Personalized intelligent hospital guide method and system, storage medium and program product

Abstract

The invention discloses a personalized intelligent hospital guide method and system, a storage medium and a program product, and the method comprises the steps: calculating an operation proficiency degree according to historical operation behavior data; dividing the hospital guide process into a plurality of hospital guide sub-processes based on the operation proficiency; sequentially executing each hospital guide sub-process, and acquiring input information of the user in the current hospital guide sub-process; storing the corresponding relationship between the input information and the current hospital guide sub-process into a hospital guide information pool; when it is detected that the user exits the current hospital guide sub-process, a hospital guide progress identifier is generated; in response to detecting that the user resends a hospital guide request, determining an uncompleted hospital guide sub-process according to the hospital guide progress identifier, and calling input information stored in the hospital guide information pool in the uncompleted hospital guide sub-process; and generating a hospital guide result according to the input information of all the hospital guide sub-processes stored in the hospital guide information pool. The method is used for improving the accuracy and efficiency of hospital guidance in outpatient service.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of Internet guidance, and particularly relates to an individualized intelligent guidance method and system, a storage medium and a program product. BACKGROUND

[0002] With the popularization of Internet medical services, intelligent guidance systems are widely used in outpatient services in large general hospitals. Due to the complex department settings in large general hospitals and the business overlap between different departments, ordinary users have difficulty in accurately selecting a medical department, and traditional manual triage often causes queuing congestion and waste of human resources. However, the traditional intelligent guidance system uses a fixed standardized diversion template, which is difficult to provide personalized guidance services according to the specific department settings of the hospital and the needs of users, resulting in low triage efficiency.

[0003] In related technologies, an intelligent guidance method based on knowledge graph and transfer learning can be used. This method constructs a hospital department knowledge graph, correlates and models multi-dimensional features such as department information and doctor expertise, and uses transfer learning technology to realize rapid adaptation of the model in different scenarios. This method can dynamically update the department information in the knowledge graph, and realize rapid transfer learning of the model with a small amount of data, improving the system's adaptability to new departments.

[0004] However, there are differences in the use of intelligent guidance systems among users of different ages. In particular, the elderly are often not proficient in operating smart devices, and often repeat input or quit and start over during the guidance process, resulting in the need for repeated input by the elderly to complete a guidance, and a portion of the elderly will give up and choose manual triage during the guidance process, reducing the accuracy and efficiency of guidance in outpatient services. SUMMARY

[0005] The application provides an individualized intelligent guidance method, system, storage medium and program product for improving the accuracy and efficiency of guidance in outpatient services.

[0006] In a first aspect, the application provides an individualized intelligent guidance method, which, in the case where there is historical operation behavior data corresponding to the user according to the guidance request sent by the user, calculates the operation proficiency according to the historical operation behavior data, and the operation proficiency is calculated according to the operation step completion time and the number of operation errors in the historical operation behavior data through a pre-designed calculation function; dividing the guidance process into a plurality of guidance sub-processes based on the operation proficiency; executing each guidance sub-process in turn and obtaining input information of the user in the current guidance sub-process; storing the corresponding relationship between the input information and the current guidance sub-process in a guidance information pool. In the case of detecting that the user exits the current triage sub-process, a triage progress identifier is generated; In response to detecting that the user resends the triage request, the uncompleted triage sub-process is determined according to the triage progress identifier, and the input information stored in the triage information pool is called in the uncompleted triage sub-process; The triage result is generated according to the input information of all triage sub-processes stored in the triage information pool.

[0007] By adopting the above technical solutions, the system evaluates the actual operation ability of the user according to the operation step completion time and the operation error times in the historical operation behavior data, avoiding the use experience difference caused by the same process for different users in the traditional fixed triage process. By dividing the triage process into multiple sub-processes and executing them one by one, the cognitive burden of the user in the triage process is reduced. The system stores the input information of the user in each triage sub-process into the triage information pool and generates a triage progress identifier, so that the user can directly continue the uncompleted triage sub-process when re-entering in the middle of the way, and automatically call the stored input information, reducing the number of times of repeated input information of the user, improving the continuity and efficiency of the triage process, and improving the use experience of the user.

[0008] In combination with some embodiments of the first aspect, in some embodiments, the operation proficiency is calculated according to the historical operation behavior data, specifically comprising: Operation time sequence features corresponding to the operation step completion time are extracted from the historical operation behavior data, and the operation time sequence features include time intervals between consecutive operations and a duration of a single operation; A time dimension score is constructed based on the operation time sequence features, and the time dimension score is negatively correlated with the time interval and the duration by a preset value; An operation error ratio corresponding to the operation error times is extracted from the historical operation behavior data, and the operation error ratio is a ratio of the operation error times to the total operation times; The time dimension score and the operation error times are input into a pre-designed calculation function to obtain the operation proficiency.

[0009] By adopting the above technical solutions, the operation time sequence features, including the time intervals between consecutive operations and the duration of a single operation, are extracted from the historical operation behavior data, and a negative correlation is established with the time interval and the duration, reflecting the influence of operation efficiency on proficiency. At the same time, the system obtains the operation error ratio by calculating the ratio of the operation error times to the total operation times, and the operation accuracy is included in the evaluation system. This evaluation method combining the time dimension score and the operation error ratio more accurately reflects the actual ability level of the user in the operation speed and operation accuracy dimensions, so that the system can more accurately evaluate the operation proficiency of the user.

[0010] In some embodiments of the first aspect, in some embodiments, the pre-designed calculation function is: S = a x (1 - b x T) + (1 - a) x (1 - E) In the above function, S is the operation proficiency, a is a time dimension weight coefficient, b is a time dimension adjustment coefficient, T is a time dimension score, and E is an operation error proportion.

[0011] By adopting the above technical solution, the function balances the influence degree of the time dimension score and the operation error proportion on the final proficiency through the weight coefficient a, and fine-tunes the time dimension score using the adjustment coefficient b, so that the evaluation result is more in line with the actual situation.

[0012] In some embodiments of the first aspect, in some embodiments, after generating the guidance result according to the input information of all guidance sub-processes stored in the guidance information pool, the method further comprises: constructing an associated index table of guidance information items according to the information items in each guidance sub-process and the corresponding feature words; identifying feature words from the input information of the user; finding the corresponding information items and guidance sub-processes according to the feature words in the associated index table; when the input information corresponds to a guidance sub-process that is not the currently executed guidance sub-process, temporarily storing the input information in the to-be-allocated information pool; determining the target guidance sub-process corresponding to the input information; when the target guidance sub-process is executed, filling the input information in the to-be-allocated information pool into the corresponding information items; statistically determining the time points at which each guidance sub-process receives input information; adjusting the execution order of the guidance sub-processes according to the order of the time points.

