WO2022160596A1 - 一种问诊信息处理方法、装置和介质 - Google Patents

一种问诊信息处理方法、装置和介质 Download PDF

Info

Publication number
WO2022160596A1
WO2022160596A1 PCT/CN2021/103667 CN2021103667W WO2022160596A1 WO 2022160596 A1 WO2022160596 A1 WO 2022160596A1 CN 2021103667 W CN2021103667 W CN 2021103667W WO 2022160596 A1 WO2022160596 A1 WO 2022160596A1
Authority
WO
WIPO (PCT)
Prior art keywords
disease
entity
feature
question
user
Prior art date
Application number
PCT/CN2021/103667
Other languages
English (en)
French (fr)
Inventor
何峻青
Original Assignee
北京搜狗科技发展有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北京搜狗科技发展有限公司 filed Critical 北京搜狗科技发展有限公司
Publication of WO2022160596A1 publication Critical patent/WO2022160596A1/zh
Priority to US18/137,960 priority Critical patent/US20230268073A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
    • G16H10/20—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for electronic clinical trials or questionnaires
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H80/00—ICT specially adapted for facilitating communication between medical practitioners or patients, e.g. for collaborative diagnosis, therapy or health monitoring
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H70/00—ICT specially adapted for the handling or processing of medical references
    • G16H70/60—ICT specially adapted for the handling or processing of medical references relating to pathologies

