artificial intelligence, characterized in that: The steps include: When receiving a user-triggered consultation request, extracting consultation content from the consultation request; wherein the consultation request carries the consultation content; Performing intent recognition on the medical inquiry content to obtain the corresponding user intent, and determining whether the user intent is for a department that meets specified conditions; If the user's intention is not a department that meets the specified conditions, then obtain the user's age and gender; Performing department prediction processing on the medical inquiry content, the age, and the gender based on a preset referral model to obtain a corresponding predicted department list result; Determine whether the current receiving department model and the predicted department list result meet the preset matching conditions; If the matching condition is not met, the target prediction department with the highest prediction value is selected from the prediction department list results; A target department model corresponding to the target predicted department is called, and a medical consultation process is performed on the user based on the target department model.

2. The request processing method based on artificial intelligence according to claim 1, characterized in that: Before the step of performing department prediction processing on the medical inquiry content, the age, and the gender based on the preset referral model to obtain a corresponding predicted department list result, the method further includes: Obtain pre-collected medical consultation text data; Annotating the medical consultation text data to obtain corresponding initial sample data; Preprocessing the initial sample data to obtain corresponding target sample data; Calling a pre-trained language model, and performing model structure optimization processing on the pre-trained language model to obtain a corresponding first specified model; Based on a preset construction strategy, the first specified model is trained and evaluated using the target sample data to obtain a second specified model that meets the preset construction requirements; The second designated model is used as the referral model.

3. The request processing method based on artificial intelligence according to claim 2, characterized in that: The step of performing model structure optimization processing on the pre-trained language model to obtain the corresponding first specified model specifically includes: Get all preset departments; Constructing a binary classification prediction head corresponding to each of the departments; Connecting the output of the pre-trained language model to the input of each of the two-class prediction heads to obtain a corresponding first language model; Applying a preset target activation function to the output of each of the two-class prediction heads to obtain a corresponding second language model; The second language model is used as the first designated model.

4. The request processing method based on artificial intelligence according to claim 1, characterized in that: The step of performing intent recognition on the medical inquiry content to obtain the corresponding user intent, and determining whether the user intent is a department that meets the specified conditions, specifically includes: Call the preset intent recognition model; Performing intent recognition on the medical inquiry content based on the intent recognition model to obtain a corresponding intent recognition label, and using the intent recognition label as the user intention; Get a predefined set of intent tags; If the user intention is included in the intention tag set, it is determined that the user intention is a department that meets the specified conditions; If the user intention is not included in the intention tag set, it is determined that the user intention is not a department that meets the specified conditions.

5. The request processing method based on artificial intelligence according to claim 1, characterized in that: Based on the preset referral model, the department prediction processing is performed on the consultation content, the age and the gender to obtain the corresponding predicted department list result: Performing department prediction processing on the consultation content, the age, and the gender based on the referral model to obtain a corresponding initial department prediction result; Get the preset department threshold; Filtering out the designated department prediction results that are greater than the department threshold from all the initial department prediction results; Call the preset sorting algorithm; Sorting the prediction results of the designated department in descending order based on the sorting algorithm to obtain corresponding sorting results; The sorting result is used as the predicted department list result.

6. The request processing method based on artificial intelligence according to claim 1, characterized in that: After the step of determining whether the user's intention is to enter a department that meets the specified conditions, the method further includes: If the user intends to search for a department that meets the specified conditions, entity recognition is performed on the inquiry content based on a preset named entity recognition model to obtain the corresponding department type and department entity; Performing standardization on the department entity to obtain a corresponding target department entity; Call the preset department support list; Performing a matching query on the department support list based on the target department entity, and determining whether the target department entity is included in the department support list; If so, call the designated department model corresponding to the department type to conduct a medical consultation process on the user.

7. The request processing method based on artificial intelligence according to claim 6, characterized in that: After the step of determining whether the target department entity is included in the department support list, the method further includes: If the target department entity is not included in the department support list, the processing of the consultation request is restricted; Get preset reminder information; Push the reminder information to the user.

8. A request processing device based on artificial intelligence, characterized in that: include: an extraction module, configured to extract the medical inquiry content from the medical inquiry request when receiving a medical inquiry request triggered by a user; wherein the medical inquiry request carries the medical inquiry content; A first judgment module is used to perform intention recognition on the medical inquiry content to obtain the corresponding user intention, and to determine whether the user intention is a department that meets the specified conditions; A first acquisition module is used to acquire the user's age and gender if the user's intention is not a department that meets the specified conditions; A prediction module, configured to perform department prediction processing on the medical inquiry content, the age, and the gender based on a preset referral model, and obtain a corresponding predicted department list result; The second judgment module is used to judge whether the current receiving department model and the predicted department list result meet the preset matching conditions; A screening module, configured to screen out a target prediction department with the highest prediction value from the prediction department list results if the matching condition is not met; The first processing module is used to call the target department model corresponding to the target predicted department, and perform a medical consultation process on the user based on the target department model.

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

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the artificial intelligence-based request processing method according to any one of claims 1 to 7.

Images