[0013] By adopting the above technical solution, by constructing the associated index table of guidance information items and identifying feature words from the input information of the user, the system realizes intelligent allocation of information and dynamic optimization of the guidance process. When the identified information does not match the current guidance sub-process, the system temporarily stores it in the to-be-allocated information pool and automatically fills it in when the corresponding sub-process is executed, reducing the repeated collection of information. The system adjusts the execution order by statistically determining the time points at which each guidance sub-process receives input information, so that the guidance process is more in line with the natural expression habits of the user, reducing the operation cost of the user repeatedly switching between multiple guidance sub-processes, improving the naturalness and efficiency of the guidance process, and making the entire guidance process more in line with the thinking habits and expression methods of the user.

[0014] In some embodiments of the first aspect, in some embodiments, identifying feature words from the input information of the user specifically comprises: extracting a symptom description phrase in the input information; removing adverbs and adjectives in the symptom description phrase to obtain a filtered symptom description phrase; determining a word in the filtered symptom description phrase as a feature word if the word has a similarity greater than a preset similarity threshold to a preset standard symptom word library.

[0015] By adopting the technical solutions described above, the symptom description is extracted from the input information, the adverbs and adjectives are removed to obtain the filtered symptom description phrase, and then the filtered symptom description phrase is compared with the preset standard symptom word library to identify the feature word, thereby improving the standardization degree of the symptom description and the accuracy of feature word identification.

[0016] In combination with some embodiments of the first aspect, in some embodiments, after adjusting the execution order of the guided diagnosis sub-processes according to the order of time points, the method further comprises: identifying an information dependency relationship between adjacent guided diagnosis sub-processes, the information dependency relationship representing whether information in a preceding guided diagnosis sub-process is a necessary condition for information in a subsequent guided diagnosis sub-process; in a case where it is determined that the information dependency relationship does not conform to the adjusted execution order, combining the guided diagnosis sub-processes having the information dependency relationship into an execution unit; generating an execution constraint identifier, the execution constraint identifier being used to mark the execution order of the guided diagnosis sub-processes in the execution unit; and controlling the adjusted execution order according to the execution constraint identifier.

[0017] By adopting the technical solutions described above, the information dependency relationship between the guided diagnosis sub-processes is identified, and the sub-processes having the dependency relationship are combined into an execution unit, and the execution constraint identifier is used to control the execution order, thereby solving the problem of information dependency conflict that may be caused by adjusting the execution order of the guided diagnosis sub-processes. When the system adjusts the execution order of the guided diagnosis sub-processes according to the time order of the user input information, there may be a case where the information required by a subsequent process has not been collected in a preceding process. By establishing the identification mechanism of the information dependency relationship, the system can find this potential conflict. The sub-processes that depend on each other are combined into an execution unit and marked using the execution constraint identifier, thereby improving the completeness and continuity of information collection. This processing mechanism maintains the flexibility of the guided diagnosis process while ensuring the logical order of information collection, avoiding information missing or collection errors caused by execution order adjustment, and improving the reliability and accuracy of the guided diagnosis process.

[0018] In combination with some embodiments of the first aspect, in some embodiments, identifying the information dependency relationship between adjacent guided diagnosis sub-processes specifically comprises: obtaining the filling rules of each information item in the guided diagnosis sub-processes; In a case where the filling rule of the information item contains a reference to any information item, the guided sub-process in which the referenced information item is located is determined as the preceding guided sub-process; The guided sub-process in which the referenced information item is located is determined as the subsequent guided sub-process; An information dependency relationship between the preceding guided sub-process and the subsequent guided sub-process is established.

[0019] By adopting the above technical solution, the automatic identification of the guided flow information dependency relationship is realized by analyzing the filling rule of the information item in the guided sub-process, identifying the information item reference relationship in the rule, and establishing the information dependency relationship between the preceding and subsequent guided sub-processes according to the reference relationship. The reference relationship in the information item filling rule directly reflects the logical association between different information items, and the system can more accurately locate the direction and dependency degree of the information flow. The sub-process containing the referenced information item is determined as the preceding process, and the sub-process containing the reference information item is determined as the subsequent process, so that the system can automatically build a complete information dependency network, and ensure that the guided flow can always maintain the correct information transmission sequence when dynamically adjusted.

[0020] In a second aspect, the embodiments of the present application provide a personalized intelligent guided system, which comprises one or more processors and a memory; the memory is coupled with the one or more processors, and the memory is used to store computer program codes, the computer program codes comprising computer instructions, and the one or more processors invoke the computer instructions to enable the system to perform the method described in the first aspect and any possible implementation manner of the first aspect.

[0021] In a third aspect, the embodiments of the present application provide a computer readable storage medium, which comprises instructions, and when the instructions run on a system, enable the system to perform the method described in the first aspect and any possible implementation manner of the first aspect.

[0022] In a fourth aspect, the embodiments of the present application provide a computer program product, and when the computer program product runs on a system, enable the system to perform the method described in any possible implementation manner of the first aspect.

[0023] The one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. The application provides a personalized intelligent guide diagnosis method. The system evaluates the actual operation ability of the user according to the operation step completion time and the number of operation errors in the historical operation behavior data, thereby avoiding the use experience difference caused by the same process of the traditional fixed guide diagnosis process for different users. By dividing the guide diagnosis process into multiple sub-processes and executing them one by one, the cognitive burden of the user in the guide diagnosis process is reduced. The system stores the input information of the user in each guide diagnosis sub-process into the guide diagnosis information pool and generates a guide diagnosis progress identifier, so that the user can directly continue the unfinished guide diagnosis sub-process when re-entering after exiting in the middle of the way, and the stored input information is automatically called, thereby reducing the number of times of repeated input information of the user, improving the continuity and efficiency of the guide diagnosis process, and improving the use experience of the user.

[0024] 2. The application provides a personalized intelligent guide diagnosis method. By constructing an associated index table of guide diagnosis information items and identifying feature words from user input information, the system realizes intelligent allocation of information and dynamic optimization of guide diagnosis process. When the identified information does not match the current guide diagnosis sub-process, the system temporarily stores it in the to-be-allocated information pool and automatically fills it when the corresponding sub-process is executed, thereby reducing the repeated collection of information. The system adjusts the execution order by counting the time points of receiving input information of each guide diagnosis sub-process, so that the guide diagnosis process is more in line with the natural expression habits of the user, reduces the operation cost of the user in repeatedly switching between multiple guide diagnosis sub-processes, improves the naturalness and efficiency of the guide diagnosis process, and makes the entire guide diagnosis process more in line with the thinking habits and expression methods of the user.