Definitions

  • the embodiments of the present application relate to the field of medical technology, and in particular, to a method, device, and medium for processing medical consultation information.
  • Inquiry is a method of diagnosing diseases by making purposeful inquiries to patients or substitute doctors to understand the occurrence, development, diagnosis and treatment process, current symptoms and other conditions related to diseases.
  • artificial intelligence technology With the continuous development of artificial intelligence technology, the way of consultation based on artificial intelligence has gradually developed, bringing a lot of convenience to users' lives.
  • a question-and-answer pair is usually preset; in this way, when the consultation data input by the user is received, the target answer data matching the consultation data is queried from the pre-configured question-and-answer pair, and Feedback the queried target answer data to the user.
  • the current Inquiry information processing method has the problem of low inquiry efficiency.
  • Embodiments of the present application provide a method, device, and medium for processing consultation information, which can improve the accuracy of target questions used for consultation, thereby improving the efficiency of consultation.
  • the embodiment of the present application discloses a method for processing medical inquiry information, including:
  • the corresponding problem entity is obtained from the knowledge graph; the problem entity is used to represent the problem related to the disease feature entity;
  • a target question is generated; the target question is used to inquire the user.
  • an embodiment of the present application discloses a device for processing medical consultation information, including:
  • a user disease characteristic determination module configured to determine the user's disease characteristic according to at least one user input
  • a user disease feature processing module configured to perform disease prediction processing on the user disease feature to obtain a corresponding candidate disease
  • the problem entity acquisition module is configured to obtain the corresponding problem entity from the knowledge graph according to the disease feature entity corresponding to the disease feature of the candidate disease; the problem entity is used to represent the problem related to the disease feature entity;
  • the first question generating module is configured to generate a target question according to the question corresponding to the question entity; the target question is used to inquire the user.
  • an embodiment of the present application discloses an apparatus for processing consultation information, comprising a memory, and one or more programs, wherein one or more programs are stored in the memory, and the programs are stored by a Alternatively, when executed by more than one processor, the steps of the foregoing method are implemented.
  • the embodiments of the present application disclose a machine-readable medium having instructions stored thereon, which, when executed by one or more processors, cause an apparatus to perform the processing of medical consultation information as described in one or more of the foregoing method.
  • the disease prediction process is dynamically performed, and the target question is dynamically generated.
  • the target question for the consultation is automatically generated by processing the consultation information, so the efficiency of the consultation can be improved.
  • the disease prediction processing and the generation of target questions on the consultation information can be a dynamic process; therefore, according to the accumulation of the user's disease characteristics during the consultation process, a more relevant disease characteristic of the user can be obtained. Therefore, the rationality of the inquiry based on the target problem can be improved; and, according to the accumulation of the user's disease characteristics during the inquiry process, a candidate disease that better matches the user's disease characteristics can be obtained, so it can improve the generation and use for disease prediction.
  • the accuracy of the target question to be processed in addition, the embodiment of the present application can dynamically determine whether to stop the inquiry information processing in advance according to the predicted disease characteristics, user disease characteristics, the number of inquiry rounds and other information, so as to improve the efficiency of inquiry and improve the user experience.
  • FIG. 1 is a schematic diagram of an application environment of a method for processing medical consultation information according to an embodiment of the present application
  • FIG. 2 is a flow chart of the steps of Embodiment 1 of a method for processing consultation information according to the present application;
  • FIG. 3 is a schematic diagram of a disease entity and its attributes according to an embodiment of the present application.
  • FIG. 4 is a schematic diagram of a disease characteristic entity and its attributes according to an embodiment of the present application.
  • FIG. 5 is a flow chart of steps of Embodiment 1 of a method for processing a knowledge graph of the present application
  • FIG. 6 is a flow chart of the steps of Embodiment 4 of a method for processing consultation information according to the present application;
  • FIG. 7 is a flow chart of the steps of Embodiment 6 of a method for processing consultation information of the present application.
  • FIG. 8 is a flow chart of the steps of Embodiment 7 of a method for processing consultation information according to the present application.
  • FIG. 9 is a flow chart of the steps of Embodiment 8 of a method for processing consultation information of the present application.
  • FIG. 10 is a structural block diagram of an embodiment of an apparatus for processing medical consultation information according to the present application.
  • FIG. 11 is a block diagram of an apparatus 1100 for processing consultation information according to the present application.
  • FIG. 12 is a schematic structural diagram of a server in some embodiments of the present application.
  • the embodiment of the present application provides a method for processing consultation information.
  • the method may include: determining a user's disease characteristics according to at least one user input; performing disease prediction on the above-mentioned user's disease characteristics processing to obtain the corresponding candidate diseases; according to the disease characteristics corresponding to the above-mentioned candidate diseases, a target question is generated, and the above-mentioned target question is used to inquire the user.
  • At least one user input of the user may include: disease characteristics that appear in the user (hereinafter referred to as user disease characteristics).
  • user disease characteristics disease characteristics that appear in the user
  • This embodiment of the present application may determine the user's disease characteristics based on the consultation, and process the user's disease characteristics.
  • Types of disease characteristics can include: symptoms, triggers, peak seasons, exposure history, family history, etc.
  • the disease prediction process is dynamically performed based on the user's disease characteristics obtained by at least one user input by processing the consultation information, and the target question is dynamically generated.
  • the target question for the consultation is automatically generated by processing the consultation information, so the efficiency of the consultation can be improved.
  • the disease prediction processing and the generation of target questions on the consultation information can be a dynamic process; therefore, according to the accumulation of the user's disease characteristics during the consultation process, a more relevant disease characteristic of the user can be obtained. Therefore, it is possible to improve the rationality of the information processing of the consultation; and, according to the accumulation of the user's disease characteristics during the consultation process, a candidate disease that better matches the user's disease characteristics can be obtained, so the generation of the target for disease prediction processing can be improved. accuracy of the question.
  • the method for processing consultation information provided in the embodiments of the present application may be applied to, for example, application scenarios of websites and/or APPs (application programs, Application).
  • application scenarios of the embodiments of the present application may include: medical-related websites, or medical-related APP, etc.
  • the method for processing diagnosis information provided in this embodiment of the present application can be applied to the application environment shown in FIG. 1 .
  • the client 100 and the server 200 are located in a wired or wireless network, and through the wired or wireless network, The client 100 and the server 200 perform data interaction.
  • the client 100 can run on a terminal, and the above-mentioned terminal specifically includes but is not limited to: a smart phone, a tablet computer, an e-book reader, MP3 (moving image expert compression standard audio layer 3, Moving Picture Experts Group Audio Layer III) ) players, MP4 (Moving Picture Experts Group Audio Layer IV) players, laptop computers, car computers, desktop computers, set-top boxes, smart TVs, wearable devices, etc.
  • a smart phone a tablet computer
  • an e-book reader MP3 (moving image expert compression standard audio layer 3, Moving Picture Experts Group Audio Layer III) ) players
  • MP4 Motion Picture Experts Group Audio Layer IV
  • the client 100 can interact with the user. Specifically, the client 100 may receive at least one user input and provide the user with the target question.
  • the client 100 may generate the target question by using the method for processing the diagnosis information in the embodiment of the present application.
  • the client 100 may send at least one user input of the user to the server 200, so that the server 200 generates the target question for the consultation by using the method for processing consultation information in the embodiment of the present application.
  • Embodiment 1 of a method for processing medical inquiry information of the present application is shown, which may specifically include the following steps:
  • Step 201 Determine the disease characteristics of the user according to at least one user input
  • Step 202 performing disease prediction processing on the above-mentioned user disease characteristics to obtain corresponding candidate diseases
  • Step 203 generating a target question according to the disease characteristic corresponding to the candidate disease, and the above target question is used to inquire the user.
  • At least one step of the method embodiment shown in FIG. 2 may be executed by a client and/or a server.
  • the embodiment of the present application does not limit the specific execution subject of each step.
  • step 201 at least one user input may be received through input methods such as keyboard input, option selection, and voice input.
  • the above-mentioned at least one user input may include:
  • the reply may include text input, selection of answer options, etc., and the form of the reply is not limited.
  • the preset questions may be pre-stored questions relative to the target questions dynamically generated according to at least one user input.
  • the user's first input is usually an active input.
  • the active input usually includes: chief complaint.
  • the chief complaint is used to characterize the main symptoms and/or signs of the patient or the representative. at least one of.
  • a preset question may be provided to the user to obtain the user's response to the preset question.
  • the preset question may be a question whose frequency exceeds the frequency threshold during the consultation.
  • the keywords of the preset question may include: duration, mental state, etc., so as to inquire about the duration of symptoms and the mental state of the patient. It can be understood that the embodiments of the present application do not limit specific preset problems.
  • the target question specifically includes: question text and answer options; the above-mentioned at least one user input specifically includes: an answer option selected by the user.
  • question text might include: “Have you vomited without nausea?”, “Which of the following symptoms?”, “Which of the following illnesses have you had before?, “What type of rash? “What is the shape of the stool?” et al. Answer options are used to characterize the answer options that can be selected.
  • the answer options can include: [Yes, No, I don't know], etc.
  • the answer options can include: answer options corresponding to disease characteristics.
  • answer options could include: [Symptom 1, Symptom 2...Symptom N, none of the above].
  • the user's disease characteristics may be determined from at least one user input.
  • the method for determining the user's disease characteristics may include, but is not limited to, an entity recognition method, or a method for matching disease characteristic tables, and the like.
  • the determined user disease characteristics may be saved to a user disease characteristic set.
  • At least one user input may include a standard description corresponding to the user's disease characteristics.
  • at least one user input may include non-standard descriptions corresponding to the user's disease characteristics, such as colloquial descriptions.
  • the non-standard descriptions in at least one user input may be converted into standard descriptions. Therefore, the embodiments of the present application will use the standard description of the user's disease characteristics to perform disease prediction processing on the standardized user's disease characteristics, so as to improve the accuracy of disease prediction.
  • disease characteristics corresponding to diseases may be determined according to medical resources such as medical books, medical databases, and medical question-and-answer data; or, disease characteristics corresponding to diseases may be determined by constructing a knowledge map.
  • medical resources such as medical books, medical databases, and medical question-and-answer data
  • disease characteristics corresponding to diseases may be determined by constructing a knowledge map.
  • the relevant content of the knowledge graph will be introduced in the following embodiments.
  • the disease prediction process may be used to determine the probability of candidate diseases corresponding to the user's disease characteristics.
  • the candidate disease may be at least one, and the candidate disease may correspond to a score, and the score may represent the probability of the candidate disease under the condition of the user's disease characteristics.
  • the above-mentioned disease prediction processing on the above-mentioned user's disease characteristics may specifically include: determining a candidate disease corresponding to the above-mentioned user's disease characteristics according to matching information between the above-mentioned user's disease characteristics and disease characteristics of the disease. For example, if the user's disease characteristics match the disease characteristics of disease A, disease A may be used as a candidate disease corresponding to the user's disease characteristics.
  • the disease prediction processing performed on the user's disease characteristics may be a dynamic process.
  • disease prediction processing may be performed on the updated user's disease feature set. In this way, a candidate disease that better matches the user's disease characteristics can be obtained according to the accumulation of the user's disease characteristics during the consultation process, so that the accuracy of the target question generated for the consultation can be improved.
  • the target question can be used to inquire about the disease characteristics corresponding to the candidate diseases, so as to help the user determine whether the corresponding disease characteristics appear.
  • the target question may include: question text and answer options.
  • the above answer options correspond to disease characteristics corresponding to the candidate diseases.
  • the embodiment of the present application can convert the answer option selected by the user into the corresponding user disease feature.
  • the target question A is "Which of the following symptoms?"
  • the answer options can include: [symptom 1, symptom 2...symptom N, none of the above], assuming the user selects the answer option [symptom 1], the "symptom 1" can be 1" is determined to be the user's disease characteristics; assuming that the user selects the answer option [none of the above], the target question A's inquiry does not hit the user's disease characteristics.
  • the embodiments of the present application may also determine and record disease characteristics that do not appear in the user according to the user's selection operation for answer options, so as to carry the disease characteristics that do not appear in the user in the inquiry information processing result.
  • a target disease feature can be determined from disease features corresponding to a candidate disease, and a target question can be generated for the target disease feature.
  • the target disease feature can be used to characterize the disease feature asked by the user in this round of consultation.
  • the user disease feature determined according to the user input may be removed from the disease feature corresponding to the candidate disease to obtain the target disease feature.
  • the target disease feature may be determined from the disease features corresponding to the candidate disease according to the importance score corresponding to the disease feature.
  • the factor features of the above importance score may specifically include at least one of the following features:
  • the target disease feature may be determined from the disease features corresponding to the candidate disease according to the correlation between the disease feature corresponding to the candidate disease and the user disease feature determined according to the user input.
  • a target disease feature whose correlation is greater than the first threshold may be determined from the disease features corresponding to the candidate diseases.
  • the correlation between the disease features can be determined according to the co-occurrence information of multiple disease features in the medical resources of one disease. It can be understood that the embodiments of the present application do not limit the specific manner of determining the correlation between disease characteristics.
  • a preset question and a target question corresponding to the user may be saved in a question queue, an inquiry question is taken from the question queue, and the inquiry question is output to the user.
  • Questions in the question queue can have corresponding priorities, and questions can be taken from the question queue according to the priority of the questions.
  • the factors for determining the priority of the questions may include: the time when the questions are queued, and/or the degree of matching between the disease characteristics corresponding to the questions and the user's disease characteristics, and the like. It can be understood that the embodiment of the present application does not limit the specific process of taking out a question from the question queue.
  • the execution body in this embodiment of the present application may be a server or a client.
  • the server may be a processing engine or an interaction engine, wherein the interaction engine may be used to interact with the client, and the processing engine may communicate with the client.
  • the interaction engine sends the consultation questions so that the interaction engine provides the consultation questions to the user. It can be understood that the embodiments of the present application do not limit the specific implementation manner of providing the user with the inquiry question.
  • step 202 after each execution of step 202, it can be determined whether to end the consultation information processing according to the score of the candidate disease, the user's disease characteristics, the number of inquiry rounds and other information; Output the processing result of the consultation information, otherwise, continue to step 203 .
  • the embodiment of the present application dynamically determines whether to stop the consultation in advance based on the predicted disease characteristics, user disease characteristics, the number of inquiry rounds and other information, so as to improve the efficiency of the inquiry and thus improve the user experience.
  • disease prediction processing is performed dynamically according to the user's disease characteristics obtained based on at least one user input, and target questions are dynamically generated. Since the embodiment of the present application automatically generates the target question for the consultation in the consultation process, the consultation efficiency can be improved.
  • the disease prediction processing and the generation of target questions on the user's disease characteristics can be a dynamic process; therefore, according to the accumulation of the user's disease characteristics during the consultation process, a more relevant user's disease characteristics can be obtained. Therefore, the rationality of the inquiry based on the target problem can be improved; and, according to the accumulation of the user's disease characteristics during the inquiry process, a candidate disease that better matches the user's disease characteristics can be obtained, so the generated disease can be improved. The accuracy of the predicted target problem.
  • This embodiment is used to illustrate the knowledge graph.
  • the knowledge graph is a structured semantic knowledge base for describing concepts and their interrelationships in the physical world.
  • an entity refers to an objectively existing and mutually distinguishable thing, including a specific person, thing, thing, abstract concept or connection, and the like.
  • An entity can be a concrete object, such as a disease, a disease feature, etc.; it can also be an abstract event, such as an inquiry for a disease feature.
  • Entities can have many properties, individual properties are called properties. Each attribute has a value range, and its type can be integer, real, or string, etc.
  • the naming unit of a tag attribute is called a field.
  • the state of the field may include a filled state or an unfilled state, wherein the filled state corresponds to the filled field content, and the unfilled state indicates that the corresponding field content is to be filled.
  • Medical entities may include: disease entities, disease characteristic entities, or problem entities, and the like.
  • Disease entities can characterize specific diseases, such as “hypertension”, “leukemia”, and the like.
  • a disease can correspond to a diseased system.
  • the disease system can correspond to the system in the anatomical sense.
  • the disease system can include: the motor system, the digestive system, the respiratory system, the urinary system, the reproductive system, the endocrine system, the immune system, the nervous system, and the circulatory system.
  • the attributes of the disease entity may include at least one of the following attributes:
  • Disease identification attribute disease system attribute, feature set attribute, clinical proportion attribute, and high incidence age attribute;
  • the feature set may include: disease features associated with the disease feature entity;
  • the clinical proportion is used to characterize the incidence probability of the disease in the disease system, and can be obtained according to the incidence number of the disease and the incidence number of the disease system.
  • the attributes of the disease entity may include: disease identification attribute, disease system attribute, feature set attribute, clinical proportion attribute, and high incidence age attribute.
  • a single attribute can have attribute parameters corresponding to it.
  • the attribute parameters of the disease system attribute include: system probability, which can characterize patients of a single disease system, and the proportion of patients in all disease systems, which can be obtained according to the ratio of patients in a single disease system to patients in all disease systems.
  • the attribute parameter of the feature set attribute may include at least one of the following parameters:
  • conditional probability of the disease feature under the condition of the disease usually includes multiple disease features, and the conditional probability can be the conditional probability corresponding to the multiple disease features under the condition of the disease;
  • the penalty factor of the disease feature under the condition of the disease can correspond to the disease feature that cannot appear under the condition of the disease, and is used to penalize the probability of the disease in the process of disease prediction.
  • Disease signature entities may characterize specific disease signatures. Types of disease characteristics can include: symptoms, triggers, peak seasons, exposure history, family history, etc.
  • the attributes of the disease feature entity may include: a hit action attribute, where the hit action attribute is used to represent the information of the problem entity that is triggered when the corresponding disease feature entity is selected.
  • the attributes of the disease feature entity may include: an affiliation attribute, which is used to represent the disease feature entity that has a parent-child relationship with the corresponding disease feature entity.
  • the attribute parameters of the affiliation attribute can include: a parent disease feature or a child disease feature.
  • the sub-disease features of the disease feature "vomiting” include: “projectile vomiting”; for another example, the sub-disease features of the disease feature "fever” include: “low fever”, “high fever” and so on.
  • the attributes of the disease feature entity may include: feature identification attribute, type attribute, affiliation attribute, frequency attribute, hit action attribute, interpretation attribute, and the like.
  • the frequency attribute can represent the number of occurrences of the corresponding disease feature in the feature set of all disease entities.
  • the named action attribute "Identification of Problem Entity 18" of the disease feature "vomiting” represents the problem entity that identifies the triggering problem entity as 18 when the disease feature "vomiting" is selected.
  • the above hit action attribute can improve the rationality of the sequence of questions in the consultation process. For example, if the user selects the "vomiting" symptom, the corresponding problem entity identifier 18 will be found according to the hit action attribute of the "vomiting" symptom, so as to further inquire about the "projective vomiting" symptom.
  • the question entity corresponds to an inquiry for disease characteristics, and is used to represent the question corresponding to an inquiry. Since a consultation may involve at least one disease characteristic, the question corresponding to the question entity may involve at least one disease characteristic.
  • the fields of the question entity may include: a question text field and an answer option field.
  • the question text field is used to characterize the question to be answered. Examples of question text fields might include: “Have you vomited without nausea?”, “Which of the following symptoms?", “Which of the following illnesses have you had before?, “What type of rash? “What is the shape of the stool?” et al.
  • the answer option field is used to characterize the answer options available for selection.
  • the field of the problem entity may further include at least one of the following fields: a disease feature field, a trigger condition field, and a jump relationship field;
  • the disease feature field is used to represent the disease feature entity
  • the trigger condition field is used to represent that the corresponding problem entity is obtained by triggering according to the disease feature entity
  • the jump relationship field is used to execute a preset jump under the condition that the answer option is selected.
  • the jump relationship field is used to jump from the first question entity to the second question entity under the condition that the answer option is selected, and the disease feature entity corresponding to the first question entity corresponds to the second question entity.
  • the disease feature entity of is a parent-child relationship.
  • the preset jump may also include: executing a preset function, and the preset function can be used to end the search of the problem entity.
  • the problem entity in this embodiment of the present application may include: a problem entity instance and/or a problem entity template.
  • Problem entity instances may correspond to preset disease characteristics.
  • the identifier of the problem entity instance is 18, which is obtained by triggering according to the disease feature "vomiting", that is, when the "vomiting" feature is selected, the problem entity instance with the identifier 18 can be triggered.
  • the instance of the problem entity corresponds to the disease feature "projectile vomiting", and is used to query the disease feature "projectile vomiting".
  • the question text may contain a definition of the disease feature "projectile vomiting" to help the user determine whether to hit the corresponding disease feature and select the corresponding answer option.
  • the question text field of the question entity template is in a populated state, and the preset fields other than the question text field of the question entity template are in an unpopulated state.
  • the preset fields may include: an answer option field, a disease characteristic field, a trigger condition field, a jump relationship field, and the like.
  • the problem entity template can correspond to a preset type of disease characteristics.
  • the corresponding problem entity template can be found in the knowledge map according to the type corresponding to the user-related disease characteristics, and the corresponding problem entity template can be obtained according to the user-related diseases.
  • fill in the fields of the problem entity template, and the problem entity template after the fields are filled can be used as a dynamic problem entity instance, and the dynamic problem entity instance can contain the questions used for consultation.
  • the question entity template corresponds to the disease characteristics of the preset type
  • the question entity template after the fields are filled can contain information of multiple preset types of disease characteristics, so the number of disease characteristics contained in the question that can be used for consultation is In turn, the number of interactive rounds of consultation can be reduced, and the efficiency of consultation can be improved.
  • the above-mentioned field filling in the question entity template may specifically include: filling the answer option fields according to the disease characteristics related to the user, and different disease characteristics may correspond to different answer options.
  • the answer option field may be filled with definitions corresponding to user-related disease characteristics, and different definitions may correspond to different answer options.
  • the above-mentioned field filling in the problem entity template may specifically include: filling the jump relationship field according to the hit action attribute corresponding to the disease feature related to the user.
  • the hit action attribute corresponding to the user-related disease feature may be filled in the jump relationship field.
  • the content of the jump relationship field may be: when the answer option is selected, jump to the preset question entity, assuming that the answer option corresponds to the first disease feature entity, and the hit action attribute of the first disease feature entity records
  • the preset problem entity may be: problem entity information corresponding to the second disease feature entity.
  • the information of the first disease feature can be filled in the answer option field of the question entity template, and the question entity information corresponding to the second disease feature entity can be filled in the jump relationship field .
  • the question entity template may correspond to the disease characteristics of the symptom type, and is used to inquire about the disease characteristics of the symptom type.
  • fields of the problem entity template can be filled according to the user-related symptom 1, symptom 2...symptom N (N can be a natural number greater than 0).
  • the explanations of symptoms such as symptom 1, symptom 2...symptom N are filled in the answer option field to help the user determine whether to hit the corresponding disease feature and select the corresponding answer option.
  • the symptom type can be populated in the disease characteristic field. Or, fill in the jump relationship field with the problem entity identifiers corresponding to the sub-symptoms of symptom 1, symptom 2...symptom N, etc.
  • the filling of other preset fields is optional, that is, the filling of the disease feature field, the triggering condition field, and the jump relationship field may not be performed.
  • problem entity template corresponding to the disease characteristics of symptom types shown in Table 5 is only an optional embodiment. In fact, those skilled in the art can also use problem entities corresponding to other types of disease characteristics according to actual application requirements.
  • template For example, a question entity template corresponding to the disease characteristics of the contact history type may also be used, and the corresponding question text may include: "Have you ever been exposed to the following pathogens, harmful factors, and disease patients?" and so on.
  • FIG. 5 a flowchart of steps of a method for processing a knowledge graph according to an embodiment of the present application is shown, which may specifically include the following steps:
  • Step 501 Determine the problem entity according to the disease characteristic entity; the above-mentioned problem entity is used to represent the problem related to the above-mentioned disease characteristic entity;
  • Step 502 in the knowledge graph, establish the association between the above-mentioned disease feature entity and the above-mentioned problem entity.
  • the disease feature entity may represent features related to the disease, which may include: disease features that appear in the disease, or disease features that cannot appear in the disease.
  • the process of determining a disease characteristic entity may include: determining a main complaint list, and determining a disease list corresponding to the main complaint list; and performing a disease diagnosis on the diseases in the disease list according to medical resources.
  • the expansion of the feature; the disease feature entity is determined according to the chief complaint in the chief complaint list and the filled disease feature.
  • the chief complaint is used to characterize the main symptoms and/or signs of the patient or the representative. at least one of.
  • the chief complaint may be obtained from medical resources such as medical query data and/or medical record data, and a list of chief complaints may be established according to the obtained chief complaint.
  • a corresponding disease may be determined according to a single main complaint in the main complaint list, and then the determined disease may be added to the disease list.
  • An implementation manner may be that the main complaint is sent to the doctor terminal, and the user corresponding to the doctor terminal determines the disease corresponding to the main complaint.
  • the user of the doctor terminal may be a doctor with more than M (M may be a natural number greater than 0, for example, M may be greater than 7) years of clinical experience, who can determine the disease corresponding to the chief complaint based on knowledge and experience.
  • Diseases in the disease list can be used as data sources for disease entities in the knowledge graph. That is, corresponding disease entities can be constructed according to the diseases in the disease list.
  • the embodiments of the present application may expand disease characteristics for the diseases in the disease list based on medical resources such as medical books, medical databases, and medical question and answer data. That is, for a disease, on the basis of the main complaint corresponding to the disease, the disease characteristics other than the main complaint are expanded.
  • the types of disease characteristics involved in the expansion may include: symptoms, predisposing factors, peak seasons, exposure history, family history, etc.
  • disease content corresponding to the disease may be acquired from medical resources, and disease characteristics corresponding to the above types are extracted from the disease content.
  • the chief complaint in the above-mentioned list of chief complaints and the filled disease features can be used as the data source of the disease feature corresponding to the disease feature entity. That is, the disease feature entity can also be constructed according to the chief complaint in the chief complaint list and the filled disease features.
  • candidate disease features main complaint and filled-in disease features
  • candidate disease features may also be sent to the doctor terminal, so that the user of the doctor terminal can update the candidate disease features.
  • the above-mentioned updating of candidate disease features may specifically include: addition of candidate disease features, deletion of candidate disease features, or modification of candidate disease features, and the like.
  • the updated candidate disease features can be used as a data source of disease features corresponding to disease feature entities. For example, a feature set corresponding to a disease may be determined for the updated candidate disease features corresponding to the disease.
  • conditional probability of the disease feature in the feature set under the condition of the disease and/or the penalty factor of the disease feature under the condition of the disease may also be determined for the disease.
  • conditional probability or the penalty factor may be determined according to the occurrence information of the disease feature in the medical resource corresponding to the disease; or, the conditional probability or the penalty factor may be determined by the user of the doctor terminal.
  • the incidence probability (clinical proportion) and/or system probability of the disease in the disease system may also be determined according to the disease and the medical resources of the disease system to which the disease belongs.
  • conditional probabilities can represent the degree of matching between disease features and diseases or the importance of disease features to diseases. Therefore, applying conditional probability to disease prediction processing can improve many candidate diseases when the user's disease features correspond to multiple candidate diseases. Discrimination of candidate diseases.
  • the above penalty factor can represent the degree of exclusion of the disease feature on the disease, and further can comprehensively determine the influence of multiple disease features on the disease. For example, if a user has a feature that should not appear in a candidate disease, the probability of the candidate disease can be reduced according to the penalty factor. The accuracy of the probability of the disease; further, in the case that the user's disease characteristics correspond to multiple candidate diseases, the degree of discrimination between multiple candidate diseases can be improved.
  • the clinical proportion can represent the incidence probability of the disease in the disease system, and can reflect the commonness of the corresponding disease.
  • Applying clinical proportions to disease prediction processing can improve the accuracy of the probability of candidate diseases under the condition of the user's disease characteristics. For example, in the case that the user's disease characteristics correspond to multiple candidate diseases, the multiple candidate diseases can be sorted according to the clinical proportions corresponding to the multiple candidate diseases; then, in the case that the user's disease characteristics correspond to multiple candidate diseases , to improve the discrimination between multiple candidate diseases.
  • the prior probability of candidate diseases can be determined according to the clinical proportion and systematic probability, and then various candidate diseases can be ranked according to the prior probability. In this way, the accuracy of the probability of a candidate disease under the condition of the user's disease characteristics can be improved; further, when the user's disease characteristics correspond to multiple candidate diseases, the degree of discrimination between multiple candidate diseases can be improved.
  • the above-mentioned determining the disease feature entity may specifically include: performing feature normalization on the main complaint and the disease features obtained by filling, so as to obtain the normalized disease feature; After normalizing the disease features, the disease feature entity is determined.
  • individualized or colloquial symptom descriptions corresponding to “headache” specifically include: “pain like a needle stick”, “pain with a twitch”, “pain when touched”, “throat and saliva hurts”, etc.
  • personalized or colloquial symptom description corresponding to “tongue pain” specifically includes: “pain on the left side of the tongue”, “pain at the tip of the tongue”, “pain at the root of the tongue”, “pain at the edge of the tongue”, etc.
  • determining a disease characteristic entity may specifically include: determining multiple attributes corresponding to the disease characteristic entity, and determining corresponding attribute values for a specific disease characteristic entity.
  • the multiple attributes corresponding to the disease feature entity may specifically include: feature identification attribute, type attribute, affiliation attribute, frequency attribute, hit action attribute, interpretation attribute, and the like.
  • the problem entity in the embodiment of the present application is used to inquire about the disease symptoms corresponding to the disease feature entity, so as to help the user determine whether the corresponding disease symptoms occur.
  • the information of the disease feature entity may be sent to the doctor terminal, so that the doctor can set the problem entity corresponding to the disease feature entity.
  • the problem entity corresponding to the disease characteristic entity may be determined according to the type corresponding to the disease characteristic entity and historical consultation data.
  • the fields of the question entity may include: a question text field and an answer option field, wherein the question text field or the answer option field may include information such as identification or interpretation of the disease feature entity.
  • the field of the problem entity may further include at least one of the following fields: a disease feature field, a trigger condition field, and a jump relationship field;
  • the disease feature field is used to represent the disease feature entity
  • the trigger condition field is used to represent that the corresponding problem entity is triggered and obtained according to the disease feature entity
  • the jump relationship field is used to execute a preset jump under the condition that the answer option is selected.
  • the jump relationship field is used to jump from the first question entity to the second question entity under the condition that the answer option is selected, and the disease feature entity corresponding to the first question entity corresponds to the second question entity.
  • the disease feature entity of is a parent-child relationship.
  • the preset jump may also include: executing a preset function, and the preset function can be used to end the search of the problem entity.
  • the information of the problem entity can also be sent to the doctor terminal, so that the user of the doctor terminal can check the information of the problem entity.
  • the problem entity may include: a problem entity instance, and the problem entity instance may correspond to a preset disease feature. Therefore, in this embodiment of the present application, a preset association between disease feature entities and problem entity instances can be established in the knowledge graph.
  • the problem entity in the embodiment of the present application may include: a problem entity template, and the problem entity template may correspond to a preset type of disease characteristics. Therefore, the embodiment of the present application can establish an association between a preset type of disease feature entity and a problem entity instance in the knowledge graph.
  • establishing the association between the disease feature entity and the problem entity may specifically include: establishing a mapping relationship between the disease feature entity and the problem entity according to the disease feature field in the problem entity, wherein the Disease signature fields are matched against disease signature entities.
  • the problem entity in the mapping relationship is an instance of the problem entity, and the disease feature entity in the mapping relationship corresponds to a preset disease feature.
  • the problem entity in the mapping relationship is the problem entity template, and the disease feature entity in the mapping relationship corresponds to multiple preset types of disease features.
  • the association between the disease entity and the disease feature entity may also be established in the knowledge graph.
  • the disease entity can be associated with the disease feature entity corresponding to the disease feature in the feature set, so that the user's disease feature can be matched with the feature set corresponding to the disease. It can be understood that the embodiments of the present application do not limit the specific association manner between the disease entity and the disease feature entity.
  • the method for processing the knowledge graph in the embodiments of the present application establishes the association between the disease feature entity and the problem entity in the knowledge graph.
  • the corresponding problem entity can be searched in the knowledge graph according to the disease characteristics related to the user, and the question for the consultation can be obtained according to the obtained problem entity.
  • the knowledge graph in the embodiment of the present application includes the association between the disease feature entity and the problem entity, the questions for the consultation can be generated in the consultation process, so the effect of the knowledge graph on the consultation can be improved, and the basis for the diagnosis can be improved.
  • the efficiency and accuracy of the knowledge graph for the processing of medical consultation information is the efficiency and accuracy of the knowledge graph for the processing of medical consultation information.
  • the embodiments of the present application represent the relationship between the disease feature and the information of the problem entity through the hit action attribute in the disease feature entity.
  • the corresponding disease feature When the corresponding disease feature is selected, the corresponding problem entity will be triggered. Since the association between disease characteristics and problem entities can be automatically performed, it can not only reduce the resource cost of manual operation of the consultation path, but also improve the rationality of the problem sequence in the consultation process.
  • the jump relationship field is used to represent the question entity: when the answer option corresponding to the disease feature is selected, it will jump to the preset question entity, which can improve the sequence of questions in the consultation process.
  • rationality Since the association between disease characteristics and problem entities can be automatically performed, it can not only reduce the resource cost of manual operation of the consultation path, but also improve the rationality of the problem sequence in the consultation process.
  • attribute parameters such as conditional probability, penalty factor, clinical proportion, and system probability are set in the disease entity.
  • conditional probability can represent the degree of matching between disease features and the disease or the importance of the disease feature to the disease
  • penalty factor can represent the exclusion degree of the disease feature to the disease, and then can comprehensively determine the impact of multiple disease features on the disease
  • clinical The proportion can represent the incidence probability of a disease in the disease system, and can reflect the commonness of the corresponding disease
  • applying one or more of the above attribute parameters to disease prediction processing can reflect the probability of the corresponding candidate disease, so it can improve the The accuracy of the probability of a candidate disease conditioned on the user's disease characteristics.
  • This embodiment describes the disease prediction processing in the process of processing the consultation information.
  • a search can be performed in the association relationship between the disease entity and the disease feature entity included in the knowledge graph according to the user's disease feature obtained based on the user's input, so as to obtain a candidate disease.
  • the disease feature of the user may be matched with the disease feature entity in the association relationship, and the disease corresponding to the successfully matched disease entity may be used as the candidate disease.
  • the user's disease characteristics may be matched with at least one candidate disease.
  • the probability of the candidate disease may be characterized according to the score of the candidate disease, and the candidate disease may be screened according to the score of the candidate disease.
  • the score of the candidate disease may be determined according to the probability feature
  • the above-mentioned probability feature may be a feature recorded in the knowledge graph, and may specifically include at least one of the following features:
  • the score of the candidate disease may be determined according to the above conditional probability.
  • the user's disease characteristics include two symptoms: "cough” and "sputum production”.
  • “acute laryngitis” and “bronchitis” can match both symptoms, but the probability of "cough” in “acute laryngitis” is 0.6, “phlegm” in “acute laryngitis” is 0.4; and “cough” in “bronchitis” The probability is 0.8 in “sputum” and 0.6 in “bronchitis”.
  • the penalty factor can correspond to disease features that cannot appear under the condition of the candidate disease, and is used to penalize the probability of the candidate disease in the process of disease prediction.
  • the candidate disease includes: candidate disease A, but the user disease features include: disease feature X that should not appear in candidate disease A.
  • the candidate disease can be reduced according to the penalty factor corresponding to disease feature X and candidate disease A. Disease A score.
  • the clinical proportion can represent the incidence probability of the disease in the disease system, and can reflect the commonness of the corresponding disease. Applying clinical proportions to disease prediction processing can improve the accuracy of disease prediction scores. In general, the higher the clinical proportion, the higher the score for the corresponding candidate disease.
  • the prior probability of the candidate disease can be determined according to the clinical proportion and the system probability, and then the score of the candidate disease can be determined according to the prior probability.
  • the higher the prior probability of the candidate disease the higher the score of the candidate disease.
  • the multiple probabilistic features can be fused, and the score of the candidate disease can be determined according to the fused probabilistic features.
  • the corresponding fusion method may include: weighted average method, or product method, and the like.
  • the candidate diseases are screened according to the scores of the candidate diseases, which may specifically include: selecting a candidate disease with a score greater than a score threshold, and/or selecting a top P with a score (P may be a natural number greater than 0) candidate disease.
  • the target question can be generated according to the disease characteristics corresponding to the target candidate disease, or the information of the target candidate disease can be output as the result of the inquiry information processing.
  • the stop condition may represent a condition corresponding to stopping the processing of the consultation information.
  • the output of the target question to the user may be stopped, and the corresponding medical consultation information processing result may also be output to the user.
  • the above method may further include: after obtaining the candidate disease, if the stopping condition is met, stopping executing the step of generating the target problem.
  • the above method may further include: after the candidate disease is obtained, if the stopping condition is met, outputting an inquiry information processing result; the inquiry information processing result may include information of the candidate disease.
  • the information of one or more target candidate diseases such as the name and score of the target candidate disease with the highest score, may be carried in the inquiry information processing result.
  • the above-mentioned processing result of the consultation information may further include: disease characteristics that appear in the user, and disease characteristics that do not appear in the user.
  • the conditional probability of the disease feature appearing in the user under the condition of the target candidate disease can also be carried in the inquiry information processing result.
  • the above-mentioned stop condition may specifically include at least one of the following conditions:
  • the score of at least one candidate disease is greater than the score threshold, which can be used to improve the quality of the target candidate disease carried in the processing result of the inquiry information;
  • Differences in the scores of multiple candidate diseases meet the difference condition can characterize that the scores of at least two candidate diseases are different. In the case that the difference condition is not met, it is indicated that the user disease characteristics correspond to multiple candidate diseases with similar scores. In this case, further user disease characteristics are required to screen multiple candidate diseases with similar scores.
  • the query ratio of the disease feature corresponding to the candidate disease complies with the ratio condition
  • the query ratio can represent the ratio of the queried disease feature relative to all disease features, and the ratio condition can be that the query ratio is greater than the ratio threshold, etc.
  • the consultation questions can be selected from the problem queue, and the selected consultation questions can be output to the user; Disease characteristics, generate target questions, and save the generated target questions to the question queue.
  • Embodiment 4 of a method for processing medical inquiry information of the present application is shown, which may specifically include the following steps:
  • Step 601 for the user, establish a question queue
  • the question queue may include: preset questions
  • Step 602 Determine the disease characteristics of the user according to at least one user input
  • Step 603 performing disease prediction processing on the above-mentioned user disease characteristics to obtain corresponding candidate diseases
  • Step 604 judge whether the stop condition is met, if yes, go to step 608, otherwise go to step 605;
  • Step 605 determine whether there are problems in the problem queue, if so, go to step 606, otherwise go to step 607;
  • Step 606 select the question from the question queue to output the question to the user
  • Step 607 generating a target question according to the disease characteristics corresponding to the candidate diseases, and adding the target question to the question queue;
  • Step 608 output the processing result of the consultation information, and end the consultation.
  • This embodiment describes the process of generating the target question.
  • the embodiment of the present application establishes the association between the disease feature entity and the problem entity in the knowledge graph.
  • the corresponding problem entity can be searched in the knowledge map according to the disease characteristics corresponding to the candidate disease, and the target question for the consultation can be obtained according to the obtained problem entity.
  • the knowledge graph in the embodiment of the present application includes the association between the disease feature entity and the problem entity, the questions for the consultation can be automatically generated in the consultation process, so the effect of the knowledge graph on the consultation can be improved, and the Consultation efficiency and accuracy of the target questions used for the interview.
  • the mapping relationship between the disease feature entity and the problem entity included in the knowledge map can be searched to obtain the problem entity corresponding to the disease feature.
  • the problem entity instance corresponding to the disease feature can be determined according to the mapping relationship between the disease feature entity and the problem entity instance.
  • the content of the question text field and the content of the answer option field may be obtained from the question entity instance corresponding to the target disease feature to obtain the target question; that is, the target question may include: the content of the question text field and the answer option the content of the field.
  • the target disease feature is "vomiting”
  • the target problem A can be determined according to the problem entity instance corresponding to "vomiting”.
  • the question text for target question A can be "whether to vomit”, and the answer option for target question A can be: [Yes, No, I don't know].
  • the corresponding question entity may be obtained from the knowledge map according to the hit action attribute in the disease feature entity corresponding to the target disease feature.
  • the target disease feature is "vomiting”
  • the answer option [Yes] corresponding to the target question A is selected, according to the hit action attribute of the disease feature entity shown in Table 2, the triggering question entity is identified as the question entity of 18 instance.
  • the target question B can be generated according to the question entity instance whose question entity identifier is 18.
  • the question text of target question B can be "Is it vomiting without nausea?”, and the answer option of target question B can be: [Yes, no, I don't know].
  • the problem entity template corresponding to the disease feature can be determined according to the mapping relationship between the type of the disease feature entity and the problem entity template.
  • the above-mentioned generating the target problem may specifically include: filling the fields of the above-mentioned problem entity template according to the above-mentioned disease characteristics, so as to obtain the target problem.
  • the fields of the question entity template specifically include: a question text field and an answer option field; the above-mentioned field filling in the question entity template specifically includes: step S1 and/or step S2.
  • the execution order of step S1 and step S2 is not specific.
  • step S1 according to at least one disease characteristic belonging to one type, fields can be filled in the question entity template corresponding to the type, so that the obtained target question carries information of at least one disease characteristic.
  • the type is "symptom”
  • the problem entity template corresponding to "symptom” can be filled with the definitions of symptoms such as symptom 1, symptom 2...symptom N to help the user determine whether to hit the corresponding disease feature and select the corresponding answer options.
  • Step S2 may fill in the jump relationship field of the problem entity template according to the hit action attribute in the disease feature entity corresponding to the disease feature. Specifically, information about the attribute of the hit action corresponding to the disease feature may be filled in the jump relationship field.
  • the content of the jump relationship field may be: when the answer option is selected, jump to the preset question entity, assuming that the answer option corresponds to the first disease feature entity, and the hit action attribute of the first disease feature entity records
  • the preset problem entity may be: problem entity information corresponding to the second disease feature entity.
  • the information of the first disease feature may be filled in the answer option field of the question entity template, and the question entity information corresponding to the second disease feature entity may be filled in the jump relationship field.
  • the corresponding question entity can be obtained from the knowledge map according to the jump relationship field in the question entity corresponding to the disease feature.
  • the jump relationship field is used to execute a preset jump under the condition that the answer option is selected.
  • the jump relationship field is used to jump from the first question entity to the second question entity under the condition that the answer option is selected, and the disease feature entity corresponding to the first question entity corresponds to the second question entity.
  • the disease feature entity of can be a parent-child relationship.
  • a question entity instance corresponding to symptoms such as "sleepiness” can be triggered according to the jump relationship field of the question entity.
  • the target question C can be generated according to the problem entity instance corresponding to the symptoms such as "drowsiness”.
  • the question text for target question C could be "Are you sleepy?” and the answer options for target question C could be: [Yes, No, I don't know].
  • the technical solution 1 determines the target problem based on the instance of the problem entity; the technical solution 2 determines the target problem based on the filling of the problem entity template; the technical solution 3 can obtain the problem entity through the jump relationship field in the problem entity. It can be understood that those skilled in the art can adopt any one or a combination of technical solutions 1 to 3 according to actual application requirements.
  • the method for processing the inquiry information in the embodiment of the present application establishes the association between the disease feature entity and the problem entity in the knowledge graph.
  • the corresponding problem entity can be searched in the knowledge map according to the disease characteristics corresponding to the candidate disease, and the target question for the consultation can be obtained according to the obtained problem entity.
  • the target question for the consultation can be generated in the consultation process, so the effect of the knowledge graph on the consultation can be improved, and the query can be improved. Efficiency of diagnosis and rationality of target problems.