[0025] 3. The application provides a personalized intelligent guide diagnosis method. By identifying the information dependency relationship between guide diagnosis sub-processes and combining sub-processes with dependency relationship into execution units, and using execution constraint identifier to control the execution order, the problem of information dependency conflict caused by adjustment of the execution order of guide diagnosis sub-processes is solved. When the system adjusts the execution order of guide diagnosis sub-processes according to the time sequence of user input information, the information required by the subsequent process may not have been collected in the previous process. By establishing an identification mechanism of information dependency relationship, the system can find this potential conflict. By combining sub-processes that depend on each other into execution units and marking them with execution constraint identifier, the integrity and continuity of information collection are improved. This processing mechanism not only maintains the flexibility of the guide diagnosis process, but also ensures the logical order of information collection, avoids information missing or collection errors caused by execution order adjustment, and improves the reliability and accuracy of the guide diagnosis process. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 It is a flowchart of a personalized intelligent guide diagnosis method in the embodiment of the application.

[0027] Figure 2This is a flowchart illustrating a method for dynamically allocating triage information based on feature word recognition in an embodiment of this application.

[0028] Figure 3 This is a schematic diagram of the physical device structure of a personalized intelligent triage system provided in an embodiment of this application. Detailed Implementation

[0029] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification and appended claims of this application, the singular expressions “a,” “an,” “the,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to any or all possible combinations including one or more of the listed items.

[0030] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0031] The following example is used in conjunction with Figure 1 The following describes a personalized intelligent triage method in an embodiment of this application: Please see Figure 1 This is a flowchart illustrating a personalized intelligent triage method in an embodiment of this application.

[0032] S101. If it is determined from the triage request sent by the user that there is historical operation behavior data corresponding to the user, calculate the operation proficiency based on the historical operation behavior data. If the existence of historical operation behavior data corresponding to the user is determined based on the triage request sent by the user, the operation proficiency is calculated based on the historical operation behavior data. The operation proficiency is calculated by a preset calculation function based on the completion time of operation steps and the number of operation errors in the historical operation behavior data. Specifically, the calculation of operation proficiency based on historical operation behavior data includes: extracting the operation time sequence features corresponding to the completion time of operation steps based on the historical operation behavior data. The operation time sequence features include the time interval between consecutive operations and the duration of a single operation. A time-dimensional score is constructed based on the operation time sequence characteristics. The time-dimensional score is negatively correlated with the time interval and duration by a preset value. According to the historical operation behavior data, an operation error ratio corresponding to the operation error times is extracted, and the operation error ratio is a ratio of the operation error times to the total operation times. The time dimension score and the operation error times are input into a pre-designed calculation function to obtain the operation proficiency.

[0033] The pre-designed calculation function is: S = a x (1 - b x T) + (1 - a) x (1 - E) In the above function, S is the operation proficiency, a is the time dimension weight coefficient, b is the time dimension adjustment coefficient, T is the time dimension score, and E is the operation error ratio.

[0034] Step S101 is to calculate the operation proficiency according to the historical operation behavior data in the case that the historical operation behavior data corresponding to the user exists according to the user sent guidance request. The historical operation behavior data here refers to the operation record of the user using the guidance system in the past, including the start and end time of each guidance, the completion time of each operation step, the input content of the user, the error or exception in the operation process and other information. The system can evaluate the familiarity and operation efficiency of the user to the guidance system, i.e. the operation proficiency, by analyzing these historical data.

[0035] Specifically, the system first judges whether the historical operation behavior data of the current user exists. This judgment can be realized through the unique identification of the user (such as user ID, user name, device ID, etc.). If there is no historical data, it means that the user is using the system for the first time, and the operation proficiency cannot be calculated, and more guidance may be needed in the future. If there is historical data, the system extracts the features related to the evaluation of operation proficiency from the historical data, mainly including two aspects: one is the completion time of the operation step, and the other is the number of errors in the operation process.

[0036] For the operation completion time, the system mainly analyzes the time spent by the user in each operation step. Generally, a skilled user can complete each step faster and the operation is more coherent. Therefore, the system extracts the user's stay time in each step, calculates the switching interval between consecutive steps, and the time spent in completing a single step, as the time features for evaluating the operation proficiency. For the operation error times, the system mainly counts the number of times of filling errors, logical errors, abnormal submissions and other situations in the guidance process of the user, as the accuracy features for evaluating the operation proficiency.

[0037] After extracting relevant features, the system comprehensively considers time and accuracy features, and calculates a score representing the user's overall operational proficiency using a preset calculation function. Operational proficiency can be a numerical value from 0 to 100, or a qualitative level description such as "proficient," "relatively proficient," "average," and "unskilled." The system can choose an appropriate representation method based on the actual situation; no specific limitations are imposed here.

[0038] It should be noted that the methods for collecting historical operation behavior data, the data storage format, the algorithms for extracting time features and accuracy features, and the selection of proficiency calculation models mentioned in step S101 can all be adjusted according to the actual design and performance requirements of the system, and are not limited to the examples given in the text. No restrictions are imposed here.

[0039] The system allows setting up front-end tracking points at various key nodes on the user guidance interface. When a user performs an action, the system automatically records the timestamp, action type, and input information, generating tracking logs. These logs are uploaded to the server in real time via the network. The server then categorizes the logs by user ID and writes them to the user action log database.

[0040] S102. Divide the triage process into several sub-processes based on the operator's proficiency. This step divides the entire triage process into several sub-processes based on the user's proficiency. Proficiency reflects the user's familiarity with the system; highly proficient users can execute more complex and longer sub-processes, while less proficient users need to break the process down into several simpler, shorter sub-processes. The granularity of the sub-process division is negatively correlated with proficiency; higher proficiency results in fewer sub-processes, but each sub-process contains more steps. The sub-process division method can be designed based on the complexity of the actual triage process and the functionality of the modules, and is not specifically limited here. By rationally dividing the sub-processes, users of different proficiency levels can adapt to the system, improving triage efficiency and user experience.

[0041] Multiple sub-process division schemes can be designed, each corresponding to a proficiency range. The appropriate sub-process scheme is selected based on the user's proficiency level. For example, scheme A is used for users with less than 30% proficiency, scheme B for users with 30%-60% proficiency, and scheme C for users with more than 60% proficiency. The granularity of the sub-processes decreases sequentially across different schemes.

[0042] A dynamic adjustment mechanism for sub-process division can also be designed. During the user's operation, the performance of each sub-process can be tracked, such as completion time and number of errors, to further refine the user's proficiency and dynamically adjust the division of sub-processes to match the user's actual proficiency level.

[0043] S103, execute each sub-guidance process in sequence, and obtain the input information of the user in the current sub-guidance process; The system executes each sub-process in sequence according to the sub-guidance processes divided in S102, usually in a dependent order. During the execution of each sub-process, the system guides the user to input information or make a selection through the interface, and obtains the input information of the user for the current sub-process. The input information can be in various forms such as text, numerical value, and options, depending on the functional design of the sub-process. The system obtains information in the following ways, but not limited to: displaying prompt information for the user to fill in the input box, displaying selection items for the user to click, providing voice input and performing voice recognition, etc. The content, form and input method of the information are not limited here.