  • the embodiments of the present application represent the relationship between the disease feature and the information of the problem entity through the hit action attribute in the disease feature entity.
  • the corresponding disease feature When the corresponding disease feature is selected, the corresponding problem entity will be triggered. Since the association between disease characteristics and problem entities can be automatically performed, it can not only reduce the resource cost of manual operation of the consultation path, but also improve the rationality of the problem sequence in the consultation process.
  • the jump relationship field is used to represent the question entity: when the answer option corresponding to the disease feature is selected, it will jump to the preset question entity, which can improve the sequence of questions in the consultation process.
  • rationality Since the association between disease characteristics and problem entities can be automatically performed, it can not only reduce the resource cost of manual operation of the consultation path, but also improve the rationality of the problem sequence in the consultation process.
  • FIG. 7 a flowchart of the steps of Embodiment 6 of a method for processing consultation information of the present application is shown, which may specifically include the following steps:
  • Step 701 obtaining a problem entity corresponding to a preset disease feature entity from a knowledge graph according to the user's disease feature obtained based on at least one user input; the above-mentioned problem entity is used to represent a problem related to the above-mentioned preset disease feature entity;
  • a preset disease feature entity matching the user's disease feature is obtained from the knowledge graph, and then a problem entity corresponding to the matching preset disease feature entity is obtained.
  • Step 702 Determine a preset disease problem according to the problem entity; the preset disease problem specifically includes: a problem text and at least one preset option;
  • Step 703 If a user selection operation for any preset option is received, output corresponding medical advice information.
  • the preset disease feature entity may represent the disease feature corresponding to the preset disease
  • the preset disease may include: a disease with a high degree of criticality and/or severity
  • the disease feature corresponding to the preset disease may include: critical Severe.
  • the preset disease problem is determined according to the problem entity corresponding to the preset disease feature entity, and the user is consulted by using the preset disease problem. Or the exclusion of pre-set diseases with higher severity, which can improve the security of consultation information processing.
  • a correspondence relationship between the disease feature and the problem entity corresponding to the preset disease feature entity may be established.
  • the user disease feature may be matched with the disease feature in the corresponding relationship to obtain the user disease feature and the predicted disease feature.
  • the disease feature entity may be characterized by a hit action attribute: the relationship between the disease feature and the information of the problem entity corresponding to the preset disease feature entity.
  • the hit action attribute in the corresponding disease feature entity can be searched according to the user's disease feature, and then the problem entity corresponding to the user's disease feature and the preset disease feature entity can be obtained.
  • the embodiment of the present application may determine the prediction related to "fever” according to the above correspondence or the hit action attribute corresponding to the "fever” entity.
  • Set the problem entity corresponding to the disease the question entity may include a preset disease question, the question text of the preset disease question may be "Which of the following symptoms?", and at least one preset option of the preset disease question may include: [Options corresponding to preset disease characteristics ,none of the above].
  • the options corresponding to the preset disease characteristics can include: [over 40 degrees, chills], etc.
  • the user's selection operation for any preset option is received, it means that the user has hit the preset disease feature, indicating that the user has a high probability of corresponding to the preset disease.
  • outputting medical advice information can improve the user's awareness of the disease. It can enhance the security of consultation information processing.
  • the preset disease problem is determined according to the problem entity corresponding to the preset disease characteristic entity, and the user is consulted by using the preset disease problem;
  • the pre-determined diseases with high criticality and/or severity are excluded, thereby improving the security of the consultation information processing.
  • Embodiment 7 of a method for processing medical inquiry information of the present application is shown, which may specifically include the following steps:
  • Step 801 receiving the chief complaint input by the user
  • Step 802 obtain the problem entity corresponding to the preset disease characteristic entity from the knowledge graph; the above-mentioned problem entity is used to represent the problem related to the above-mentioned preset disease characteristic entity;
  • Step 803 Determine a preset disease problem according to the problem entity; the preset disease problem specifically includes: a problem text and at least one preset option;
  • Step 804 If a user selection operation for any preset option is received, output corresponding medical advice information;
  • Step 805 If the user does not select any preset option, perform disease prediction processing on the user's disease characteristics to obtain corresponding candidate diseases;
  • Step 806 generating a target question according to the disease characteristics corresponding to the above-mentioned candidate diseases, and the above-mentioned target question is used to inquire the user.
  • the pre-set disease may be excluded first according to the main complaint. If the user's selection operation for any pre-set option is received, it means that the user suffers from the pre-set disease. In this case, the corresponding medical advice information can be output, so as to improve the security of the inquiry information processing.
  • the user does not select any of the preset options, it means that the probability of the user suffering from the preset disease is low, and the user can be consulted.
  • Embodiment 8 of a method for processing medical inquiry information of the present application is shown, which may specifically include the following steps:
  • Step 901 providing a knowledge graph
  • Step 902 converting at least one user input into a user symptom feature to obtain a user symptom feature set
  • Step 903 Search the knowledge graph to obtain the critical and critical problem corresponding to the user's symptoms, and determine whether there is a critical or critical problem according to the user's reply, if so, perform step 904, otherwise, perform step 905;
  • Step 904 outputting corresponding medical advice information
  • Step 905 performing disease prediction processing on the user symptom feature set to obtain corresponding candidate diseases and their scores;
  • Step 906 according to the candidate disease and its score, determine whether the stop condition is met, if yes, go to step 907, otherwise, go to step 908;
  • Step 907 output the processing result of the consultation information, and end the processing of the consultation information
  • Step 908 according to the symptom features corresponding to the candidate diseases and the knowledge map, sort the symptom features corresponding to the candidate diseases;
  • Step 909 Determine the target symptom feature from the symptom features corresponding to the candidate diseases according to the sorting result, and generate a target question according to the target symptom feature, and the target question is used to collect information such as user symptoms.
  • At least one user input may be converted into corresponding user symptom characteristics by using an entity identification method, or a symptom identification method such as a disease characteristic table matching method.
  • step 903 obtains the critical and critical problem corresponding to the user's symptoms, which may specifically include: obtaining the problem entity corresponding to the corresponding critical and critical problem from the knowledge graph according to the user's symptom feature set; the problem entity represents a problem.
  • Various types of information contained including question text, answer options, and jump relationships (that is, the corresponding actions when an answer option is selected, etc.);
  • the answer options for the critical critical problem may include: at least one critical critical option;
  • the user's selection operation for any critical illness option is received, it can be indicated that the above-mentioned user symptom feature set corresponds to a critical illness.
  • the process of disease prediction in step 905 specifically includes: according to the matching information between the user symptom feature set and the symptoms and other features corresponding to the disease in the medical knowledge graph, determining the candidate disease corresponding to the current user symptom feature set .
  • the process of performing disease prediction in step 905 specifically includes: determining the score of the candidate disease according to the probability feature;
  • the above-mentioned probability feature may specifically include at least one of the following features:
  • conditional probability of the disease feature matching the user symptom feature set under the condition of the candidate disease
  • the stop condition may include at least one of the following conditions:
  • At least one candidate disease has a score greater than a score threshold
  • the number of query rounds exceeds the round threshold.
  • the process of sorting the symptom features corresponding to the candidate diseases in step 908 specifically includes: according to the conditional probability of the disease feature entities corresponding to the symptom features included in the candidate diseases in each disease, and the priority of the candidate diseases.
  • the symptom features included in the candidate diseases are sorted.
  • the target symptom features that meet the criteria may be selected for inquiry according to the sorting result according to preset criteria.
  • the preset criteria may include: the importance score is greater than the second threshold, or the top X (X may be a natural number greater than 0) bits in the order of the importance score from high to low, etc.
  • step 909 generates the target question according to the target symptom feature, which may specifically include: merging the target symptom feature according to the type, and obtaining the problem entity of the corresponding type from the knowledge graph; according to the problem entity and belonging to the type
  • the target symptom features are generated to generate the target problem.
  • fields of the question entity template corresponding to the type may be filled, so that the obtained target question carries the information of the at least one target symptom characteristic.
  • the method for processing consultation information in the embodiment of the present application has the following technical effects:
  • the embodiment of the present application supports multiple input methods such as selection input of answer options, symptom input, short sentence input, etc.
  • the symptom recognition method can be used to convert user input into knowledge graphs with standard Describe the symptom characteristics, so that the user experience can be improved.
  • the embodiment of the present application can exclude critical illnesses before inquiring the user.
  • the user's symptom characteristics correspond to critical illnesses, the user is advised to seek medical treatment, which can improve the security of inquiry information processing.
  • the disease prediction and generation target questions in the embodiments of the present application may be dynamic processes; therefore, the target questions more related to the user's disease characteristics can be obtained according to the accumulation of the user's disease characteristics during the consultation process, thus improving the performance of the consultation process. rationality.
  • a stop condition is set according to the inquiry ratio of the symptom characteristics corresponding to the candidate disease, the number of inquiry rounds, and the sufficiency of the user's symptom set to the determination basis of the candidate disease, and when the stop condition is met, the consultation is ended, It can not only improve the flexibility of consultation, but also shorten the number of inquiry rounds, thereby improving the efficiency of consultation and improving user experience.
  • FIG. 10 a structural block diagram of an embodiment of an apparatus for processing consultation information according to the present application is shown, which may specifically include: a user disease feature determination module 1001 , a user disease feature processing module 1002 and a question generation module 1003 .
  • the user disease feature determination module 1001 is configured to determine the user disease feature according to at least one user input
  • a user disease feature processing module 1002 configured to perform disease prediction processing on the user disease feature to obtain a corresponding candidate disease
  • the question generation module 1003 is configured to generate a target question according to the disease characteristics corresponding to the candidate disease, and the target question is used to inquire the user.
  • the user disease feature processing module 1002 may include:
  • the candidate disease determination module is configured to determine the candidate disease corresponding to the user's disease characteristic according to the matching information between the above-mentioned user's disease characteristic and the disease characteristic corresponding to the disease.
  • the question generation module 1003 may include:
  • the problem entity obtaining module is configured to obtain the corresponding problem entity from the knowledge map according to the disease feature entity corresponding to the above disease feature; the above problem entity is used to represent the problem related to the above disease feature entity;
  • the first question generation module is configured to generate the target question according to the question corresponding to the above question entity.
  • the above-mentioned question entity may include: a question entity template; the above-mentioned first question generation module may include:
  • the field filling module is configured to perform field filling on the above problem entity template according to the above disease characteristics, so as to obtain the target problem.
  • the above field filling module may include:
  • the first field filling module is configured to perform field filling on the question entity template corresponding to the above type according to at least one disease characteristic belonging to one type, so that the obtained target question carries information of at least one disease characteristic.
  • the fields of the above question entity template may include: a question text field and an answer option field;
  • the second field filling module is configured to fill in the information of the above disease characteristics in the answer option field of the above question entity template.
  • the above field filling module may include:
  • the third field filling module is configured to fill in the jump relationship field of the problem entity template according to the hit action attribute in the disease feature entity corresponding to the disease feature.
  • the above-mentioned problem entity may include: a problem entity instance; the above-mentioned problem generation module may include:
  • the second question generation module is configured to obtain the target question from the question entity instance corresponding to the above-mentioned disease feature.
  • the above problem entity acquisition module may include:
  • the first question entity acquisition module is configured to obtain the corresponding question entity from the knowledge map according to the hit action attribute in the disease feature entity corresponding to the above disease feature when the answer option corresponding to the above disease feature is selected; and/ or,
  • the second question entity obtaining module is configured to obtain the corresponding question entity from the knowledge map according to the jump relationship field in the question entity corresponding to the disease feature when the answer option corresponding to the disease feature is selected.
  • the above-mentioned question generation module 1003 may include:
  • the target disease feature determination module is configured to determine the target disease feature from the disease features corresponding to the above-mentioned candidate diseases according to the importance score corresponding to the disease feature; the above-mentioned target disease feature is configured to represent the disease feature inquired by the user in this round of consultation;
  • the entity determination module is configured to obtain the corresponding type of problem entity from the knowledge map according to the type corresponding to the above-mentioned target disease characteristic; the above-mentioned problem entity is used to represent the problem related to the above-mentioned disease characteristic entity;
  • the third question generation module is configured to generate a target question according to the above-mentioned problem entity and the target disease characteristics belonging to the above-mentioned type.
  • the above-mentioned at least one user input may include:
  • the above target question may include: question text and answer options;
  • the above-mentioned at least one user input may include: an answer option selected by the user.
  • the above device may also include:
  • the preset problem entity obtaining module is configured to obtain the problem entity corresponding to the preset disease feature entity from the knowledge map according to the user disease feature; the problem entity is used to represent the problem related to the preset disease feature entity;
  • the preset disease problem determination module is configured to determine the preset disease problem according to the above-mentioned problem entity; the above-mentioned preset disease problem may include: problem text and at least one preset option;
  • the suggestion output module is configured to output the corresponding medical advice information if a user selection operation for any preset option is received.
  • the above device may also include:
  • a score determination module configured to determine the score of the above-mentioned candidate disease according to the probability feature
  • the above-mentioned probability feature may include at least one of the following features:
  • the above device may also include:
  • the stopping module is configured to notify the above-mentioned problem generation module to stop executing the above-mentioned generation target problem if the stopping condition is met after the candidate disease is obtained.
  • the above device may also include:
  • the processing result output module is configured to output the processing result of the consultation information if the stopping condition is met after the candidate disease is obtained; the above-mentioned processing result of the consultation information may include: the information of the above-mentioned candidate disease.
  • the above-mentioned processing result of the consultation information may further include: disease characteristics that appear in the user, and disease characteristics that do not appear in the user.
  • the above stop condition may include at least one of the following conditions:
  • At least one candidate disease has a score greater than a score threshold
  • the number of query rounds exceeds the round threshold.
  • the device for processing medical information has the following technical effects:
  • the embodiment of the present application supports multiple input methods such as selection input of answer options, symptom input, short sentence input, etc.
  • the symptom recognition method can be used to convert user input into knowledge graphs with standard Describe the symptom characteristics, so that the user experience can be improved.
  • the embodiment of the present application can exclude critical illnesses before inquiring the user.
  • the user's symptom characteristics correspond to critical illnesses, the user is advised to seek medical treatment, which can improve the security of inquiry information processing.
  • the disease prediction and generation target questions in the embodiments of the present application may be dynamic processes; therefore, the target questions more related to the user's disease characteristics can be obtained according to the accumulation of the user's disease characteristics during the consultation process, thus improving the performance of the consultation process. rationality.
  • a stop condition is set according to the inquiry ratio of the symptom characteristics corresponding to the candidate disease, the number of inquiry rounds, and the sufficiency of the user's symptom set to the determination basis of the candidate disease, and when the stop condition is met, the consultation is ended, It can not only improve the flexibility of consultation, but also shorten the number of inquiry rounds, thereby improving the efficiency of consultation and improving user experience.
  • An embodiment of the present application provides an apparatus for processing consultation information, including a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be processed by the one or more programs
  • the execution of the one or more programs by the device includes instructions for performing the following operations: determining the disease characteristics of the user according to at least one user input; performing disease prediction processing on the user disease characteristics to obtain corresponding candidate diseases; The disease feature entity corresponding to the disease feature of the candidate disease, and the corresponding problem entity is obtained from the knowledge map; the problem entity is used to represent the problem related to the disease feature entity; according to the problem corresponding to the problem entity, the target problem is generated ; The target question is used to inquire the user.
  • FIG. 11 is a block diagram of an apparatus 1100 for processing consultation information according to an exemplary embodiment.
  • apparatus 1100 may be a mobile phone, computer, digital broadcast terminal, messaging device, game console, tablet device, medical device, fitness device, personal digital assistant, and the like.
  • the apparatus 1100 may include one or more of the following components: a processing component 1102, a memory 1104, a power supply component 1106, a multimedia component 1108, an audio component 1110, an input/output (I/O) interface 1112, a sensor component 1114, and communication component 1116.
  • a processing component 1102 a memory 1104, a power supply component 1106, a multimedia component 1108, an audio component 1110, an input/output (I/O) interface 1112, a sensor component 1114, and communication component 1116.
  • the processing component 1102 generally controls the overall operation of the device 1100, such as operations associated with display, phone calls, data communications, camera operations, and recording operations.
  • the processing element 1102 may include one or more processors 1120 to execute instructions to perform all or part of the steps of the methods described above. Additionally, processing component 1102 may include one or more modules that facilitate interaction between processing component 1102 and other components. For example, processing component 1102 may include a multimedia module to facilitate interaction between multimedia component 1108 and processing component 1102.
  • Memory 1104 is configured to store various types of data to support operation at device 1100 . Examples of such data include instructions for any application or method operating on the device 1100, contact data, phonebook data, messages, pictures, videos, and the like. Memory 1104 may be implemented by any type of volatile or non-volatile storage device or combination thereof, such as static random access memory (SRAM), electrically erasable programmable read only memory (EEPROM), erasable Programmable Read Only Memory (EPROM), Programmable Read Only Memory (PROM), Read Only Memory (ROM), Magnetic Memory, Flash Memory, Magnetic or Optical Disk.
  • SRAM static random access memory
  • EEPROM electrically erasable programmable read only memory
  • EPROM erasable Programmable Read Only Memory
  • PROM Programmable Read Only Memory
  • ROM Read Only Memory
  • Magnetic Memory Flash Memory
  • Magnetic or Optical Disk Magnetic Disk
  • Power component 1106 provides power to various components of device 1100 .
  • Power components 1106 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to device 1100 .
  • Multimedia component 1108 includes a screen that provides an output interface between the device 1100 and the user.
  • the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from a user.
  • the touch panel includes one or more touch sensors to sense touch, swipe, and gestures on the touch panel. The touch sensor may not only sense the boundaries of a touch or swipe action, but also detect the duration and pressure associated with the touch or swipe action.
  • the multimedia component 1108 includes a front-facing camera and/or a rear-facing camera. When the device 1100 is in an operation mode, such as a shooting mode or a video mode, the front camera and/or the rear camera may receive external multimedia data. Each of the front and rear cameras can be a fixed optical lens system or have focal length and optical zoom capability.
  • Audio component 1110 is configured to output and/or input audio signals.
  • audio component 1110 includes a microphone (MIC) that is configured to receive external audio signals when device 1100 is in operating modes, such as calling mode, recording mode, and voice data processing mode. The received audio signal may be further stored in memory 1104 or transmitted via communication component 1116 .
  • audio component 1110 also includes a speaker for outputting audio signals.
  • the I/O interface 1112 provides an interface between the processing component 1102 and a peripheral interface module, which may be a keyboard, a click wheel, a button, or the like. These buttons may include, but are not limited to: home button, volume buttons, start button, and lock button.
  • Sensor assembly 1114 includes one or more sensors for providing status assessment of various aspects of device 1100 .
  • the sensor assembly 1114 can detect the on/off state of the device 1100, the relative positioning of components, such as the display and keypad of the device 1100, and the sensor assembly 1114 can also detect a change in the position of the device 1100 or a component of the device 1100 , the presence or absence of user contact with the device 1100 , the device 1100 orientation or acceleration/deceleration and the temperature change of the device 1100 .
  • Sensor assembly 1114 may include a proximity sensor configured to detect the presence of nearby objects in the absence of any physical contact.
  • Sensor assembly 1114 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications.
  • the sensor assembly 1114 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.
  • Communication component 1116 is configured to facilitate wired or wireless communication between apparatus 1100 and other devices.
  • Device 1100 may access wireless networks based on communication standards, such as WiFi, 2G or 3G, or a combination thereof.
  • the communication component 1116 receives broadcast signals or broadcast related information from an external broadcast management system via a broadcast channel.
  • the communication component 1116 also includes a near field communication (NFC) module to facilitate short-range communication.
  • NFC near field communication
  • the NFC module may be implemented based on radio frequency data processing (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.
  • RFID radio frequency data processing
  • IrDA infrared data association
  • UWB ultra-wideband
  • Bluetooth Bluetooth
  • apparatus 1100 may be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable A gate array (FPGA), controller, microcontroller, microprocessor or other electronic component implementation is used to perform the above method.
  • ASICs application specific integrated circuits
  • DSPs digital signal processors
  • DSPDs digital signal processing devices
  • PLDs programmable logic devices
  • FPGA field programmable A gate array
  • controller microcontroller, microprocessor or other electronic component implementation is used to perform the above method.
  • non-transitory computer-readable storage medium including instructions, such as a memory 1104 including instructions, which are executable by the processor 1120 of the apparatus 1100 to perform the method described above.
  • the non-transitory computer-readable storage medium may be ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, optical data storage device, and the like.
  • FIG. 12 is a schematic structural diagram of a server in some embodiments of the present application.
  • the server 1900 may vary greatly due to different configurations or performance, and may include one or more central processing units (CPU) 1922 (eg, one or more processors) and memory 1932, one or more One or more storage media 1930 (eg, one or more mass storage devices) that store applications 1942 or data 1944.
  • the memory 1932 and the storage medium 1930 may be short-term storage or persistent storage.
  • the program stored in the storage medium 1930 may include one or more modules (not shown in the figure), and each module may include a series of instructions to operate on the server.
  • the central processing unit 1922 may be configured to communicate with the storage medium 1930 to execute a series of instruction operations in the storage medium 1930 on the server 1900 .
  • Server 1900 may also include one or more power supplies 1926, one or more wired or wireless network interfaces 1950, one or more input and output interfaces 1958, one or more keyboards 1956, and/or, one or more operating systems 1941, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, etc.
  • a non-transitory computer-readable storage medium when the instructions in the storage medium are executed by the processor of the device (server or terminal), the device can execute the consultation information shown in any one of FIG. 2 to FIG. 9 .
  • a non-transitory computer-readable storage medium when an instruction in the storage medium is executed by a processor of a device (server or terminal), the device can execute a method for processing consultation information, the method comprising: Determine the user's disease characteristics according to at least one user input; perform disease prediction processing on the user's disease characteristics to obtain corresponding candidate diseases; obtain the corresponding disease characteristic entities from the knowledge map according to the disease characteristic entities corresponding to the disease characteristics of the candidate diseases.
  • Question entity the question entity is used to represent the question related to the disease feature entity; according to the question corresponding to the question entity, a target question is generated; the target question is used to inquire the user.