[0044] The system can use a wizard-style interaction, each sub-process corresponds to a wizard page, guiding the user to fill in and select information in the page, and setting the previous step and next step buttons to allow the user to switch between sub-processes.

[0045] For some complex sub-processes, the system can further split them into multiple steps, each step corresponding to a page or dialog box, guiding the user to input different information step by step. Verification rules can be set between steps, and only when the information of the current step is verified, can the next step be entered.

[0046] S104, store the corresponding relationship between the input information and the current sub-guidance process in the guidance information pool; After obtaining the input information of the user in the current sub-guidance process, the system needs to establish a corresponding relationship between the input information and the current sub-process, and store it in the guidance information pool. The guidance information pool can be a data structure in memory, or an external persistent storage (such as a database). The corresponding relationship between the input information and the sub-process can be represented in various ways, such as stored in the form of Key-Value, where Key is the unique identifier of the sub-process, and Value is the input information set of the sub-process. The purpose of the system to store the input information in the information pool is to facilitate the acquisition and use of these information in subsequent processes, especially after the guidance is interrupted, the previous input state can be restored to avoid information loss. The design and storage method of the information pool can be flexibly selected according to the actual system requirements, which is not limited here.

[0047] Use the Key-Value form of memory-level data structure such as HashMap to realize the guidance information pool, Key is the sub-process identifier, Value is the input information object, and each time the input information of a sub-process is obtained, it is put into HashMap.

[0048] The persistent storage of the guiding information pool is implemented using a relational database. A sub-process table and an input information table are designed to store sub-process meta-information and input information, respectively, and the two tables are associated by a sub-process ID. After obtaining the input information, it is inserted into the input information table, and the state of the corresponding sub-process record is updated.

[0049] S105. In the case where it is detected that the user exits the current guiding sub-process, a guiding progress identifier is generated. This step is mainly used to record the guiding progress in the case where the user abnormally exits the guiding, so as to be recovered subsequently. The system monitors the behavior of the user to determine whether the user exits the current guiding sub-process. Common exit scenarios include, but are not limited to, the user actively closing the guiding page, system crash, network interruption, etc. Once it is detected that the user exits, the system generates a guiding progress identifier to indicate which sub-process the guiding process has executed to and the completion state of the sub-process. Specifically, the progress identifier can include the current sub-process ID, the completion state (completed / unfinished), and the information items that have been filled in, etc. The generated progress identifier can be stored in the guiding information pool or a separate persistent storage.

[0050] AOP aspect programming is used to set a pointcut at a key execution point (such as entering a sub-process, completing input, etc.) of each sub-process, and the guiding progress identifier is recorded and updated at the pointcut.

[0051] The page lifecycle events (such as onunload and onbeforeunload) and AJAX exception callbacks are used in the web front end to capture the cases where the user exits the page and the request is abnormal, and send a notification to the back end to generate the progress identifier.

[0052] S106. In response to detecting that the user resends the guiding request, the guiding sub-process that is not completed is determined according to the guiding progress identifier, and the input information stored in the guiding information pool is called in the guiding sub-process that is not completed. When the user reopens the guiding system and sends a guiding request, the system first checks whether there is a guiding progress identifier generated previously. If there is a progress identifier, it indicates that the previous guiding process is not completed and needs to continue the execution according to the previous progress. The system analyzes the guiding progress identifier to determine which sub-process should be executed currently and whether the sub-process has been completed. If the current sub-process is not completed, the system extracts the input information that has been filled in and stored in the guiding information pool when the sub-process is executed again, and directly loads it onto the interface for the user to further edit and modify without re-inputting. If the previous input information is incomplete, the user can continue to fill in the remaining information. If the current sub-process has been completed, the system automatically jumps to the next sub-process. Through the processing of this step, the user can continue the guiding process from the last interruption, which improves the user experience.

[0053] At the entrance of the triage request, the acquisition and analysis logic of the triage progress is set. First, check if there is a progress identifier, if there is, locate to the specific sub-process according to the sub-process ID recorded in the identifier.

[0054] For the unfinished sub-process, according to the input information index recorded in the progress identifier, batch get the corresponding input information from the triage information pool. Fill these information into the interface control of the sub-process, and realize the automatic loading of information through data binding.

[0055] This step may exist the problem that the input information in the triage information pool is inconsistent with the actual demand, such as the information entered in the last triage has become outdated, or new content needs to be added on the basis of the original information. For this, the system needs to provide convenient information editing function after loading the existing information, allowing users to modify the existing information. For example, for single / multiple selection type information, the system should allow users to recheck; for date type information, allow users to reselect date; for open question information, give users the right to edit freely. At the same time, it is necessary to clearly identify which information has been entered before on the interface, prompt the user to review and confirm, in order to ensure the accuracy of the information. The system can also set the mandatory recording system for information, for some sensitive or time-sensitive information, reacquire in this triage, rather than follow the previous information.

[0056] S107, generate the triage result according to the input information of all triage sub-processes stored in the triage information pool.

[0057] When all the triage sub-processes are executed, the system will start to generate the final triage result according to all the input information of the sub-processes accumulated in the triage information pool. Specifically, the system will traverse the triage information pool, extract the input information of each sub-process in turn according to the execution order of the sub-processes, and integrate these information according to the predetermined format and rules to form a complete triage report. The content and format of the triage report can be customized according to the actual needs of the hospital, usually including patient basic information, chief complaint, present illness history, past history, preliminary diagnosis, examination / treatment suggestion and other parts. The generated triage result can be output in the form of paper print or electronic document, provided to the doctor and patient. The doctor can quickly understand the basic situation of the patient according to the triage result, and make further diagnosis and treatment decision.

[0058] Use the join table query of the database, join the triage sub-process table and the input information table, arrange according to the execution order of the sub-processes, and extract all the input information at once. Then convert the query result into a fixed format triage report.

[0059] On the basis of the triage information pool, a set of XML or JSON format triage result template is designed. The KV pairs in the information pool are traversed, and the input information is filled into the corresponding template tags. Through XSLT or template engine technology, the filled template is converted into a triage report.

[0060] The triage result generated in this step may not be accurate and comprehensive, especially for more complex diseases. The information collected by triage alone is not enough to draw a diagnosis conclusion, and further doctor consultation, physical examination, and inspection are needed. Therefore, when generating the triage result, the system needs to clearly inform the patient that the result is only for reference and cannot replace the doctor's diagnosis. In the triage result, the doctor's suggestions and the next examination can be set to guide the patient for subsequent treatment.

[0061] In the above embodiment, the system evaluates the user's actual operation ability according to the operation step completion time and operation error times in the historical operation behavior data, avoiding the difference in use experience caused by the same flow for different users in the traditional fixed triage process. By dividing the triage process into multiple sub-processes and executing them one by one, the cognitive burden of the user in the triage process is reduced. The system stores the input information of the user in each triage sub-process into the triage information pool and generates a triage progress identifier, so that the user can directly continue the unfinished triage sub-process when re-entering after exiting in the middle of the way, and the stored input information is automatically called, reducing the number of times the user repeats the input information, improving the continuity and efficiency of the triage process, and improving the user's use experience.

[0062] The above embodiment realizes the personalized configuration of the triage process by calculating the operation proficiency and dividing the triage sub-process, and solves the information retention problem after the triage process is interrupted through the design of the triage information pool and the progress identifier. However, in actual application, the user often describes the symptoms according to his own thinking habits and expression methods. Such natural language input may not completely match the preset triage process sequence. In order to better adapt to the input habits of the user and improve the intelligent level of the triage system, the present application also provides a triage information dynamic allocation method based on feature word recognition. This method realizes intelligent recognition and flexible allocation of input information by establishing an association index between information items and feature words, and can dynamically adjust the triage process according to the actual input sequence of the user, so that the triage system better adapts to the expression habits of the user. The following will be described in combination with Figure 2 A triage information dynamic allocation method based on feature word recognition in the embodiment of the present application is described: Please refer to Figure 2 A flowchart of a triage information dynamic allocation method based on feature word recognition in the embodiment of the present application is shown.

[0063] S201, construct an associated index table of the information items according to the information items in each sub-guidance process and the corresponding feature words; This step is a preprocessing work completed in the system design phase, and the purpose is to establish the mapping relationship between the information items and the feature words, and to provide the basis for subsequent input information recognition and distribution. Specifically, the system first sorts out the information items that need to be filled in by the user in each sub-guidance process, such as name, gender, date of birth, chief complaint, symptom description, etc. Then, the system defines one or more sets of feature words for each information item. Feature words are key words that can represent the characteristics of the information item, and are usually entity words, symptom words, body part words, etc. that are highly related to the content of the information item. Feature words can be selected from medical dictionaries, specialty word libraries, or defined by medical experts based on guidance experience, or obtained by mining and analyzing medical-patient conversation corpus. The system does not limit the number and type of feature words, and can be flexibly set according to actual conditions to obtain more comprehensive and accurate information item representation. After obtaining the corresponding relationship between the information items and the feature words, the system constructs a hash index table with the information item as the key and the feature word as the value. The index table can also include meta information such as the sub-guidance process to which the information item belongs and the position of the information item in the interface, facilitating quick retrieval and positioning of information.

[0064] The system can use a forward index mode to establish a one-to-many mapping from information items to feature words with the information item as the primary key. For example, the "chief complaint" information item corresponds to feature words such as "unwell, pain, uncomfortable", and the "symptom duration" information item corresponds to time feature words such as "days, weeks, months". When matching, if the input information contains these feature words, the corresponding primary keys "chief complaint" and "symptom duration" can be quickly retrieved.

[0065] The system can also use an inverted index mode to establish a one-to-many mapping from feature words to information items with the feature word as the primary key. For example, the feature words "cough, expectoration" correspond to the "chief complaint" and "companion symptoms" information items, and the feature words "abdominal pain, diarrhea" correspond to the "chief complaint" and "medical history" information items. When matching, the system extracts the key words from the input information, and then uses these key words as the primary key to find the corresponding set of information items.

[0066] S202, identify feature words from the user's input information; The system identifies feature words from the user's input information, specifically including: extracting symptom description phrases from the input information; removing adverbs and modifiers from the symptom description phrases to obtain filtered symptom description phrases; determining words in the filtered symptom description phrases that have a similarity greater than a preset similarity threshold to a preset standard symptom word library as feature words.

[0067] This step is performed in real time during the process of guiding. When the user inputs information in the form of natural language in the guiding interface, the system analyzes the input content in real time and identifies the feature words that may be contained therein. Specifically, the system first preprocesses the text input by the user, including word segmentation, removal of stop words, part-of-speech tagging, etc. Then, the system matches the preprocessed words with a pre-constructed index table of feature words. If the input text contains a feature word registered in the index table, the identification is successful. It should be noted that the input of the user may be colloquial, non-standard, and different from the standard medical feature words. Therefore, the system needs to consider the similarity of words when matching, and can calculate the edit distance and semantic similarity between the input words and the feature words, and set a similarity threshold. When the similarity is greater than the threshold, it is considered that the matching is successful. As for the specific algorithm of similarity calculation and the setting of the threshold, it can be optimized according to the actual effect of the system, which is not limited here. The input information that cannot be matched to the feature words can be stored in the to-be-processed queue for subsequent manual review.

[0068] The system can use string matching algorithms such as KMP, Boyer-Moore, etc. to directly search for feature words in the input text. This way is simple to implement, but has high requirements for the format of the input text and the consistency of the feature words.

[0069] The system can use machine learning methods such as conditional random fields (CRF), recurrent neural networks (RNN), etc. to perform sequence labeling on the input text and automatically identify the feature words therein. This way can better handle colloquial and non-standard inputs and has higher recognition accuracy. However, it requires pre-labeled training corpus and optimization of the machine learning model, which is relatively complex to implement.

[0070] S203, find the corresponding information item and guiding sub-process according to the feature word in the association index table; After identifying the feature words contained in the input information in step S202, the system needs to further determine the information items corresponding to these feature words and the guiding sub-processes to which they belong, in order to subsequently classify and fill the information. Specifically, the system uses the identified feature words as clues to search in the pre-constructed association index table. If the index table is indexed by information items, the system first finds the information item containing the feature word, and then obtains the guiding sub-process corresponding to the information item. If the index table is indexed by feature words, the system can directly obtain the information item and guiding sub-process associated with the feature word. It should be noted that the search result of the index table may not be unique, i.e. the same feature word may appear in multiple information items, and an information item may also belong to multiple sub-processes at the same time. This one-to-many association relationship is allowed to exist, reflecting the flexibility of the system. The system can take all the found information items and sub-processes as candidates and leave them for screening and confirmation in the next step, and the number of candidate results is not limited here.

[0071] If the association index table is stored in a hash table, the system can quickly locate the corresponding information item and sub-process through the hash value of the feature word, and the search complexity is O(1).

[0072] If the association index table is stored in a B+ tree or other structure, the system can match the key value of the feature word in the tree to obtain the corresponding information item and sub-process, and the search complexity is O(logn).