Landscapes

  • Health & Medical Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Medical Informatics (AREA)
  • Public Health (AREA)
  • Biomedical Technology (AREA)
  • Epidemiology (AREA)
  • General Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • Pathology (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • Medical Treatment And Welfare Office Work (AREA)

Abstract

本申请实施例提供了一种问诊信息处理方法、装置和介质。其中的方法具体包括：依据至少一次用户输入，确定用户疾病特征；对所述用户疾病特征进行疾病预测处理，以得到对应的候选疾病；依据所述候选疾病的疾病特征对应的疾病特征实体，从知识图谱中获取对应的问题实体；所述问题实体用于表征与所述疾病特征实体相关的问题；依据所述问题实体对应的问题，生成目标问题；所述目标问题用于对用户进行问诊。本申请实施例可以提高用于问诊的目标问题的准确度，从而提高问诊的效率。

Description

一种问诊信息处理方法、装置和介质
本申请要求在2021年1月26日提交中国专利局、申请号为202110105880.9、发明名称为“一种问诊信息处理方法、装置和介质”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请实施例涉及医疗技术领域，尤其涉及一种问诊信息处理方法、装置和介质。
背景技术
问诊是通过对患者或代诊者进行有目的的询问，了解疾病的发生、发展、诊治经过、现在症状及其他一切与疾病有关的情况，以诊察疾病的一种方法。随着人工智能技术的不断发展，基于人工智能的问诊方式逐渐发展起来，给用户的生活带来了诸多便捷。
目前的问诊信息处理方法，通常预置问答对；这样，在接收到用户输入的问诊数据的情况下，从预配置的问答对中查询与该问诊数据相匹配的目标解答数据，并将所查询到的目标解答数据反馈给用户。
在实际应用中，在未查询到与该问诊数据相匹配的目标解答数据的情况下，需要针对该问诊数据触发人工问诊流程，以通过人工来解答该问诊数据；因此，目前的问诊信息处理方法存在问诊效率低的问题。
发明内容
本申请实施例提供一种问诊信息处理方法、装置和介质，能够提高用于问诊的目标问题的准确度，从而提高问诊的效率。
为了解决上述问题，本申请实施例公开了一种问诊信息处理方法，包括：
依据至少一次用户输入，确定用户疾病特征；
对所述用户疾病特征进行疾病预测处理，以得到对应的候选疾病；
依据所述候选疾病的疾病特征对应的疾病特征实体，从知识图谱中获取对应的问题实体；所述问题实体用于表征与所述疾病特征实体相关的问题；
依据所述问题实体对应的问题，生成目标问题；所述目标问题用于对用户进行问诊。
另一方面，本申请实施例公开了一种问诊信息处理装置，包括：
用户疾病特征确定模块，配置为依据至少一次用户输入，确定用户疾病特征；
用户疾病特征处理模块，配置为对所述用户疾病特征进行疾病预测处理，以得到对应的候选疾病；
问题实体获取模块，配置为依据所述候选疾病的疾病特征对应的疾病特征实体，从知识图谱中获取对应的问题实体；所述问题实体用于表征与所述疾病特征实体相关的问题；
第一问题生成模块，配置为依据所述问题实体对应的问题，生成目标问题；所述目标问题用于对用户进行问诊。
再一方面，本申请实施例公开了一种用于处理问诊信息的装置，包括有存储器，以及一个或者一个以上的程序，其中一个或者一个以上程序存储于存储器中，且所述程序被一个或者一个以上处理器执行时，实现前述方法的步骤。
又一方面，本申请实施例公开了一种机器可读介质，其上存储有指令，当由一个或多个处理器执行时，使得装置执行如前述一个或多个所述的问诊信息处理方法。
本申请实施例包括以下优点：
本申请实施例在问诊过程中，通过对问诊信息的处理，基于至少一次用户输入确定的用户疾病特征，动态进行疾病预测处理，并动态生成目标问题。由于本申请实施例在问诊流程中，通过对问诊信息的处理，自动生成用于问诊的目标问题，因此能够提高问诊效率。
并且，由于本申请实施例中，对问诊信息进行的疾病预测处理和生成目标问题，可以为动态过程；因此，能够依据问诊过程中用户疾病特征的累积，得到与用户疾病特征更相关的目标问题，因此可以提高依据该目标问题进行问诊的合理性；以及，能够依据问诊过程中用户疾病特征的累积，得到与用户疾病特征更匹配的候选疾病，因此可以提高生成用于疾病预测处理的目标问题的准确度；另外，本申请实施例可依据预测得到的疾病特征、用户疾病特征、询问轮数等信息，动态决定问诊信息处理是否提前停止，提高问诊效率，从而提高用户体验。
附图说明
为了更清楚地说明本申请实施例的技术方案，下面将对本申请实施例的描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。
图1是本申请实施例的一种问诊信息处理方法的应用环境的示意；
图2是本申请的一种问诊信息处理方法实施例一的步骤流程图；
图3是本申请实施例的一种疾病实体及其属性的示意图；
图4是本申请实施例的一种疾病特征实体及其属性的示意图；
图5是本申请的一种知识图谱的处理方法实施例一的步骤流程图；
图6是本申请的一种问诊信息处理方法实施例四的步骤流程图；
图7是本申请的一种问诊信息处理方法实施例六的步骤流程图；
图8是本申请的一种问诊信息处理方法实施例七的步骤流程图；
图9是本申请的一种问诊信息处理方法实施例八的步骤流程图；
图10是本申请的一种问诊信息处理装置实施例的结构框图；
图11是本申请的一种用于处理问诊信息的装置1100的框图；及
图12是本申请的一些实施例中服务端的结构示意图。
具体实施方式
下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。
针对传统技术中问诊效率低的技术问题，本申请实施例提供了一种问诊信息处理方法，该方法可以包括：依据至少一次用户输入，确定用户疾病特征；对上述用户疾病特征进行疾病预测处理，以得到对应的候选疾病；依据上述候选疾病对应的疾病特征，生成目标问题，上述目标问题用于对用户进行问诊。
本申请实施例中，用户的至少一次用户输入中可以包括：用户出现的疾病特征(以下简称用户疾病特征)，本申请实施例可以基于问诊确定用户疾病特征，并对用户疾病特征进行处理。疾病特征的类型可以包括：症状、诱因、高发季节、接触史、家族史等。
本申请实施例在问诊过程中，通过对问诊信息的处理，基于至少一次用户输入得到的用户疾病特征，动态进行疾病预测处理，并动态生成目标问题。由于本申请实施例在问诊流程中，通过对问诊信息的处理，自动生成用于问诊的目标问题，因此能够提高问诊效率。
并且，由于本申请实施例中，对问诊信息进行的疾病预测处理和生成目标问题，可以为动态过程；因此，能够依据问诊过程中用户疾病特征的累积，得到与用户疾病特征更相关的目标问题，因此可以提高问诊信息处理的合理性；以及，能够依据问诊过程中 用户疾病特征的累积，得到与用户疾病特征更匹配的候选疾病，因此可以提高生成用于疾病预测处理的目标问题的准确度。
本申请实施例提供的问诊信息处理方法可以应用于例如网站和/或APP(应用程序，Application)的应用场景中，例如，本申请实施例的应用场景可以包括：医疗相关网站、或者医疗相关APP等。
本申请实施例提供的问诊信息处理方法可应用于图1所示的应用环境中，如图1所示，客户端100与服务端200位于有线或无线网络中，通过该有线或无线网络，客户端100与服务端200进行数据交互。
可选地，客户端100可以运行在终端上，上述终端具体包括但不限于：智能手机、平板电脑、电子书阅读器、MP3(动态影像专家压缩标准音频层面3，Moving Picture Experts Group Audio Layer III)播放器、MP4(动态影像专家压缩标准音频层面4，Moving Picture Experts Group Audio Layer IV)播放器、膝上型便携计算机、车载电脑、台式计算机、机顶盒、智能电视机、可穿戴设备等等。
在实际应用中，客户端100可以与用户进行交互。具体地，客户端100可以接收至少一次用户输入，并向用户提供目标问题。
客户端100可以利用本申请实施例的问诊信息处理方法，生成目标问题。或者，客户端100可以向服务端200发送用户的至少一次用户输入，以使服务端200利用本申请实施例的问诊信息处理方法，生成用于问诊的目标问题。
方法实施例一
参照图2，示出了本申请的一种问诊信息处理方法实施例一的步骤流程图，具体可以包括如下步骤：
步骤201、依据至少一次用户输入，确定用户疾病特征；
步骤202、对上述用户疾病特征进行疾病预测处理，以得到对应的候选疾病；
步骤203、依据上述候选疾病对应的疾病特征，生成目标问题，上述目标问题用于对用户进行问诊。
图2所示方法实施例的至少一个步骤可由客户端和/或服务端执行，当然本申请实施例对于各个步骤的具体执行主体不加以限制。
步骤201中，可以通过键盘输入、选项选择、语音输入等输入方式，接收至少一次用户输入。
可选地，上述至少一次用户输入可以包括：
主动输入；或者
主动输入和针对预设问题的回复；或者
主动输入和针对所述目标问题的回复；或者
主动输入、以及针对预设问题和目标问题的回复。其中，所述回复可以包括文本输入、答案选项的选择等，不限定回复的形式。
相对于依据至少一次用户输入动态生成的目标问题，预设问题可以为预先存储的问题。
例如，在问诊流程中，用户的首次输入通常为主动输入。该主动输入中通常包括：主诉。在医疗技术领域，主诉用于表征患者或代诊者对最主要的症状和(或)体征的叙述，主诉通常包括：患者或代诊者陈述的症状、体征、性质、以及持续时间等内容中的至少一种。
在接收到主动输入后，可以向用户提供预设问题，以获取用户针对预设问题的回复。
预设问题可以为问诊中频率超过频率阈值的问题，如预设问题的关键词可以包括：持续时间、精神状态等，以对症状的持续时间、患者的精神状态进行询问。可以理解，本申请实施例对于具体的预设问题不加以限制。
可选地，目标问题具体包括：问题文本和答案选项；则上述至少一次用户输入，具体包括：用户选择的答案选项。
问题文本的示例可以包括：“是否没有恶心动作直接呕吐？”、“有以下哪些症状？”、“之前得过以下哪些疾病？、“皮疹是什么类型的？“大便的形状是怎样的？”等。答案选项用于表征可供选择的答案选项。
对于“是否没有恶心动作直接呕吐？”等是非类型的问题，答案选项可以包括：[是，否，我不知道]等。
对于“有以下哪些症状？”等非是非类型的问题，答案选项可以包括：疾病特征对应的答案选项。例如，答案选项可以包括：[症状1，症状2…症状N，以上都没有]。
在本申请的一种可选实施例中，可以从至少一次用户输入中确定出用户疾病特征。用户疾病特征的确定方法可以包括但不限于：实体识别方法、或疾病特征表的匹配方法等。可以将确定出的用户疾病特征保存至用户疾病特征集合。
需要说明的是，至少一次用户输入中可以包括用户疾病特征对应的标准描述。或者，至少一次用户输入中可以包括用户疾病特征对应的非标准描述，如口语化描述等，此种情况下，可以将至少一次用户输入中的非标准描述转换为标准描述。因此，本申请实施例将使用用户疾病特征的标准描述，对标准化后的用户疾病特征进行疾病预测处理，以提高疾病预测的准确度。
本申请实施例中，可以依据医学书籍、医学数据库、医学问答数据等医疗资源，确定疾病对应的疾病特征；或者，可以利用构建知识图谱的方式，确定疾病对应的疾病特征。知识图谱的相关内容，将在后续的实施例进行介绍。
步骤202中，疾病预测处理可用于确定用户疾病特征对应的候选疾病的概率。该候选疾病可以为至少一个，该候选疾病可以对应有得分，该得分可以表征在用户疾病特征的条件下候选疾病的概率。
上述对上述用户疾病特征进行疾病预测处理，具体可以包括：依据上述用户疾病特征与疾病的疾病特征之间的匹配信息，确定上述用户疾病特征对应的候选疾病。例如，若用户疾病特征与疾病A的疾病特征相匹配，则可以将疾病A作为用户疾病特征对应的候选疾病。
本申请实施例中，对用户疾病特征进行的疾病预测处理，可以为动态过程。在用户疾病特征集合发生更新的情况下，可以对更新后的用户疾病特征集合进行疾病预测处理。这样，能够依据问诊过程中用户疾病特征的累积，得到与用户疾病特征更匹配的候选疾病，因此可以提高据此生成的用于问诊的目标问题的准确度。
步骤203中，目标问题可用于对候选疾病对应的疾病特征进行问诊，以帮助用户确定是否出现对应的疾病特征。
可选地，目标问题可以包括：问题文本和答案选项，上述答案选项与候选疾病对应的疾病特征相应，这样，可以根据用户针对答案选项的选择操作，确定用户是否出现对应的疾病特征。在确定用户出现对应的疾病特征的情况下，本申请实施例可以将用户选择的答案选项，转换为对应的用户疾病特征。
例如，目标问题A为“有以下哪些症状？”，答案选项可以包括：[症状1，症状2…症状N，以上都没有]，假设用户选择了答案选项[症状1]，则可以将“症状1”确定为用户疾病特征；假设用户选择了答案选项[以上都没有]，则目标问题A的问诊未命中用户疾病特征。
本申请实施例还可以依据用户针对答案选项的选择操作，确定并记录用户未出现的疾病特征，以在问诊信息处理结果中携带用户未出现的疾病特征。
本申请实施例可以从候选疾病对应的疾病特征中确定出目标疾病特征，并针对目标疾病特征生成目标问题。所述目标疾病特征可用于表征本轮问诊向用户询问的疾病特征。
根据一种可选实施例，可以从候选疾病对应的疾病特征中去除依据用户输入确定的用户疾病特征，以得到目标疾病特征。
根据另一种可选实施例，可以依据疾病特征对应的重要性得分，从所述候选疾病对 应的疾病特征中确定出目标疾病特征。
上述重要性得分的因子特征具体可以包括如下特征中的至少一种：
疾病特征在疾病的条件下的条件概率；
疾病在疾病系统中的发病概率；
疾病系统的系统概率；
候选疾病对应的疾病特征与用户疾病特征之间的相关性。
例如，可以依据候选疾病对应的疾病特征与依据用户输入确定的用户疾病特征之间的相关性，从候选疾病对应的疾病特征中确定出目标疾病特征。例如，可以从候选疾病对应的疾病特征中确定出相关性大于第一阈值的目标疾病特征。可选地，可以依据多种疾病特征在一种疾病的医疗资源中的共现信息，确定疾病特征之间的相关性。可以理解，本申请实施例对于疾病特征之间的相关性的具体确定方式不加以限制。
在本申请的一种可选实施例中，可以将用户对应的预设问题和目标问题保存至问题队列，从问题队列中取出一个问诊问题，并向用户输出该问诊问题。
问题队列中的问题可以对应有优先级，可以依据问题的优先级，从问题队列中取出问诊问题。问题的优先级的确定因子可以包括：问题的入队时间、和/或、问题对应的疾病特征与用户疾病特征之间的匹配度等。可以理解，本申请实施例对于从问题队列中取出一个问诊问题的具体过程不加以限制。
本申请实施例的执行主体可以为服务端或客户端，在执行主体为服务端的情况下，服务端可以为处理引擎或交互引擎，其中，交互引擎可用于与客户端进行交互，处理引擎可以向交互引擎发送问诊问题，以使交互引擎向用户提供问诊问题。可以理解，本申请实施例对于向用户提供问诊问题的具体实现方式不加以限制。
在本申请的一种可选实施例中，在每次执行步骤202后，可以根据候选疾病的得分、用户疾病特征、询问轮数等信息，判断是否结束问诊信息处理；若结束问诊则输出问诊信息处理结果，否则继续步骤203。
本申请实施例依据预测得到的疾病特征、用户疾病特征、询问轮数等信息，动态决定问诊是否提前停止，提高问诊效率，从而提高用户体验。
综上，本申请实施例的问诊信息处理方法，在问诊过程中，依据基于至少一次用户输入得到的用户疾病特征，动态进行疾病预测处理，并动态生成目标问题。由于本申请实施例在问诊流程中自动生成用于问诊的目标问题，因此能够提高问诊效率。
并且，由于本申请实施例中，对用户疾病特征进行的疾病预测处理和生成目标问题，可以为动态过程；因此，能够依据问诊过程中用户疾病特征的累积，得到与用户疾病特 征更相关的目标问题，因此可以提高依据该目标问题进行问诊的合理性；以及，能够依据问诊过程中用户疾病特征的累积，得到与用户疾病特征更匹配的候选疾病，因此可以提高生成的用于疾病预测的目标问题的准确度。
方法实施例二
本实施例用于说明知识图谱。
本申请实施例中，知识图谱是结构化的语义知识库，用于描述物理世界中的概念及其相互关系。
本申请实施例中，实体(Entity)是指客观存在并可相互区别的事物，包括具体的人、事、物、抽象的概念或联系等。实体可以是具体的对象，如：一种疾病，一种疾病特征等；也可以是抽象的事件，如：针对疾病特征的一次问诊等。
实体可以有很多特性，单个特性称为属性。每个属性有一个值域，其类型可以是整数型、实数型、或字符串型等。标记属性的命名单位称为字段。字段的状态可以包括：已填充状态或未填充状态，其中，已填充状态对应已填充的字段内容，未填充状态表征对应的字段内容待填充。
医疗领域中的实体可以称为医疗实体。医疗实体可以包括：疾病实体、疾病特征实体、或问题实体等。
疾病实体可以表征具体的疾病，如“高血压”、“白血病”等。疾病可以对应有疾病系统。疾病系统可以对应解剖学意义上的系统，如疾病系统可以包括：运动系统、消化系统、呼吸系统、泌尿系统、生殖系统、内分泌系统、免疫系统、神经系统和循环系统等。
可选地，疾病实体的属性可以包括如下属性中的至少一种：
疾病标识属性、疾病系统属性、特征集合属性、临床占比属性、以及高发年龄属性；
所述特征集合中可以包括：与该疾病特征实体相关联的疾病特征；
所述临床占比用于表征疾病在疾病系统中的发病概率，可以依据疾病的发病数量与疾病系统的发病数量得到。
参照图3，示出了本申请实施例的一种疾病实体及其属性的示意图。其中，疾病实体的属性可以包括：疾病标识属性、疾病系统属性、特征集合属性、临床占比属性、以及高发年龄属性。
单个属性可以对应有属性参数。
例如，疾病系统属性的属性参数包括：系统概率，可以表征单个疾病系统的患者，占所有疾病系统的患者的比例，可以依据单个疾病系统的病人和所有疾病系统的病人的比值得到。
又如，特征集合属性的属性参数可以包括如下参数中的至少一种：
疾病特征在疾病的条件下的条件概率；特征集合中通常包括多个疾病特征，条件概率可以为多个疾病特征在疾病的条件下分别对应的条件概率；
疾病特征在疾病的条件下的惩罚因子，该惩罚因子可以对应在疾病的条件下不能出现的疾病特征，用于在疾病预测的过程中，对疾病的概率进行惩罚。
参照表1，示出了本申请实施例的一种疾病实体的实例的示意，其中，疾病名称“急性喉炎”和“支气管炎”均为“呼吸系统”的疾病，其分别对应多种疾病特征，每种疾病特征分别对应有条件概率。
表1
Figure PCTCN2021103667-appb-000001
疾病特征实体可以表征具体的疾病特征。疾病特征的类型可以包括：症状、诱因、高发季节、接触史、家族史等。
疾病特征实体的属性可以包括：命中动作属性，该命中动作属性用于表征在对应疾病特征实体被选中的情况下，触发的问题实体的信息。
疾病特征实体的属性可以包括：从属关系属性，用于表征与对应疾病特征实体具有父子关系的疾病特征实体。从属关系属性的属性参数可以包括：父疾病特征或子疾病特征。
例如，疾病特征“呕吐”的子疾病特征包括：“喷射性呕吐”；又如，疾病特征“发热”的子疾病特征包括：“低热”、“高热”等。
参照图4，示出了本申请实施例的一种疾病特征实体及其属性的示意图。其中，疾病特征实体的属性可以包括：特征标识属性、类型属性、从属关系属性、频次属性、命中动作属性、释义属性等。其中，频次属性可以表征对应疾病特征在所有疾病实体的特征集合中出现的次数。
参照表2，示出了本申请实施例的一种疾病特征实体的实例的示意。其中，疾病特征“呕吐”的命名动作属性“问题实体的标识18”，表征在疾病特征“呕吐”被选中的情况下，将触发问题实体标识为18的问题实体。
表2
Figure PCTCN2021103667-appb-000002
上述命中动作属性可以提高问诊流程中问题接序的合理性。例如，若用户选中了“呕吐”症状，则会根据该“呕吐”症状的命中动作属性，查找到对应的问题实体标识18，以进一步针对“喷射性呕吐”症状进行问诊。
问题实体与针对疾病特征的一次问诊相应，用于表征一次问诊对应的问题。由于一次问诊可以涉及至少一个疾病特征，因此，问题实体对应的问题可以涉及至少一个疾病特征。
可选地，问题实体的字段可以包括：问题文本字段和答案选项字段。问题文本字段用于表征待回复的问题。问题文本字段的示例可以包括：“是否没有恶心动作直接呕吐？”、“有以下哪些症状？”、“之前得过以下哪些疾病？、“皮疹是什么类型的？“大便的形状是怎样的？”等。答案选项字段用于表征可供选择的答案选项。
可选地，问题实体的字段还可以包括如下字段中的至少一种：疾病特征字段、触发条件字段、以及跳转关系字段；
其中，所述疾病特征字段用于表征疾病特征实体；
所述触发条件字段用于表征对应的问题实体为依据疾病特征实体触发得到；
所述跳转关系字段用于选中答案选项的条件下执行预设跳转。
可选地，所述跳转关系字段用于在选中答案选项的条件下，从第一问题实体跳转至第二问题实体，第一问题实体对应的疾病特征实体与所述第二问题实体对应的疾病特征实体为父子关系。
当然，本领域技术人员可以根据实际应用需求，确定预设跳转，例如，预设跳转还可以包括：执行预设函数，预设函数可用于结束问题实体的查找等。
参照表3，示出了问题实体的字段的意义和取值。
表3
Figure PCTCN2021103667-appb-000003
本申请实施例的问题实体可以包括：问题实体实例、和/或、问题实体模板。
其中，问题实体实例的所有字段处于已填充状态。问题实体实例可以与预设的疾病特征相应。
参照表4，示出了本申请实施例的一种问题实体实例的示意。该问题实体实例的标识为18，为依据疾病特征“呕吐”触发得到，也即，在“呕吐”特征被选中的情况下，可以触发标识为18的问题实体实例。
该问题实体实例与疾病特征“喷射性呕吐”相应，用于对疾病特征“喷射性呕吐”进行问诊。问题文本中可以包含疾病特征“喷射性呕吐”的释义，以帮助用户确定是否命中对应的疾病特征，并选择对应的答案选项。
表4
标识 18
问题文本 是否没有恶心动作直接呕吐？
答案选项 [是，否，我不知道]
疾病特征 喷射性呕吐
触发条件 呕吐
问题实体模板的问题文本字段处于已填充状态，问题实体模板的除了问题文本字段之外的预设字段处于未填充状态。预设字段可以包括：答案选项字段、疾病特征字段、触发条件字段、以及跳转关系字段等。
问题实体模板可以与预设类型的疾病特征相应，这样，在问诊流程中，可以依据用户相关的疾病特征对应的类型，在知识图谱中查找得到对应的问题实体模板，并依据用户相关的疾病特征，对问题实体模板进行字段填充，字段填充后的问题实体模板可以作为动态的问题实体实例，动态的问题实体实例可以包含用于问诊的问题。由于问题实体模板与预设类型的疾病特征相应，故字段填充后的问题实体模板中可以包含预设类型的多种疾病特征的信息，因此能够用于问诊的问题所包含的疾病特征数量，进而能够减少问诊的交互轮数，以及能够提高问诊效率。
上述对问题实体模板进行字段填充，具体可以包括：依据用户相关的疾病特征，对答案选项字段进行填充，不同的疾病特征可以对应不同的答案选项。具体地，可以在答案选项字段中填充用户相关的疾病特征对应的释义，不同的释义可以对应不同的答案选项。