[0073] S204, when the input information corresponds to a guidance sub-process that is not the currently executed guidance sub-process, temporarily storing the input information into a to-be-allocated information pool; This step is to cache and manage the user input information. Due to the flexibility of natural language input, the user may input information belonging to other sub-processes in a certain guidance sub-process. In order to avoid the loss and confusion of information, the system does not directly fill these information into the current sub-process, but temporarily stores it in a to-be-allocated information pool. The to-be-allocated information pool can be a separate storage structure such as a list, a queue, etc., which is separated from the storage of the guidance sub-process. Each information in the pool is labeled with the information item label and the sub-process label found in step S203, which is used for subsequent information classification. It should be noted that for the case where the guidance sub-process corresponding to the input information is exactly the currently executed sub-process, the system can directly fill the input information into the corresponding information item on the interface without storing it in the to-be-allocated information pool, which is not required here.

[0074] The system can maintain a private to-be-allocated information pool for each user, using a List data structure, and add each received input information that cannot be directly classified to the end of the List.

[0075] If multiple users share the same to-be-allocated information pool, in order to avoid information conflict, the system can add a user identifier when storing, and implement the to-be-allocated information pool as a Hashmap<UserId, List <inputinfo>the structure of

[0076] S205, determining a target guiding sub-process corresponding to the input information; In step S203, the system has preliminarily found out the guiding sub-processes to which the input information may correspond, but the result may not be unique, and needs to be finally confirmed in this step. Specifically, the system first sorts all information in the to-be-assigned information pool in chronological order, and then checks the sub-process labels carried by each piece of information. For information containing only one sub-process label, the sub-process can be directly determined as the target sub-process; for information containing multiple sub-process labels, the system needs to perform a correlation analysis to calculate the correlation degree of the information with each candidate sub-process, and determine the sub-process with the highest correlation degree as the target sub-process. The correlation analysis can utilize the content of the information item filled in the candidate sub-process to perform semantic matching with the current information, or can investigate the coincidence degree of the feature words in the current information and the feature words corresponding to the candidate sub-process. Information that cannot determine the explicit target sub-process can continue to be retained in the to-be-assigned information pool, waiting for subsequent re-analysis.

[0077] The system can use the VSM (Vector Space Model) algorithm to represent the input information and the candidate sub-processes as vectors, and calculate the correlation between the two by calculating the cosine similarity between the vectors.

[0078] The system can also use a word embedding model such as Word2Vec to map the input information and the candidate sub-processes into the same semantic space, and calculate the semantic similarity by using Euclidean distance, Manhattan distance, etc.

[0079] This step may exist the situation that the input information cannot be closely related to any guiding sub-process, resulting in difficulty in determining its explicit attribution. The reason for this situation may be that the input information contains new symptoms or new expressions that are not defined by the system, or the input information is too simple and general. For this kind of information that cannot be classified, the system temporarily retains it in the to-be-assigned information pool, and at the same time can send feedback and prompts to the user, requiring them to provide supplementary information related to the defined guiding sub-process. After accumulating a certain amount of unclassified information, the system can also find out whether there are new symptom patterns in it through clustering and other methods, and expand and refine the original guiding sub-processes accordingly.

[0080] S206, filling the input information in the to-be-assigned information pool into the corresponding information item when executing the target guiding sub-process; In the process of system performing the guided diagnosis process, when it is the turn of the target sub-process determined in step S205, the system takes out the input information belonging to the sub-process in the to-be-allocated information pool, and fills the content thereof into the corresponding control of the interface according to the information item label, to complete the final classification of the information. It should be noted that due to the randomness of natural language input, the user may include the content of multiple information items in one input segment. Therefore, when filling, the system needs to further perform semantic segmentation on the segment to identify the text interval corresponding to different information items, and then fill the content thereof into the corresponding control. After filling, the system deletes the filled information from the to-be-allocated pool to avoid repeated filling. For the target sub-process that has not been executed, the corresponding input information continues to be retained in the to-be-allocated pool until it is taken out and filled when the corresponding sub-process is executed.

[0081] If the interface control of the target sub-process is a text box, the system can directly fill the content of the corresponding information; if the interface control is a single / multiple selection box, the system can match the information content with the option text, and automatically check the option with the highest matching degree.

[0082] For long text containing multiple information items, the system can use sequence labeling models such as BiLSTM-CRF to identify the text segments corresponding to different information items. Syntax analysis can also be used to identify the boundaries of clauses and sentences according to punctuation marks, conjunctions, etc., and then perform semantic classification.

[0083] This step may have the case that the user subsequently modifies the filled information, resulting in inconsistency between the filled information and the final submitted information. The user may add, delete or modify the information after the system automatically fills part of the information. To ensure the synchronization of the information, the system should reacquire the latest filled content on the interface when the user actively submits the current sub-process, and update the related entries in the to-be-allocated information pool and the allocated information storage accordingly. The system can also provide the functions of undo and restore on the interface to allow the user to conveniently revert to the state filled by the system. In addition, for the information newly added by the user, if it still belongs to the current sub-process, the system directly stores it in the local storage; if it belongs to other executed or unexecuted sub-processes, the system re-stores it in the to-be-allocated information pool for subsequent classification.

[0084] S207, statistics the time points at which each guided diagnosis sub-process receives input information; Due to the introduction of the mechanism of the information pool to be allocated, the time when each sub-process receives the input information can not be completely consistent with the execution time set in the triage process. When the system performs the information filling operation in step S206, it records the time stamp of each information in the information pool to be allocated, which represents the time when the sub-process actually receives the information. After completing all information filling, the system summarizes the time stamps recorded by each sub-process, and can obtain the time distribution of each sub-process receiving the input information. The receiving time distribution reflects the order of filling various types of information by the user, and embodies the user's thinking habit. The statistical method of the distribution can be flexibly set, such as calculating the earliest receiving time and the average receiving time of each sub-process, which is not limited here. It should be noted that for the information filled directly on the interface without passing through the information pool to be allocated, its receiving time can be directly recorded as the filling time.

[0085] The system can set a time stamp field for each information in the information pool to be allocated, which records the time when the information enters the information pool. Each time the information filling is performed, the receiving time stamp list of the target sub-process is updated.

[0086] The system can also record a classification log in the database in real time when performing information filling, including information content, belonging sub-process, filling time and other fields, and the receiving time distribution can be obtained by subsequent aggregation query.

[0087] S208, adjust the execution order of the triage sub-process according to the order of the time points.

[0088] The system sorts the average receiving time of each sub-process to obtain a sub-process order reflecting the user's habit. Then, the system compares this order with the pre-set sub-process order, and calculates the similarity between the two. If the similarity is lower than a certain threshold, it means that the original order deviates greatly from the user's habit and needs to be adjusted. If the similarity is higher than the threshold, it means that the original order has already well met the user's habit and can be continued to be used. The calculation of the order similarity can use algorithms such as edit distance and Kendall Tau distance, and the setting of the threshold can be adjusted according to the actual effect, which is not limited here. After deciding to adjust the order, the system needs to further check whether the adjusted order will introduce logical errors, such as whether the dependency relationship of the information items still holds. The system can predefine a set of constraint rules for the sub-process and its information items, and automatically verify the adjusted order to ensure its rationality. After the order adjustment is completed, the system can guide the user to fill in the information according to the new order in the subsequent execution of the triage process, better fit the user's habit, and optimize the triage experience.