上述对问题实体模板进行字段填充，具体可以包括：依据用户相关的疾病特征对应的命中动作属性，对跳转关系字段进行填充。具体地，可以在跳转关系字段中填充用户相关的疾病特征对应的命中动作属性。
例如，跳转关系字段的内容可以为：答案选项被选中的情况下，跳转至预设的问题实体，假设答案选项对应第一疾病特征实体，第一疾病特征实体的命中动作属性中记录有第二疾病特征实体，则预设的问题实体可以为：第二疾病特征实体对应的问题实体信息。
假设用户相关的疾病特征为第一疾病特征，则可以在问题实体模板的答案选项字段中填充第一疾病特征的信息，以及，在跳转关系字段中填充第二疾病特征实体对应的问题实体信息。
参照表5，示出了本申请实施例的一种问题实体模板的示意。该问题实体模板可以与症状类型的疾病特征相应，用于对症状类型的疾病特征进行问诊。
在问诊流程中，可以依据用户相关的症状1、症状2…症状N(N可以为大于0的自然数)，对该问题实体模板进行字段填充。
例如，在答案选项字段中填充症状1、症状2…症状N等症状的释义，以帮助用户确定是否命中对应的疾病特征，并选择对应的答案选项。
需要说明的是，本申请实施例中，在答案选项与疾病特征一一对应的情况下，“选中答案选项”与“选中疾病特征”可以为等同特征。例如，选中症状1对应的答案选项，可以等同于选中症状1。
又如，可以在疾病特征字段中填充症状类型。或者，在跳转关系字段中填充症状1、症状2…症状N的子症状对应的问题实体标识等。
可以理解，除了答案选项字段的填充之外，其他预设字段的填充是可选的，也即，可以不进行疾病特征字段、触发条件字段、以及跳转关系字段的填充。
可以理解，表5所示与症状类型的疾病特征相应的问题实体模板只是作为可选实施例，实际上，本领域技术人员还可以根据实际应用需求，采用与其他类型的疾病特征相应的问题实体模板。例如，还可以采用与接触史类型的疾病特征相应的问题实体模板，对应的问题文本可以包括：“是否接触过如下病原、有害因素、疾病患者？”等。
表5
Figure PCTCN2021103667-appb-000004
参照图5，示出了本申请实施例的一种知识图谱的处理方法的步骤流程图，具体可以包括如下步骤：
步骤501、依据疾病特征实体，确定问题实体；上述问题实体用于表征与上述疾病特征实体相关的问题；
步骤502、在知识图谱中，建立上述疾病特征实体与上述问题实体之间的关联。
步骤501中，疾病特征实体可以表征与疾病相关的特征，其可以包括：疾病中出现的疾病特征，也可以包括：疾病中不能出现的疾病特征。
在本申请的一种可选实施例中，确定疾病特征实体的过程可以包括：确定主诉列表，并确定所述主诉列表对应的疾病列表；依据医疗资源，对所述疾病列表中的疾病进行疾病特征的扩充；依据所述主诉列表中主诉、以及填充得到的疾病特征，确定疾病特征实体。
在医疗技术领域，主诉用于表征患者或代诊者对最主要的症状和(或)体征的叙述，主诉通常包括：患者或代诊者陈述的症状、体征、性质、以及持续时间等内容中的至少一种。
本申请实施例可以从医疗查询数据和/或病历数据等医疗资源中获取主诉，并依据获取的主诉建立主诉列表。
在具体实现中，可以根据主诉列表中的单个主诉，确定对应的疾病，进而将确定的疾病添加至疾病列表。一种实现方式可以为，向医生终端发送主诉，由医生终端对应的用户确定主诉对应的疾病。其中，医生终端的用户可以为具有M(M可以为大于0的自然数，例如M可以大于7)年以上临床经验的医生，其可以依据知识和经验，确定主诉对应的疾病。
疾病列表中的疾病可以作为知识图谱中疾病实体的数据源。也即，可以依据疾病列表中的疾病，构建对应的疾病实体。
在确定疾病列表后，本申请实施例可以依据医学书籍、医学数据库、医学问答数据等医疗资源，对疾病列表中的疾病进行疾病特征的扩充。也即，针对疾病，在疾病对应的主诉的基础上，对除了主诉之外的疾病特征进行扩充。扩充所涉及疾病特征的类型可以包括：症状、诱因、高发季节、接触史、家族史等。
可选地，可以从医疗资源中获取疾病对应的疾病内容，并从疾病内容中抽取上述类型对应的疾病特征。
上述主诉列表中主诉、以及填充得到的疾病特征，可以作为疾病特征实体对应疾病特征的数据源。也即，还可以依据主诉列表中主诉、以及填充得到的疾病特征，构建疾病特征实体。
在本申请的一种可选实施例中，还可以向医生终端发送疾病对应的候选疾病特征(主诉和填充得到的疾病特征)，以使医生终端的用户对候选疾病特征进行更新。上述对候选疾病特征进行更新具体可以包括：候选疾病特征的增加、候选疾病特征的删除、或者候选疾病特征的修改等。
更新后的候选疾病特征，可以作为疾病特征实体对应疾病特征的数据源。例如，可以针对疾病对应的更新后的候选疾病特征，确定疾病对应的特征集合。
在本申请的另一种可选实施例中，还可以针对疾病，确定特征集合中疾病特征在疾病的条件下的条件概率、和/或、疾病特征在疾病的条件下的惩罚因子。可选地，可以依据疾病特征在疾病对应医疗资源中的出现信息，确定上述条件概率或惩罚因子；或者，可由医生终端的用户确定上述条件概率或惩罚因子。
在本申请的再一种可选实施例中，还可以依据疾病及疾病所属疾病系统的医疗资源，确定疾病在疾病系统中的发病概率(临床占比)、和/或、系统概率。
上述条件概率能够表征疾病特征与疾病之间的匹配程度或疾病特征对疾病的重要程度，因此，将条件概率应用于疾病预测处理，能够在用户疾病特征对应多种候选疾病的情况下，提高多种候选疾病的区分度。
在本申请实施例的一种应用示例1中，对于出现“咳嗽”症状、“咳痰”症状的病人，在进行疾病预测的过程中，虽然“急性喉炎”和“支气管炎”都能与这两个症状匹配，但是“咳嗽”在“急性喉炎”中的概率为0.6，“咳痰”在“急性喉炎”中的概率为0.4；而“咳嗽”在“支气管炎”中的概率为0.8，“咳痰”在“支气管炎”中的概率为0.6。由于上述条件概率，可以确定这两个症状与“支气管炎”之间的匹配程度高于这两个症状与“急性喉炎”之间的匹配程度，进而能够在用户疾病特征对应多种候选疾病的情况下，提高多种候选疾病之间的区分度。
上述惩罚因子能够表征疾病特征对于疾病的排除程度，进而能够综合确定多个疾病特征对于疾病的影响。例如，用户出现了不应该出现在某候选疾病中的特征，可以依据该惩罚因子降低该候选疾病的概率，例如可以依据该惩罚因子降低该疾病的得分，能够提高在用户疾病特征的条件下候选疾病的概率的准确度；进而能够在用户疾病特征对应多种候选疾病的情况下，提高多种候选疾病之间的区分度。
临床占比能够表征疾病在疾病系统中的发病概率，可以反映对应疾病的常见性。将临床占比应用于疾病预测处理，能够提高在用户疾病特征的条件下候选疾病的概率的准确度。例如，在用户疾病特征对应多种候选疾病的情况下，可以依据多种候选疾病分别对应的临床占比，对多种候选疾病进行排序；进而能够在用户疾病特征对应多种候选疾病的情况下，提高多种候选疾病之间的区分度。
在将临床占比和系统概率应用于疾病预测的过程中，可以依据临床占比和系统概率确定候选疾病的先验概率，进而可以依据该先验概率对多种候选疾病进行排序。这样，可以提高在用户疾病特征的条件下候选疾病的概率的准确度；进而能够在用户疾病特征 对应多种候选疾病的情况下，提高多种候选疾病之间的区分度。
在本申请的一种可选实施例中，上述确定疾病特征实体，具体可以包括：对主诉、以及填充得到的疾病特征进行特征归一化，以得到归一化后的疾病特征；并依据归一化后的疾病特征，确定疾病特征实体。
上述特征归一化可以将语义相同且描述不同的疾病特征，统一为标准描述。例如，“头疼”对应的个性化或口语化的症状描述具体包括：“针扎一样疼”、“一抽一抽的痛”、“一摸就疼”、“咽口水跟着疼”等。又如，“舌痛””对应的个性化或口语化的症状描述具体包括：“舌头左边疼”、“舌尖疼痛”、“舌头根疼”、“舌头边缘痛”等。
本申请实施例中，确定疾病特征实体，具体可以包括：确定疾病特征实体对应的多个属性，并针对特定的疾病特征实体，确定对应的属性值。疾病特征实体对应的多个属性具体可以包括：特征标识属性、类型属性、从属关系属性、频次属性、命中动作属性、释义属性等。
本申请实施例的问题实体，用于对疾病特征实体对应的疾病症状进行问诊，以帮助用户确定是否出现对应的疾病症状。
本领域技术人员可以根据实际应用需求，确定疾病特征实体对应的问题实体。根据一种实施例，可以向医生终端发送疾病特征实体的信息，以使医生设置疾病特征实体对应的问题实体。
根据另一种实施例，可以依据疾病特征实体对应的类型、以及历史问诊数据，确定疾病特征实体对应的问题实体。
本申请实施例中，问题实体的字段可以包括：问题文本字段和答案选项字段，其中，问题文本字段或答案选项字段中可以包括疾病特征实体的标识或释义等信息。
可选地，问题实体的字段还可以包括如下字段中的至少一种：疾病特征字段、触发条件字段、以及跳转关系字段；
其中，所述疾病特征字段用于表征疾病特征实体；
所述触发条件字段用于表征对应的问题实体依据疾病特征实体触发得到；
所述跳转关系字段用于选中答案选项的条件下执行预设跳转。
可选地，所述跳转关系字段用于在选中答案选项的条件下，从第一问题实体跳转至第二问题实体，第一问题实体对应的疾病特征实体与所述第二问题实体对应的疾病特征实体为父子关系。
当然，本领域技术人员可以根据实际应用需求，确定预设跳转，例如，预设跳转还可以包括：执行预设函数，预设函数可用于结束问题实体的查找等。
本申请实施例还可以向医生终端发送问题实体的信息，以使医生终端的用户对问题实体的信息进行审核。
步骤502中，问题实体可以包括：问题实体实例，问题实体实例可以与预设的疾病特征相应。因此，本申请实施例可以在知识图谱中，建立预设的疾病特征实体与问题实体实例之间的关联。
本申请实施例的问题实体可以包括：问题实体模板，问题实体模板可以与预设类型的疾病特征相应。因此，本申请实施例可以在知识图谱中，建立预设类型的疾病特征实体与问题实体实例之间的关联。
本申请实施例中，建立疾病特征实体与问题实体之间的关联，具体可以包括：依据问题实体中的疾病特征字段，建立疾病特征实体与问题实体之间的映射关系，其中，问题实体中的疾病特征字段与疾病特征实体相匹配。
若问题实体中的疾病特征字段表征预设的疾病特征，则映射关系中的问题实体为问题实体实例，映射关系中的疾病特征实体对应一种预设的疾病特征。
若问题实体中的疾病特征字段表征预设类型的疾病特征，则映射关系中的问题实体为问题实体模板，映射关系中的疾病特征实体对应预设类型的多种疾病特征。
在本申请的一种可选实施例中，还可以在知识图谱中建立疾病实体与疾病特征实体之间的关联。具体地，可以将疾病实体与其特征集合中疾病特征对应的疾病特征实体进行关联，这样，可以对用户疾病特征与疾病对应的特征集合进行匹配。可以理解，本申请实施例对于疾病实体与疾病特征实体之间的具体关联方式不加以限制。
综上，本申请实施例的知识图谱的处理方法，在知识图谱中建立疾病特征实体与问题实体之间的关联。这样，在问诊流程中，可以依据用户相关的疾病特征，在知识图谱中查找对应的问题实体，并依据查找得到的问题实体，得到用于问诊的问题。由于本申请实施例的知识图谱中包括疾病特征实体与问题实体之间的关联，可以在问诊流程中生成用于问诊的问题，因此能够提升知识图谱对问诊的作用，进而能够提高依据该知识图谱进行问诊信息处理的效率和准确率。
并且，本申请实施例在疾病特征实体中通过命中动作属性表征了：疾病特征与问题实体的信息之间的关系，在对应疾病特征被选中的情况下，将触发对应的问题实体。由于可以自动进行疾病特征与问题实体之间的关联，因此不仅能够降低人工运营问诊路径所花费的资源成本，而且可以提高问诊流程中问题接序的合理性。
另外，本申请实施例在问题实体中通过跳转关系字段表征了：疾病特征对应的答案选项被选中的情况下、将跳转至预设的问题实体，这样可以提高问诊流程中问题接序的 合理性。由于可以自动进行疾病特征与问题实体之间的关联，因此不仅能够降低人工运营问诊路径所花费的资源成本，而且可以提高问诊流程中问题接序的合理性。
再者，本申请实施例在疾病实体中设置了条件概率、惩罚因子、临床占比、系统概率等属性参数。其中，条件概率能够表征疾病特征与疾病之间的匹配程度或疾病特征对疾病的重要程度；惩罚因子能够表征疾病特征对于疾病的排除程度，进而能够综合确定多个疾病特征对于疾病的影响；临床占比能够表征疾病在疾病系统中的发病概率，可以反映对应疾病的常见性；将上述属性参数中的一种或多种应用于疾病预测处理，可以反映对应候选疾病的概率，因此可以提高在用户疾病特征的条件下候选疾病的概率的准确度。
方法实施例三
本实施例对问诊信息处理过程中的疾病预测处理进行说明。
本申请实施例可以依据基于用户输入得到的用户疾病特征，在知识图谱中包括的疾病实体与疾病特征实体之间的关联关系中进行查找，以得到候选疾病。具体地，可以将用户疾病特征与关联关系中的疾病特征实体进行匹配，将匹配成功的疾病实体对应的疾病作为候选疾病。
本申请实施例中，用户疾病特征可以与至少一种候选疾病相匹配，本申请实施例可以依据候选疾病的得分表征候选疾病的概率，并依据候选疾病的得分，对候选疾病进行筛选。
在本申请的一种可选实施例中，可以依据概率特征，确定所述候选疾病的得分；
上述概率特征可以为知识图谱中记录的特征，具体可以包括如下特征中的至少一种：
与所述用户疾病特征相匹配的疾病特征在候选疾病的条件下的条件概率；
与所述用户疾病特征相匹配的疾病特征在候选疾病的条件下的惩罚因子；
候选疾病在疾病系统中的发病概率；
以及疾病系统的发病概率。
根据一种实施例，可以依据上述条件概率，确定候选疾病的得分。
在本申请实施例的一种应用示例1中，用户疾病特征包括：“咳嗽”和“咳痰”这两个症状，在进行疾病预测的过程中，虽然“急性喉炎”和“支气管炎”都能与这两个症状匹配，但是“咳嗽”在“急性喉炎”中的概率为0.6，“咳痰”在“急性喉炎”中的概率为0.4；而“咳嗽”在“支气管炎”中的概率为0.8，“咳痰”在“支气管炎”中的概率为0.6。依据上述条件概率，可以确定这两个症状与“支气管炎”之间的匹配程度高于这两个症状与“急性喉炎”之间的匹配程度，因此可以确定“支气管炎”的得分高于“急 性喉炎”的得分。
该惩罚因子可以对应在候选疾病的条件下不能出现的疾病特征，用于在疾病预测的过程中，对候选疾病的概率进行惩罚。例如，候选疾病包括：候选疾病A，但用户疾病特征包括：不应该出现在候选疾病A中的疾病特征X，此种情况下，可以依据疾病特征X和候选疾病A对应的惩罚因子，减少候选疾病A的得分。
临床占比能够表征疾病在疾病系统中的发病概率，可以反映对应疾病的常见性。将临床占比应用于疾病预测处理，能够提高疾病预测的得分的准确度。通常，临床占比越高，则对应候选疾病的得分越高。
在将临床占比和系统概率应用于疾病预测的过程中，可以依据临床占比和系统概率确定候选疾病的先验概率，进而可以依据该先验概率，确定候选疾病的得分。通常候选疾病的先验概率越高，则候选疾病的得分越高。
在采用多种概率特征的情况下，可以对多种概率特征进行融合，并依据融合概率特征，确定候选疾病的得分。对应的融合方式可以包括：加权平均方式、或者乘积方式等。
本申请实施例依据候选疾病的得分，对候选疾病进行筛选，具体可以包括：选取得分大于得分阈值的候选疾病，和/或，选取得分排在前P(P可以为大于0的自然数)位的候选疾病。
假设筛选后的候选疾病为目标候选疾病，则可以依据目标候选疾病对应的疾病特征，生成目标问题，或者，可以将目标候选疾病的信息，作为问诊信息处理结果进行输出。
方法实施例四
本实施例对问诊信息处理的停止条件进行说明。
停止条件可以表征停止问诊信息处理对应的条件。在停止问诊信息处理的情况下，可以停止向用户输出目标问题，还可以向用户输出对应的问诊信息处理结果。
相应地，上述方法还可以包括：在得到候选疾病后，若符合停止条件，则停止执行所述生成目标问题的步骤。
相应地，上述方法还可以包括：在得到候选疾病后，若符合停止条件，则输出问诊信息处理结果；所述问诊信息处理结果可以包括：候选疾病的信息。可以在问诊信息处理结果中携带一个或多个目标候选疾病的信息，如得分最高的目标候选疾病的名称及得分。
可选地，上述问诊信息处理结果还可以包括：用户出现的疾病特征、以及用户未出现的疾病特征。还可以在问诊信息处理结果中携带用户出现的疾病特征在目标候选疾病的条件下的条件概率。
可以理解，本领域技术人员可以根据实际应用需求，在问诊信息处理结果中携带所需的信息，本申请实施例对于在问诊信息处理结果中携带的具体信息不加以限制。
本申请实施例中，可选的是，上述停止条件具体可以包括如下条件中的至少一种：
至少一种候选疾病的得分大于得分阈值，可用于提高问诊信息处理结果中携带的目标候选疾病的质量；
多种候选疾病的得分差异符合差异条件；差异条件可以表征至少两种候选疾病的得分存在差异。在不符合差异条件的情况下，说明用户疾病特征对应得分接近的多种候选疾病，此种情况下，还需要进一步的用户疾病特征对得分接近的多种候选疾病进行筛选。
候选疾病对应疾病特征的询问比例符合比例条件，询问比例可以表征已询问的疾病特征相对于所有疾病特征的比例，比例条件可以为询问比例大于比例阈值等；以及
询问轮数超过轮数阈值，在询问轮数超过轮数阈值的情况下，停止问诊，可以节省问诊时间，提升用户体验。
需要说明的是，在不符合停止条件的情况下，可以从问题队列中选取问诊问题，并向用户输出选取的问诊问题；若问题队列中不包含问题，则可以依据上述候选疾病对应的疾病特征，生成目标问题，并将生成的目标问题保存至问题队列。
参照图6，示出了本申请的一种问诊信息处理方法实施例四的步骤流程图，具体可以包括如下步骤：
步骤601、针对用户，建立问题队列，该问题队列中可以包括：预设问题；
步骤602、依据至少一次用户输入，确定用户疾病特征；
步骤603、对上述用户疾病特征进行疾病预测处理，以得到对应的候选疾病；
步骤604、判断是否符合停止条件，若是，则执行步骤608，否则执行步骤605；
步骤605、判断问题队列中是否包含问题，若是，则执行步骤606，否则执行步骤607；
步骤606、从问题队列中选取问诊问题，以向用户输出问诊问题；
步骤607、依据候选疾病对应的疾病特征，生成目标问题，并将该目标问题加入问题队列；
步骤608、输出问诊信息处理结果，并结束问诊。
方法实施例五
本实施例对生成目标问题的过程进行说明。
本申请实施例在知识图谱中建立疾病特征实体与问题实体之间的关联。这样，在问诊流程中，可以依据候选疾病对应的疾病特征，在知识图谱中查找对应的问题实体，并依据查找得到的问题实体，得到用于问诊的目标问题。由于本申请实施例的知识图谱中 包括疾病特征实体与问题实体之间的关联，可以在问诊流程中自动生成用于问诊的问题，因此能够提升知识图谱对问诊的作用，进而能够提高问诊效率和用于问诊的目标问题的准确度。
本申请实施例可以依据疾病特征对应的疾病特征实体，在知识图谱包括的疾病特征实体与问题实体之间的映射关系中进行查找，以得到疾病特征对应的问题实体。
本申请实施例可以提供生成目标问题的如下技术方案：
技术方案1、
技术方案1中，可以依据疾病特征实体与问题实体实例之间的映射关系，确定疾病特征对应的问题实体实例。
在具体实现中，可以从目标疾病特征对应的问题实体实例中获取问题文本字段的内容和答案选项字段的内容，以得到目标问题；也即，目标问题可以包括：问题文本字段的内容和答案选项字段的内容。例如，在目标疾病特征为“呕吐”的情况下，可以依据“呕吐”对应的问题实体实例，确定目标问题A。目标问题A的问题文本可以为“是否呕吐”，目标问题A的答案选项可以为：[是，否，我不知道]。
在目标疾病特征对应的答案选项被选中的情况下，可以依据所述目标疾病特征对应的疾病特征实体中的命中动作属性，从知识图谱中获取对应的问题实体。
例如，目标疾病特征为“呕吐”，在目标问题A对应的答案选项[是]被选中的情况下，根据表2所示疾病特征实体的命中动作属性，将触发问题实体标识为18的问题实体实例。此种情况下，可以依据问题实体标识为18的问题实体实例，生成目标问题B。目标问题B的问题文本可以为“是否没有恶心动作直接呕吐？”，目标问题B的答案选项可以为：[是，否，我不知道]。
技术方案2、
技术方案2中，可以依据疾病特征实体的类型与问题实体模板之间的映射关系，确定疾病特征对应的问题实体模板。
此种情况下，上述生成目标问题，具体可以包括：依据上述疾病特征，对上述问题实体模板进行字段填充，以得到目标问题。
问题实体模板的字段具体包括：问题文本字段和答案选项字段；上述对所述问题实体模板进行字段填充，具体包括：步骤S1和/或步骤S2。步骤S1和步骤S2的执行顺序不分先后。
其中，步骤S1可以依据属于一种类型的至少一种疾病特征，对所述类型对应的问题实体模板进行字段填充，以在得到的目标问题中携带至少一种疾病特征的信息。例如， 类型为“症状”，则可以在“症状”对应的问题实体模板中填充症状1、症状2…症状N等症状的释义，以帮助用户确定是否命中对应的疾病特征，并选择对应的答案选项。
步骤S2可以依据所述疾病特征对应疾病特征实体中的命中动作属性，对问题实体模板的跳转关系字段进行填充。具体地，可以在跳转关系字段中填充疾病特征对应命中动作属性的信息。
例如，跳转关系字段的内容可以为：答案选项被选中的情况下，跳转至预设的问题实体，假设答案选项对应第一疾病特征实体，第一疾病特征实体的命中动作属性中记录有第二疾病特征实体，则预设的问题实体可以为：第二疾病特征实体对应的问题实体信息。
假设疾病特征为第一疾病特征，则可以在问题实体模板的答案选项字段中填充第一疾病特征的信息，以及，在跳转关系字段中填充第二疾病特征实体对应的问题实体信息。
技术方案3、
技术方案3中，可以在疾病特征对应的答案选项被选中的情况下，依据所述疾病特征对应问题实体中的跳转关系字段，从知识图谱中获取对应的问题实体。
所述跳转关系字段用于选中答案选项的条件下执行预设跳转。可选地，所述跳转关系字段用于在选中答案选项的条件下，从第一问题实体跳转至第二问题实体，第一问题实体对应的疾病特征实体与所述第二问题实体对应的疾病特征实体可以为父子关系。
例如，在目标问题B对应的答案选项[是]被选中的情况下，可以根据问题实体的跳转关系字段，触发“嗜睡”等症状对应的问题实体实例。此种情况下，可以依据“嗜睡”等症状对应的问题实体实例，生成目标问题C。目标问题C的问题文本可以为“是否嗜睡？”，目标问题C的答案选项可以为：[是，否，我不知道]。
技术方案1基于问题实体实例确定目标问题；技术方案2基于问题实体模板的填充确定目标问题；技术方案3可以通过问题实体中的跳转关系字段，获取问题实体。可以理解，本领域技术人员可以根据实际应用需求，采用技术方案1至技术方案3中的任一或组合。
综上，本申请实施例的问诊信息处理方法，在知识图谱中建立疾病特征实体与问题实体之间的关联。这样，在问诊流程中，可以依据候选疾病对应的疾病特征，在知识图谱中查找对应的问题实体，并依据查找得到的问题实体，得到用于问诊的目标问题。由于本申请实施例的知识图谱中疾病特征实体与问题实体之间的关联，可以在问诊流程中生成用于问诊的目标问题，因此能够提升知识图谱对问诊的作用，进而能够提高问诊效率和目标问题的合理性。
并且，本申请实施例在疾病特征实体中通过命中动作属性表征了：疾病特征与问题实体的信息之间的关系，在对应疾病特征被选中的情况下，将触发对应的问题实体。由于可以自动进行疾病特征与问题实体之间的关联，因此不仅能够降低人工运营问诊路径所花费的资源成本，而且可以提高问诊流程中问题接序的合理性。
另外，本申请实施例在问题实体中通过跳转关系字段表征了：疾病特征对应的答案选项被选中的情况下、将跳转至预设的问题实体，这样可以提高问诊流程中问题接序的合理性。由于可以自动进行疾病特征与问题实体之间的关联，因此不仅能够降低人工运营问诊路径所花费的资源成本，而且可以提高问诊流程中问题接序的合理性。
方法实施例六
参照图7，示出了本申请的一种问诊信息处理方法实施例六的步骤流程图，具体可以包括如下步骤：
步骤701、依据基于至少一次用户输入得到的用户疾病特征，从知识图谱中获取预设疾病特征实体对应的问题实体；上述问题实体用于表征与上述预设疾病特征实体相关的问题；
具体的，从知识图谱中获取与用户疾病特征相匹配的预设疾病特征实体，然后，获取与相匹配的预设疾病特征实体对应的问题实体。
步骤702、依据上述问题实体，确定预设疾病问题；上述预设疾病问题具体包括：问题文本、以及至少一个预设选项；
步骤703、若接收到用户针对任一预设选项的选择操作，则输出对应的就医建议信息。
本申请实施例中，预设疾病特征实体可以表征预设疾病对应的疾病特征，预设疾病可以包括：危急程度和/或严重程度较高的疾病，预设疾病对应的疾病特征可以包括：危急重症。
本申请实施例在问诊过程中，依据预设疾病特征实体对应的问题实体，确定预设疾病问题，并利用预设疾病问题对用户进行问诊；可以在问诊过程中进行危急程度和/或严重程度较高的预设疾病的排除，进而可以提高问诊信息处理的安全性。
根据一种实施例，可以建立疾病特征与预设疾病特征实体对应的问题实体之间的对应关系，这样，可以将用户疾病特征与对应关系中的疾病特征进行匹配，以得到用户疾病特征和预设疾病特征实体对应的问题实体。
根据另一种实施例，可以在疾病特征实体中通过命中动作属性表征：疾病特征与预设疾病特征实体对应的问题实体的信息之间的关系。这样，可以依据用户疾病特征，查找对应疾病特征实体中的命中动作属性，进而可以得到用户疾病特征和预设疾病特征实体对应的问题实体。
在本申请的一种应用示例2中，假设用户疾病特征为“发热”，则本申请实施例可以 依据上述对应关系、或者“发热”实体对应的命中动作属性，确定与“发热”相关的预设疾病对应的问题实体。例如，问题实体中可以包括预设疾病问题，预设疾病问题的问题文本可以为“有以下哪些症状？”，预设疾病问题的至少一个预设选项可以包括：[预设疾病特征对应的选项，以上都不是]。预设疾病特征对应的选项可以包括：[超过40度，寒战]等。
若接收到用户针对任一预设选项的选择操作，则说明用户命中了预设疾病特征，说明用户对应预设疾病的概率较高，此种情况下，输出就医建议信息，可以提高用户对于病情的重视程度，进而可以增强问诊信息处理的安全性。
综上，本申请实施例的知识图谱的处理方法，在问诊过程中，依据预设疾病特征实体对应的问题实体，确定预设疾病问题，并利用预设疾病问题对用户进行问诊；可以在问诊过程中进行危急程度和/或严重程度较高的预设疾病的排除，进而可以提高问诊信息处理的安全性。
方法实施例七
参照图8，示出了本申请的一种问诊信息处理方法实施例七的步骤流程图，具体可以包括如下步骤：
步骤801、接收用户输入的主诉；
步骤802、依据上述主诉，从知识图谱中获取预设疾病特征实体对应的问题实体；上述问题实体用于表征与上述预设疾病特征实体相关的问题；
步骤803、依据上述问题实体，确定预设疾病问题；上述预设疾病问题具体包括：问题文本、以及至少一个预设选项；
步骤804、若接收到用户针对任一预设选项的选择操作，则输出对应的就医建议信息；
步骤805、若用户未选择任一预设选项，则对用户疾病特征进行疾病预测处理，以得到对应的候选疾病；
步骤806、依据上述候选疾病对应的疾病特征，生成目标问题，上述目标问题用于对用户进行问诊。
本申请实施例在问诊流程中，在向用户问诊之前，可以首先依据主诉进行预设疾病的排除，若接收到用户针对任一预设选项的选择操作，则说明用户患有预设疾病的概率较高，此种情况下，可以输出对应的就医建议信息，以提高问诊信息处理的安全性。
若用户未选择任一预设选项，则说明用户患有预设疾病的概率较低，可以向用户问诊。
方法实施例八
参照图9，示出了本申请的一种问诊信息处理方法实施例八的步骤流程图，具体可以包括如下步骤：
步骤901、提供知识图谱；
步骤902、将至少一次用户输入转化为用户症状特征，以得到用户症状特征集合；
步骤903、查找知识图谱，获得用户症状对应的危急重症问题，根据用户的回复判断是否对应有危急重症，若有则执行步骤904，否则执行步骤905；
步骤904、输出对应的就医建议信息；
步骤905、对用户症状特征集合进行疾病预测处理，以得到对应的候选疾病及其得分；
步骤906、依据候选疾病及其得分，判断是否符合停止条件，若是，则执行步骤907，否则，执行步骤908；
步骤907、输出问诊信息处理结果，并结束问诊信息处理；
步骤908、依据候选疾病对应的症状特征、以及知识图谱，对候选疾病对应的症状特征进行排序；
步骤909、依据排序结果从候选疾病对应的症状特征中确定出目标症状特征，并依据目标症状特征，生成目标问题，该目标问题用于收集用户症状等信息。
步骤902中，可以利用实体识别方法、或疾病特征表的匹配方法等症状识别方法，将至少一次用户输入转化为对应的用户症状特征。
在实际应用中，步骤903获得用户症状对应的危急重症问题，具体可以包括：依据用户症状特征集合，从知识图谱中获取对应的危急重症问题对应的问题实体；所述问题实体表示了一个问题所包含的各类信息，包括问题文本、答案选项、跳转关系(也即，答案选项被选中时的对应动作等)；
依据所述问题实体，确定危急重症问题；该危急重症问题的答案选项可以包括：至少一个危急重症选项；
若接收到用户针对任一危急重症选项的选择操作，则可以说明上述用户症状特征集合对应有危急重症。
在实际应用中，步骤905进行疾病预测的过程，具体包括：依据用户症状特征集合与医疗知识图谱中疾病对应的症状等特征之间的匹配信息，确定到当前的用户症状特征集合对应的候选疾病。
在实际应用中，步骤905进行疾病预测的过程，具体包括：依据概率特征，确定所述候选疾病的得分；
上述概率特征具体可以包括如下特征中的至少一种：
与所述用户症状特征集合相匹配的疾病特征在候选疾病的条件下的条件概率；
与所述用户症状特征集合相匹配的疾病特征在候选疾病的条件下的惩罚因子；
候选疾病在解剖学系统的所有疾病中的临床占比；
以及各解剖学系统对应疾病的临床占比。
步骤906中，停止条件可以包括如下条件中的至少一种：
至少一种候选疾病的得分大于得分阈值；
多种候选疾病的得分差异符合差异条件；
所述候选疾病对应症状特征的询问比例符合比例条件；以及
询问轮数超过轮数阈值。
在实际应用中，步骤908对候选疾病对应的症状特征进行排序的过程，具体包括：依据所述候选疾病所包含的症状特征对应的疾病特征实体在各个疾病中的条件概率、以及候选疾病的先验概率等信息，计算各个症状特征的重要性得分，并依据重要性得分，对候选疾病所包含的症状特征进行排序。可选地，可以按照预设的标准，依据排序结果选择符合标准的目标症状特征进行询问。例如，预设的标准可以包括：重要性得分大于第二阈值，或者按照重要性得分从高到低的顺序，排在前X(X可以为大于0的自然数)位等。
在实际应用中，步骤909依据目标症状特征，生成目标问题，具体可以包括：按照类型，对目标症状特征进行归并，从知识图谱中获取对应类型的问题实体；依据所述问题实体和属于该类型的目标症状特征，生成目标问题。
具体地，可以依据属于一种类型的至少一种目标症状特征，对该类型对应的问题实体模板进行字段填充，以在得到的目标问题中携带至少一种目标症状特征的信息。
综上，本申请实施例的问诊信息处理方法，具有如下技术效果：
首先，本申请实施例支持答案选项的选择输入、症状输入、短句输入等多种输入方式，对于任意输入方式的用户输入，都能通过症状识别方法，将用户输入转化为知识图谱中具有标准描述的症状特征，这样，可以提升用户体验。
其次，本申请实施例在向用户问诊之前，可以进行危急重症的排除，在用户症状特征对应危急重症的情况下，建议用户就医，能够提高问诊信息处理的安全性。
再者，本申请实施例的疾病预测和生成目标问题，可以为动态过程；因此，能够依据问诊过程中用户疾病特征的累积，得到与用户疾病特征更相关的目标问题，因此可以提高问诊的合理性。
进一步，本申请实施例根据候选疾病对应症状特征的询问比例、询问轮数、以及用户症状集合到候选疾病的确定依据的充分性，设置停止条件，在符合停止条件的情况下，结束问诊，不仅能够提高问诊的灵活性，而且能够缩短询问轮数，进而能够提高问诊效率和提升用户体验。
需要说明的是，对于方法实施例，为了简单描述，故将其都表述为一系列的动作组合，但是本领域技术人员应该知悉，本申请实施例并不受所描述的动作顺序的限制，因为依据本申请实施例，某些步骤可以采用其他顺序或者同时进行。其次，本领域技术人员也应该知悉，说明书中所描述的实施例均属于优选实施例，所涉及的动作并不一定是本申请实施例所必须的。
装置实施例
参照图10，示出了本申请的一种问诊信息处理装置实施例的结构框图，具体可以包括：用户疾病特征确定模块1001、用户疾病特征处理模块1002和问题生成模块1003。
其中，用户疾病特征确定模块1001，配置为依据至少一次用户输入，确定用户疾病特征；
用户疾病特征处理模块1002，配置为对所述用户疾病特征进行疾病预测处理，以得到对应的候选疾病；以及
问题生成模块1003，配置为依据所述候选疾病对应的疾病特征，生成目标问题，所述目标问题用于对用户进行问诊。
可选地，用户疾病特征处理模块1002可以包括：
候选疾病确定模块，配置为依据上述用户疾病特征与疾病对应的疾病特征之间的匹配信息，确定上述用户疾病特征对应的候选疾病。
可选地，问题生成模块1003可以包括：
问题实体获取模块，配置为依据上述疾病特征对应的疾病特征实体，从知识图谱中获取对应的问题实体；上述问题实体用于表征与上述疾病特征实体相关的问题；
第一问题生成模块，配置为依据上述问题实体对应的问题，生成目标问题。
可选地，上述问题实体可以包括：问题实体模板；上述第一问题生成模块可以包括：
字段填充模块，配置为依据上述疾病特征，对上述问题实体模板进行字段填充，以得到目标问题。
可选地，上述字段填充模块可以包括：
第一字段填充模块，配置为依据属于一种类型的至少一种疾病特征，对上述类型对应的问题实体模板进行字段填充，以在得到的目标问题中携带至少一种疾病特征的信息。
可选地，上述问题实体模板的字段可以包括：问题文本字段和答案选项字段；
上述字段填充模块可以包括：
第二字段填充模块，配置为在上述问题实体模板的答案选项字段中填充上述疾病特征的信息。
可选地，上述字段填充模块可以包括：
第三字段填充模块，配置为依据上述疾病特征对应疾病特征实体中的命中动作属性，对上述问题实体模板的跳转关系字段进行填充。
可选地，上述问题实体可以包括：问题实体实例；上述问题生成模块可以包括：
第二问题生成模块，配置为从上述疾病特征对应的问题实体实例中，获取目标问题。
可选地，上述问题实体获取模块可以包括：
第一问题实体获取模块，配置为在上述疾病特征对应的答案选项被选中的情况下，依据上述疾病特征对应的疾病特征实体中的命中动作属性，从知识图谱中获取对应的问 题实体；和/或，
第二问题实体获取模块，配置为在上述疾病特征对应的答案选项被选中的情况下，依据上述疾病特征对应问题实体中的跳转关系字段，从知识图谱中获取对应的问题实体。
可选地，上述问题生成模块1003可以包括：
目标疾病特征确定模块，配置为依据疾病特征对应的重要性得分，从上述候选疾病对应的疾病特征中确定出目标疾病特征；上述目标疾病特征配置为表征本轮问诊向用户询问的疾病特征；
实体确定模块，配置为依据上述目标疾病特征对应的类型，从知识图谱中获取对应类型的问题实体；上述问题实体用于表征与上述疾病特征实体相关的问题；
第三问题生成模块，配置为依据上述问题实体、以及属于上述类型的目标疾病特征，生成目标问题。
可选地，上述至少一次用户输入可以包括：
主动输入；或者
主动输入和针对预设问题的回复；或者
主动输入和针对上述目标问题的回复；或者
主动输入、以及针对预设问题和目标问题的回复。
可选地，上述目标问题可以包括：问题文本和答案选项；
上述至少一次用户输入，可以包括：用户选择的答案选项。
可选地，上述装置还可以包括：
预设问题实体获取模块，配置为依据上述用户疾病特征，从知识图谱中获取预设疾病特征实体对应的问题实体；上述问题实体用于表征与上述预设疾病特征实体相关的问题；
预设疾病问题确定模块，配置为依据上述问题实体，确定预设疾病问题；上述预设疾病问题可以包括：问题文本、以及至少一个预设选项；
建议输出模块，配置为若接收到用户针对任一预设选项的选择操作，则输出对应的就医建议信息。
可选地，上述装置还可以包括：
得分确定模块，配置为依据概率特征，确定上述候选疾病的得分；
上述概率特征可以包括如下特征中的至少一种：
与上述用户疾病特征相匹配的疾病特征在候选疾病的条件下的条件概率；
与上述用户疾病特征相匹配的疾病特征在候选疾病的条件下的惩罚因子；
候选疾病在疾病系统中的发病概率；
以及疾病系统的发病概率。
可选地，上述装置还可以包括：
停止模块，配置为在得到候选疾病后，若符合停止条件，则通知上述问题生成模块停止执行上述生成目标问题。
可选地，上述装置还可以包括：
处理结果输出模块，配置为在得到候选疾病后，若符合停止条件，则输出问诊信息处理结果；上述问诊信息处理结果可以包括：上述候选疾病的信息。
可选地，上述问诊信息处理结果还可以包括：用户出现的疾病特征、以及用户未出现的疾病特征。
可选地，上述停止条件可以包括如下条件中的至少一种：
至少一种候选疾病的得分大于得分阈值；
多种候选疾病的得分差异符合差异条件；
上述候选疾病对应疾病特征的询问比例符合比例条件；以及
询问轮数超过轮数阈值。
综上，本申请实施例的问诊信息处理装置，具有如下技术效果：
首先，本申请实施例支持答案选项的选择输入、症状输入、短句输入等多种输入方式，对于任意输入方式的用户输入，都能通过症状识别方法，将用户输入转化为知识图谱中具有标准描述的症状特征，这样，可以提升用户体验。
其次，本申请实施例在向用户问诊之前，可以进行危急重症的排除，在用户症状特征对应危急重症的情况下，建议用户就医，能够提高问诊信息处理的安全性。
再者，本申请实施例的疾病预测和生成目标问题，可以为动态过程；因此，能够依据问诊过程中用户疾病特征的累积，得到与用户疾病特征更相关的目标问题，因此可以提高问诊的合理性。
进一步，本申请实施例根据候选疾病对应症状特征的询问比例、询问轮数、以及用户症状集合到候选疾病的确定依据的充分性，设置停止条件，在符合停止条件的情况下，结束问诊，不仅能够提高问诊的灵活性，而且能够缩短询问轮数，进而能够提高问诊效率和提升用户体验。
对于装置实施例而言，由于其与方法实施例基本相似，所以描述的比较简单，相关之处参见方法实施例的部分说明即可。
本说明书中的各个实施例均采用递进的方式描述，每个实施例重点说明的都是与其他实施例的不同之处，各个实施例之间相同相似的部分互相参见即可。
关于上述实施例中的装置，其中各个模块执行操作的具体方式已经在有关该方法的实施例中进行了详细描述，此处将不做详细阐述说明。
本申请实施例提供了一种用于处理问诊信息的装置，包括有存储器，以及一个或者一个以上的程序，其中一个或者一个以上程序存储于存储器中，且经配置以由一个或者 一个以上处理器执行所述一个或者一个以上程序包含用于进行以下操作的指令：依据至少一次用户输入，确定用户疾病特征；对所述用户疾病特征进行疾病预测处理，以得到对应的候选疾病；依据所述候选疾病的疾病特征对应的疾病特征实体，从知识图谱中获取对应的问题实体；所述问题实体用于表征与所述疾病特征实体相关的问题；依据所述问题实体对应的问题，生成目标问题；所述目标问题用于对用户进行问诊。
图11是根据一示例性实施例示出的一种用于处理问诊信息的装置1100的框图。例如，装置1100可以是移动电话，计算机，数字广播终端，消息收发设备，游戏控制台，平板设备，医疗设备，健身设备，个人数字助理等。
参照图11，装置1100可以包括以下一个或多个组件：处理组件1102，存储器1104，电源组件1106，多媒体组件1108，音频组件1110，输入/输出(I/O)的接口1112，传感器组件1114，以及通信组件1116。
处理组件1102通常控制装置1100的整体操作，诸如与显示，电话呼叫，数据通信，相机操作和记录操作相关联的操作。处理元件1102可以包括一个或多个处理器1120来执行指令，以完成上述的方法的全部或部分步骤。此外，处理组件1102可以包括一个或多个模块，便于处理组件1102和其他组件之间的交互。例如，处理组件1102可以包括多媒体模块，以方便多媒体组件1108和处理组件1102之间的交互。
存储器1104被配置为存储各种类型的数据以支持在设备1100的操作。这些数据的示例包括用于在装置1100上操作的任何应用程序或方法的指令，联系人数据，电话簿数据，消息，图片，视频等。存储器1104可以由任何类型的易失性或非易失性存储设备或者它们的组合实现，如静态随机存取存储器(SRAM)，电可擦除可编程只读存储器(EEPROM)，可擦除可编程只读存储器(EPROM)，可编程只读存储器(PROM)，只读存储器(ROM)，磁存储器，快闪存储器，磁盘或光盘。
电源组件1106为装置1100的各种组件提供电力。电源组件1106可以包括电源管理系统，一个或多个电源，及其他与为装置1100生成、管理和分配电力相关联的组件。
多媒体组件1108包括在所述装置1100和用户之间的提供一个输出接口的屏幕。在一些实施例中，屏幕可以包括液晶显示器(LCD)和触摸面板(TP)。如果屏幕包括触摸面板，屏幕可以被实现为触摸屏，以接收来自用户的输入信号。触摸面板包括一个或多个触摸传感器以感测触摸、滑动和触摸面板上的手势。所述触摸传感器可以不仅感测触摸或滑动动作的边界，而且还检测与所述触摸或滑动操作相关的持续时间和压力。在一些实施例中，多媒体组件1108包括一个前置摄像头和/或后置摄像头。当设备1100处于操作模式，如拍摄模式或视频模式时，前置摄像头和/或后置摄像头可以接收外部的多媒体数据。每个前置摄像头和后置摄像头可以是一个固定的光学透镜系统或具有焦距和光学变焦能力。
音频组件1110被配置为输出和/或输入音频信号。例如，音频组件1110包括一个麦 克风(MIC)，当装置1100处于操作模式，如呼叫模式、记录模式和语音数据处理模式时，麦克风被配置为接收外部音频信号。所接收的音频信号可以被进一步存储在存储器1104或经由通信组件1116发送。在一些实施例中，音频组件1110还包括一个扬声器，用于输出音频信号。
I/O接口1112为处理组件1102和外围接口模块之间提供接口，上述外围接口模块可以是键盘，点击轮，按钮等。这些按钮可包括但不限于：主页按钮、音量按钮、启动按钮和锁定按钮。
传感器组件1114包括一个或多个传感器，用于为装置1100提供各个方面的状态评估。例如，传感器组件1114可以检测到设备1100的打开/关闭状态，组件的相对定位，例如所述组件为装置1100的显示器和小键盘，传感器组件1114还可以检测装置1100或装置1100一个组件的位置改变，用户与装置1100接触的存在或不存在，装置1100方位或加速/减速和装置1100的温度变化。传感器组件1114可以包括接近传感器，被配置用来在没有任何的物理接触时检测附近物体的存在。传感器组件1114还可以包括光传感器，如CMOS或CCD图像传感器，用于在成像应用中使用。在一些实施例中，该传感器组件1114还可以包括加速度传感器，陀螺仪传感器，磁传感器，压力传感器或温度传感器。
通信组件1116被配置为便于装置1100和其他设备之间有线或无线方式的通信。装置1100可以接入基于通信标准的无线网络，如WiFi，2G或3G，或它们的组合。在一个示例性实施例中，通信组件1116经由广播信道接收来自外部广播管理系统的广播信号或广播相关信息。在一个示例性实施例中，所述通信组件1116还包括近场通信(NFC)模块，以促进短程通信。例如，在NFC模块可基于射频数据处理(RFID)技术，红外数据协会(IrDA)技术，超宽带(UWB)技术，蓝牙(BT)技术和其他技术来实现。
在示例性实施例中，装置1100可以被一个或多个应用专用集成电路(ASIC)、数字信号处理器(DSP)、数字信号处理设备(DSPD)、可编程逻辑器件(PLD)、现场可编程门阵列(FPGA)、控制器、微控制器、微处理器或其他电子元件实现，用于执行上述方法。
在示例性实施例中，还提供了一种包括指令的非临时性计算机可读存储介质，例如包括指令的存储器1104，上述指令可由装置1100的处理器1120执行以完成上述方法。例如，所述非临时性计算机可读存储介质可以是ROM、随机存取存储器(RAM)、CD-ROM、磁带、软盘和光数据存储设备等。
图12是本申请的一些实施例中服务端的结构示意图。该服务端1900可因配置或性能不同而产生比较大的差异，可以包括一个或一个以上中央处理器(central processing units，CPU)1922(例如，一个或一个以上处理器)和存储器1932，一个或一个以上存储应用程序1942或数据1944的存储介质1930(例如一个或一个以上海量存储设备)。其中，存储器1932和存储介质1930可以是短暂存储或持久存储。存储在存储介质1930的 程序可以包括一个或一个以上模块(图示没标出)，每个模块可以包括对服务端中的一系列指令操作。更进一步地，中央处理器1922可以设置为与存储介质1930通信，在服务端1900上执行存储介质1930中的一系列指令操作。
服务端1900还可以包括一个或一个以上电源1926，一个或一个以上有线或无线网络接口1950，一个或一个以上输入输出接口1958，一个或一个以上键盘1956，和/或，一个或一个以上操作系统1941，例如Windows ServerTM，Mac OS XTM，UnixTM,LinuxTM，FreeBSDTM等等。
一种非临时性计算机可读存储介质，当所述存储介质中的指令由装置(服务端或者终端)的处理器执行时，使得装置能够执行图2至图9任一所示的问诊信息处理方法。
一种非临时性计算机可读存储介质，当所述存储介质中的指令由装置(服务端或者终端)的处理器执行时，使得装置能够执行一种问诊信息处理方法，所述方法包括：依据至少一次用户输入，确定用户疾病特征；对所述用户疾病特征进行疾病预测处理，以得到对应的候选疾病；依据所述候选疾病的疾病特征对应的疾病特征实体，从知识图谱中获取对应的问题实体；所述问题实体用于表征与所述疾病特征实体相关的问题；依据所述问题实体对应的问题，生成目标问题；所述目标问题用于对用户进行问诊。
本领域技术人员在考虑说明书及实践这里公开的发明后，将容易想到本申请的其它实施方案。本申请旨在涵盖本申请的任何变型、用途或者适应性变化，这些变型、用途或者适应性变化遵循本申请的一般性原理并包括本公开未公开的本技术领域中的公知常识或惯用技术手段。说明书和实施例仅被视为示例性的，本申请的真正范围和精神由下面的权利要求指出。
应当理解的是，本申请并不局限于上面已经描述并在附图中示出的精确结构，并且可以在不脱离其范围进行各种修改和改变。本申请的范围仅由所附的权利要求来限制。
以上所述仅为本申请的较佳实施例，并不用以限制本申请，凡在本申请的精神和原则之内，所作的任何修改、等同替换、改进等，均应包含在本申请的保护范围之内。
以上对本申请所提供的一种问诊信息处理方法、一种问诊信息处理装置和一种用于处理问诊信息的装置，进行了详细介绍，本文中应用了具体个例对本申请的原理及实施方式进行了阐述，以上实施例的说明只是用于帮助理解本申请的方法及其核心思想；同时，对于本领域的一般技术人员，依据本申请的思想，在具体实施方式及应用范围上均会有改变之处，综上所述，本说明书内容不应理解为对本申请的限制。