[0089] The system can use a greedy algorithm to select the subprocess that is closest to the average receiving time of the current subprocess and has not been selected yet, based on the user's preferred order, as the next subprocess to be executed, until all subprocesses have been selected.

[0090] The system can also employ integer programming algorithms to model the order reordering problem as a minimum cut problem, obtaining the optimal reordering scheme by solving the integer programming model. The objective function can be to minimize the difference between the conventional order and the reordered order.

[0091] In the above embodiments, by constructing an associated index table of triage information items and identifying feature words from user input information, the system achieves intelligent information allocation and dynamic optimization of the triage process. When the identified information does not match the current triage sub-process, the system temporarily stores it in the information pool to be allocated and automatically fills it when the corresponding sub-process is executed, reducing the duplication of information collection. The system adjusts the execution order by statistically analyzing the time points when each triage sub-process receives input information, making the triage process more in line with the user's natural expression habits, reducing the operational cost of users repeatedly switching between multiple triage sub-processes, improving the naturalness and efficiency of the triage process, and making the entire triage process more in line with the user's thinking habits and expression methods.

[0092] Furthermore, in another embodiment, after adjusting the execution order of the triage sub-processes according to the time sequence, the system also identifies the information dependency relationship between adjacent triage sub-processes. The information dependency relationship characterizes whether the information in the preceding triage sub-process is a necessary condition for the information in the subsequent triage sub-process. Specifically, this includes: obtaining the filling rules for each information item in the triage sub-process; if it is determined that the filling rules for an information item contain a reference to any information item, determining the triage sub-process containing the referenced information item as the preceding triage sub-process; determining the triage sub-process containing the referenced information item as the subsequent triage sub-process; and establishing the information dependency relationship between the preceding and subsequent triage sub-processes. If it is determined that there are information dependencies that do not conform to the adjusted execution order, the triage sub-processes with information dependencies will be combined into an execution unit; Generate execution constraint identifiers, which are used to mark the execution order of each diagnostic sub-process in the execution unit; The execution order is adjusted based on the execution constraint identifier.

[0093] The system analyzes the filling rules of information items in each sub-process to determine whether there is information dependency between sub-processes. The filling rules can be expressions or functions used to check and calculate information items. If the rule of an information item refers to other information items, it means that the information item depends on the referenced information items. By parsing the cross-reference relationship in the rules, the system can establish a dependency directed graph, where the nodes represent the sub-processes and the directed edges (A, B) represent that the sub-process A depends on the sub-process B, i.e., the information items in A refer to the information items in B.

[0094] For the identified dependency relationship, the system checks whether it is consistent with the adjusted sub-process execution order. If the post-order sub-process is adjusted to be before the pre-order sub-process, it means that the new execution order violates the existing information dependency logic. In this case, instead of simply exchanging the execution order of the two, the system organizes a group of sub-processes with dependency relationships into an indivisible execution unit. The execution unit can be a linear sequence or a sub-DAG (Directed Acyclic Graph). The execution order of sub-processes within the unit is arranged according to the original dependency relationship and is not affected by external adjustments.

[0095] For the assembled execution unit, the system generates an execution constraint identifier, which can formally represent the execution order constraints of each sub-process within the execution unit. One representation method is to use a sequence pattern similar to regular expressions, such as "A.B+.C?" which means that A must be executed first, then one or more B, and finally optionally C. Another representation method is to use a partial order relationship, such as "{A < B, A < C}" which means that A must be executed before B and C in the unit. The partial order relationship here can be derived from the sub-process dependency graph by topological sorting algorithms, etc.

[0096] After determining the execution unit and its internal sub-process order constraints, the system reconsiders the adjusted overall sub-process execution order, extracts those execution units that violate the constraints, adjusts their positions in the order, and continues until all constraints are met. The optimization goal of adjusting the position is to make the execution order of the execution unit consistent with the average receiving time order of its internal sub-processes as much as possible to take into account user habits. The adjustment can use heuristic optimization algorithms such as local search, simulated annealing, etc.

[0097] In the above embodiment, by identifying the information dependency relationship between the sub-processes of the guided diagnosis, and combining the sub-processes with the dependency relationship into an execution unit, and using the execution constraint identifier to control the execution order, the problem of information dependency conflict caused by the adjustment of the execution order of the guided diagnosis sub-processes is solved. When the system adjusts the execution order of the guided diagnosis sub-processes according to the time sequence of the user input information, there may be a case that the information required by the subsequent process has not been collected in the previous process. By establishing the identification mechanism of the information dependency relationship, the system can find this potential conflict. Combining the sub-processes that depend on each other into an execution unit and using the execution constraint identifier to mark, the integrity and continuity of information collection are improved. This processing mechanism maintains the flexibility of the guided diagnosis process while ensuring the logical order of information collection, avoiding the missing or collection error caused by the adjustment of the execution order, and improving the reliability and accuracy of the guided diagnosis process.

[0098] The system in the embodiment of the present application will be described from the perspective of hardware processing. Please refer to Figure 3 FIG. 1 is a schematic diagram of the entity device structure of the personalized intelligent guided diagnosis system provided by the embodiment of the present application.

[0099] It should be noted that Figure 3 The structure of the system shown is only an example, and should not bring any limitation to the function and use range of the embodiment of the present application.

[0100] As Figure 3 shown, the system includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 302 or the program loaded from the storage part 308 into the random access memory (RAM) 303, such as performing the method in the above embodiment. In the RAM 303, various programs and data required for system operation are also stored. The CPU 301, the ROM 302, and the RAM 303 are connected to each other through a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0101] The following components are connected to the I / O interface 305: an input section 306 including a camera, a microphone, and the like; an output section 307 including a liquid crystal display (LCD), a speaker, and the like; a storage section 308 including a hard disk and the like; and a communication section 309 including a network interface card such as a LAN (Local Area Network) card, a modem, and the like. The communication section 309 performs a communication process via a network such as the Internet. A drive 310 is also connected to the I / O interface 305 as necessary. A removable medium 311 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, and the like is attached to the drive 310 as necessary, so that a computer program read out therefrom is installed in the storage section 308 as necessary.

[0102] In particular, the processes described above with reference to the flowcharts can be implemented as a computer software program according to embodiments of the present application. For example, embodiments of the present application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing a computer program for executing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network by the communication section 309, and / or installed from the removable medium 311. When the computer program is executed by the central processing unit (CPU) 301, various functions defined in the present application are executed.

[0103] It should be noted that the computer-readable medium in the embodiments of the present application can be a computer-readable signal medium or a computer-readable storage medium or any combination of the two. The computer-readable storage medium may, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or apparatus, or any combination of the above. More specific examples of the computer-readable storage medium can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (Compact Disc Read-Only Memory, CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device or apparatus. In the present application, the computer-readable signal medium can include a data signal carrying computer-readable computer programs in a baseband or as a part of a carrier wave. Such a propagated data signal can take various forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above.

[0104] The flowcharts and block diagrams in the drawings illustrate the possible implementation architectures, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. In the flowcharts or block diagrams, each block can represent a module, a program segment or a part of code containing one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur in different orders than that shown in the drawings. For example, two blocks represented in succession can actually be executed substantially in parallel, and sometimes in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams or flowcharts, and the combination of blocks in the block diagrams or flowcharts, can be implemented by a dedicated hardware-based system that performs the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0105] As another aspect, the present application also provides a computer readable storage medium, which can be included in the system described in the above embodiments, or can exist independently without being assembled into the system. The above storage medium carries one or more computer programs, which, when executed by a processor of a system, enable the system to implement the method provided in the above embodiments.

[0106] The above embodiments are only used to illustrate the technical solutions of the present application, but not limit the present application; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

[0107] In the above embodiments, according to the context, the term "when" can be interpreted as "if" or "after" or "in response to determining" or "in response to detecting". Similarly, according to the context, the phrase "upon determining" or "if detecting (the stated condition or event)" can be interpreted as "if determining" or "in response to determining" or "upon detecting (the stated condition or event)" or "in response to detecting (the stated condition or event)".

[0108] In the above embodiments, all or part of the embodiments can be realized by software, hardware, firmware or any combination thereof. When realized by software, all or part of the embodiments can be realized in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center through wired (such as coaxial cable, optical fiber, digital subscriber line) or wireless (such as infrared, wireless, microwave, etc.) mode. The computer readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. integrated with one or more available media. The available media can be magnetic media (for example, floppy disk, hard disk, magnetic tape), optical media (for example, DVD), or semiconductor media (for example, solid state disk) and the like.

[0109] Those skilled in the art can understand that all or part of the processes in the above-mentioned method embodiments can be implemented by a computer program instructing relevant hardware to complete, the program can be stored in a computer readable storage medium, and the program can include the processes of the above-mentioned method embodiments when executed. The aforementioned storage medium includes ROM or random storage memory RAM, magnetic disc or optical disc and various storage code medium.< / inputinfo>

Claims

1. A personalized intelligent guiding method, characterized in that, Comprise: In the case where it is determined according to the user's sent guidance request that there is historical operation behavior data corresponding to the user, calculate operation proficiency according to the historical operation behavior data, wherein the operation proficiency is calculated according to operation step completion time and operation error times in the historical operation behavior data through a pre-designed calculation function; Divide the guidance process into several guidance sub-processes based on the operation proficiency; Execute each guidance sub-process in turn and obtain the user's input information in the current guidance sub-process; Store the correspondence between the input information and the current guidance sub-process in a guidance information pool; In the case where it is detected that the user exits the current guidance sub-process, generate a guidance progress identifier; In response to detecting that the user resends a guidance request, determine the unfinished guidance sub-process according to the guidance progress identifier, and call the stored input information in the guidance information pool in the unfinished guidance sub-process; Generate a guidance result according to the input information of all guidance sub-processes stored in the guidance information pool.

2. The method of claim 1, wherein, The calculation of operation proficiency according to the historical operation behavior data specifically comprises: Extract operation timing features corresponding to the operation step completion time from the historical operation behavior data, wherein the operation timing features include time intervals between consecutive operations and duration of a single operation; Construct a time dimension score based on the operation timing features, wherein the time dimension score is negatively correlated with the time intervals and the duration by a pre-set value; Extract an operation error ratio corresponding to the operation error times from the historical operation behavior data, wherein the operation error ratio is the ratio of the operation error times to the total operation times; Input the time dimension score and the operation error times into a pre-designed calculation function to obtain the operation proficiency.

3. The method of claim 1, wherein, The pre-designed calculation function is: S = α × (1 - β × T) + (1 - α) × (1 - E) In the above function, S is the operation proficiency, α is the time dimension weight coefficient, β is the time dimension adjustment coefficient, T is the time dimension score, and E is the operation error ratio.

4. The method of claim 1, wherein, After generating a guidance result according to the input information of all guidance sub-processes stored in the guidance information pool, the method further comprises: Construct an associated index table of guidance information items according to information items in each guidance sub-process and their corresponding feature words; Identify feature words from the user's input information; Find corresponding information items and guidance sub-processes in the associated index table according to the feature words; When the input information corresponds to a guidance sub-process that is not the currently executed guidance sub-process, temporarily store the input information in a to-be-allocated information pool; Determine the target guidance sub-process corresponding to the input information; When the target guidance sub-process is executed, fill the input information in the to-be-allocated information pool into the corresponding information items; and According to the order of the time points, adjust the execution order of the guidance sub-processes.

5. The method of claim 4, wherein, The identification of feature words from the user's input information specifically comprises: Extract symptom description phrases from the input information; remove the adverbs and adjectives in the symptom description phrase to obtain a filtered symptom description phrase; determine a feature word in the filtered symptom description phrase as a word having a similarity greater than a preset similarity threshold to a preset standard symptom word library.

6. The method of claim 4, wherein, After the execution order of the guided diagnosis sub-processes is adjusted according to the sequence of the time points, the method further comprises: identifying an information dependency relationship between adjacent guided diagnosis sub-processes, the information dependency relationship representing whether information in a preceding guided diagnosis sub-process is a necessary condition for information in a subsequent guided diagnosis sub-process; combining guided diagnosis sub-processes having the information dependency relationship into an execution unit when it is determined that the information dependency relationship does not conform to the adjusted execution order; generating an execution constraint identifier for marking the execution order of each guided diagnosis sub-process in the execution unit; controlling the adjusted execution order according to the execution constraint identifier.

7. The method of claim 6, wherein, The identifying of the information dependency relationship between adjacent guided diagnosis sub-processes specifically comprises: obtaining a filling rule of each information item in the guided diagnosis sub-processes; determining a guided diagnosis sub-process in which an information item referenced in the filling rule of the information item is located as a preceding guided diagnosis sub-process when it is determined that the filling rule of the information item contains a reference to any information item; determining a guided diagnosis sub-process in which the referenced information item is located as a subsequent guided diagnosis sub-process; establishing an information dependency relationship between the preceding guided diagnosis sub-process and the subsequent guided diagnosis sub-process.

8. A personalized smart guidance system characterized in that, The system comprises: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code comprises computer instructions, and the one or more processors invoke the computer instructions to enable the system to perform the method according to any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that, When the instructions run on the system, the system is enabled to perform the method according to any one of claims 1-7.

10. A computer program product, characterised in that, When the computer program product runs on the system, the system is enabled to perform the method according to any one of claims 1-7.

Images