Claims (43)

  1. 一种问诊信息处理方法，其特征在于，所述方法包括：
    依据至少一次用户输入，确定用户疾病特征；
    对所述用户疾病特征进行疾病预测处理，以得到对应的候选疾病；
    依据所述候选疾病的疾病特征对应的疾病特征实体，从知识图谱中获取对应的问题实体；所述问题实体用于表征与所述疾病特征实体相关的问题；
    依据所述问题实体对应的问题，生成目标问题；所述目标问题用于对用户进行问诊。
  2. 根据权利要求1所述的方法，其特征在于，所述对所述用户疾病特征进行疾病预测处理，包括：
    依据所述用户疾病特征与疾病对应的疾病特征之间的匹配信息，确定所述用户疾病特征对应的候选疾病。
  3. 根据权利要求1所述的方法，其特征在于，所述问题实体包括：问题实体模板；所述依据所述问题实体对应的问题，生成目标问题，包括：
    依据所述候选疾病的疾病特征，对所述问题实体模板进行字段填充，以得到目标问题。
  4. 根据权利要求3所述的方法，其特征在于，所述对所述问题实体模板进行字段填充，包括：
    依据属于一种类型的至少一种疾病特征，对所述类型对应的问题实体模板进行字段填充，以在得到的目标问题中携带至少一种疾病特征的信息。
  5. 根据权利要求3所述的方法，其特征在于，所述问题实体模板的字段包括：问题文本字段和答案选项字段；
    所述对所述问题实体模板进行字段填充，包括：
    在所述问题实体模板的答案选项字段中填充所述候选疾病的疾病特征的信息。
  6. 根据权利要求3所述的方法，其特征在于，所述对所述问题实体模板进行字段填充，包括：
    依据所述候选疾病的疾病特征对应疾病特征实体中的命中动作属性，对所述问题实体模板的跳转关系字段进行填充。
  7. 根据权利要求1所述的方法，其特征在于，所述问题实体包括：问题实体实例；所述方法还包括：
    从所述候选疾病的疾病特征对应的问题实体实例中，获取目标问题。
  8. 根据权利要求1所述的方法，其特征在于，所述从知识图谱中获取对应的问题实体，包括：
    在所述候选疾病的疾病特征对应的答案选项被选中的情况下，依据所述候选疾病的疾病特征对应的疾病特征实体中的命中动作属性，从知识图谱中获取对应的问题实体；和/或，
    在所述候选疾病的疾病特征对应的答案选项被选中的情况下，依据所述疾病特征对应问题实体中的跳转关系字段，从知识图谱中获取对应的问题实体。
  9. 根据权利要求1所述的方法，其特征在于，所述从知识图谱中获取对应的问题实体，包括：
    依据疾病特征对应的重要性得分，从所述候选疾病对应的疾病特征中确定出目标疾病特征；所述目标疾病特征用于表征本轮问诊向用户询问的疾病特征；
    依据所述目标疾病特征对应的类型，从知识图谱中获取对应类型的问题实体；所述问题实体用于表征与所述疾病特征实体相关的问题；
    则所述生成目标问题，包括：
    依据所述问题实体、以及属于所述类型的目标疾病特征，生成目标问题。
  10. 根据权利要求1至9中任一所述的方法，其特征在于，所述至少一次用户输入包括：
    主动输入；或者
    主动输入和针对预设问题的回复；或者
    主动输入和针对所述目标问题的回复；或者
    主动输入、以及针对预设问题和目标问题的回复。
  11. 根据权利要求1至9中任一所述的方法，其特征在于，所述目标问题包括：问题文本和答案选项；
    所述至少一次用户输入，包括：用户选择的答案选项。
  12. 根据权利要求1至9中任一所述的方法，其特征在于，所述方法还包括：
    依据所述用户疾病特征，从知识图谱中获取预设疾病特征实体对应的问题实体；
    依据所述问题实体，确定预设疾病问题；所述预设疾病问题包括：问题文本、以及至少一个预设选项；
    若接收到用户针对任一预设选项的选择操作，则输出对应的就医建议信息。
  13. 根据权利要求1至9中任一所述的方法，其特征在于，所述方法还包括：
    依据概率特征，确定所述候选疾病的得分；
    所述概率特征包括如下特征中的至少一种：
    与所述用户疾病特征相匹配的疾病特征在候选疾病的条件下的条件概率；
    与所述用户疾病特征相匹配的疾病特征在候选疾病的条件下的惩罚因子；
    候选疾病在疾病系统中的发病概率；
    以及疾病系统的发病概率。
  14. 根据权利要求1所述的方法，其特征在于，所述方法还包括：
    在得到候选疾病后，若符合停止条件，则停止执行所述生成目标问题的步骤。
  15. 根据权利要求1所述的方法，其特征在于，所述方法还包括：
    在得到候选疾病后，若符合停止条件，则输出问诊信息处理结果；所述问诊信息处理结果包括：所述候选疾病的信息。
  16. 根据权利要求15所述的方法，其特征在于，所述问诊信息处理结果还包括：用户出现的疾病特征、以及用户未出现的疾病特征。
  17. 根据权利要求14或15所述的方法，其特征在于，所述停止条件包括如下条件中的至少一种：
    至少一种候选疾病的得分大于得分阈值；
    多种候选疾病的得分差异符合差异条件；
    所述候选疾病对应疾病特征的询问比例符合比例条件；以及
    询问轮数超过轮数阈值。
  18. 一种问诊信息处理装置，其特征在于，包括：
    用户疾病特征确定模块，配置为依据至少一次用户输入，确定用户疾病特征；
    用户疾病特征处理模块，配置为对所述用户疾病特征进行疾病预测处理，以得到对应的候选疾病；
    问题实体获取模块，配置为依据所述候选疾病的疾病特征对应的疾病特征实体，从知识图谱中获取对应的问题实体；所述问题实体用于表征与所述疾病特征实体相关的问题；
    第一问题生成模块，配置为依据所述问题实体对应的问题，生成目标问题；所述目标问题用于对用户进行问诊。
  19. 根据权利要求18所述的装置，其特征在于，所述用户疾病特征处理模块包括：
    候选疾病确定模块，配置为依据所述用户疾病特征与疾病对应的疾病特征之间的匹配信息，确定所述用户疾病特征对应的候选疾病。
  20. 根据权利要求18所述的装置，其特征在于，所述问题实体包括：问题实体模板；所述第一问题生成模块包括：
    字段填充模块，配置为依据所述候选疾病的疾病特征，对所述问题实体模板进行字段填充，以得到目标问题。
  21. 根据权利要求20所述的装置，其特征在于，所述字段填充模块包括：
    第一字段填充模块，配置为依据属于一种类型的至少一种疾病特征，对所述类型对应的问题实体模板进行字段填充，以在得到的目标问题中携带至少一种疾病特征的信息。
  22. 根据权利要求20所述的装置，其特征在于，所述问题实体模板的字段包括：问题文本字段和答案选项字段；
    所述字段填充模块包括：
    第二字段填充模块，配置为在所述问题实体模板的答案选项字段中填充所述候选疾病的疾病特征的信息。
  23. 根据权利要求20所述的装置，其特征在于，所述字段填充模块包括：
    第三字段填充模块，配置为依据所述候选疾病的疾病特征对应疾病特征实体中的命中动作属性，对所述问题实体模板的跳转关系字段进行填充。
  24. 根据权利要求18所述的装置，其特征在于，所述问题实体包括：问题实体实例；所述装置还包括：
    第二问题生成模块，配置为从所述候选疾病的疾病特征对应的问题实体实例中，获取目标问题。
  25. 根据权利要求18所述的装置，其特征在于，所述问题实体获取模块包括：
    第一问题实体获取模块，配置为在所述候选疾病的疾病特征对应的答案选项被选中的情况下，依据所述候选疾病的疾病特征对应的疾病特征实体中的命中动作属性，从知识图谱中获取对应的问题实体；和/或，
    第二问题实体获取模块，配置为在所述候选疾病的疾病特征对应的答案选项被选中的情况下，依据所述疾病特征对应问题实体中的跳转关系字段，从知识图谱中获取对应的问题实体。
  26. 根据权利要求18所述的装置，其特征在于，所述问题实体获取模块包括：
    目标疾病特征确定模块，配置为依据疾病特征对应的重要性得分，从所述候选疾病对应的疾病特征中确定出目标疾病特征；所述目标疾病特征用于表征本轮问诊向用户询问的疾病特征；
    实体确定模块，配置为依据所述目标疾病特征对应的类型，从知识图谱中获取对应类型的问题实体；
    则所述第一问题生成模块，具体配置为依据所述问题实体、以及属于所述类型的目标疾病特征，生成目标问题。
  27. 根据权利要求18至26中任一所述的装置，其特征在于，所述至少一次用户输入包括：
    主动输入；或者
    主动输入和针对预设问题的回复；或者
    主动输入和针对所述目标问题的回复；或者
    主动输入、以及针对预设问题和目标问题的回复。
  28. 根据权利要求18至26中任一所述的装置，其特征在于，所述目标问题包括：问题文本和答案选项；
    所述至少一次用户输入，包括：用户选择的答案选项。
  29. 根据权利要求18至26中任一所述的装置，其特征在于，所述装置还包括：
    预设问题实体获取模块，配置为依据所述用户疾病特征，从知识图谱中获取预设疾病特征实体对应的问题实体；
    预设疾病问题确定模块，配置为依据所述问题实体，确定预设疾病问题；所述预设疾病问题包括：问题文本、以及至少一个预设选项；
    建议输出模块，配置为若接收到用户针对任一预设选项的选择操作，则输出对应的就医建议信息。
  30. 根据权利要求18至26中任一所述的装置，其特征在于，所述装置还包括：
    得分确定模块，配置为依据概率特征，确定所述候选疾病的得分；
    所述概率特征包括如下特征中的至少一种：
    与所述用户疾病特征相匹配的疾病特征在候选疾病的条件下的条件概率；
    与所述用户疾病特征相匹配的疾病特征在候选疾病的条件下的惩罚因子；
    候选疾病在疾病系统中的发病概率；
    以及疾病系统的发病概率。
  31. 根据权利要求18所述的装置，其特征在于，所述装置还包括：
    停止模块，配置为在得到候选疾病后，若符合停止条件，则通知所述问题生成模块停止执行所述生成目标问题。
  32. 根据权利要求18所述的装置，其特征在于，所述装置还包括：
    处理结果输出模块，配置为在得到候选疾病后，若符合停止条件，则输出问诊信息处理结果；所述问诊信息处理结果包括：所述候选疾病的信息。
  33. 根据权利要求32所述的装置，其特征在于，所述问诊信息处理结果还包括：用户出现的疾病特征、以及用户未出现的疾病特征。
  34. 根据权利要求31或32所述的装置，其特征在于，所述停止条件包括如下条件中的至少一种：
    至少一种候选疾病的得分大于得分阈值；
    多种候选疾病的得分差异符合差异条件；
    所述候选疾病对应疾病特征的询问比例符合比例条件；以及
    询问轮数超过轮数阈值。
  35. 一种用于处理问诊信息的装置，其特征在于，包括有存储器，以及一个或者一个以上的程序，其中一个或者一个以上程序存储于存储器中，且所述程序被一个或者一个以上处理器执行时，实现权利要求1至17中任一所述方法的步骤。
  36. 一种机器可读介质，其上存储有指令，当由一个或多个处理器执行时，使得装置执行如权利要求1至17中一个或多个所述的问诊信息处理方法。
  37. 一种知识图谱的构建方法，其特征在于，所述方法包括：
    依据疾病特征实体，确定问题实体；所述问题实体用于表征与所述疾病特征实体相关的问题；
    在知识图谱中，建立所述疾病特征实体与所述问题实体之间的关联。
  38. 根据权利要求37所述的方法，其特征在于，所述问题实体的字段包括：问题文本字段和答案选项字段。
  39. 根据权利要求38所述的方法，其特征在于，所述问题实体包括：
    问题实体实例；所述问题实体实例的所有字段处于已填充状态；和/或
    问题实体模板；所述问题实体模板的问题文本字段处于已填充状态，所述问题实体模板的除了问题文本字段之外的预设字段处于未填充状态。
  40. 根据权利要求37所述的方法，其特征在于，所述方法还包括：
    在知识图谱中，建立所述疾病特征实体与疾病实体之间的关联；
    所述疾病实体的属性包括如下属性中的至少一种：
    疾病标识属性、疾病系统属性、特征集合属性、临床占比属性、以及高发年龄属性；
    其中，所述特征集合中包括：与所述疾病特征实体相关联的疾病特征；
    所述临床占比属性用于表征疾病在疾病系统中的发病概率。
  41. 根据权利要求40所述的方法，其特征在于，所述疾病系统属性对应有系统概率，用于表征疾病系统的疾病的发病概率之和。
  42. 根据权利要求40所述的方法，其特征在于，所述特征集合属性的属性参数包括如下参数中的至少一种：
    疾病特征在疾病的条件下的条件概率；
    疾病特征在疾病的条件下的惩罚因子。
  43. 一种问诊信息处理方法，其特征在于，所述方法包括：
    依据基于至少一次用户输入得到的用户疾病特征，从知识图谱中获取预设疾病特征实体对应的问题实体；所述问题实体用于表征与所述预设疾病特征实体相关的问题；
    依据所述问题实体，确定预设疾病问题；所述预设疾病问题具体包括：问题文本、以及至少一个预设选项；
    若接收到用户针对任一预设选项的选择操作，则输出对应的就医建议信息。
PCT/CN2021/103667 2021-01-26 2021-06-30 一种问诊信息处理方法、装置和介质 WO2022160596A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US18/137,960 US20230268073A1 (en) 2021-01-26 2023-04-21 Inquiry information processing method and apparatus, and medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110105880.9 2021-01-26
CN202110105880.9A CN112768091A (zh) 2021-01-26 2021-01-26 一种问诊信息处理方法、装置和介质

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US18/137,960 Continuation US20230268073A1 (en) 2021-01-26 2023-04-21 Inquiry information processing method and apparatus, and medium

Publications (1)

Publication Number Publication Date
WO2022160596A1 true WO2022160596A1 (zh) 2022-08-04

Family

ID=75705838

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/103667 WO2022160596A1 (zh) 2021-01-26 2021-06-30 一种问诊信息处理方法、装置和介质

Country Status (3)

Country Link
US (1) US20230268073A1 (zh)
CN (1) CN112768091A (zh)
WO (1) WO2022160596A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116612906A (zh) * 2023-07-20 2023-08-18 北方健康医疗大数据科技有限公司 一种基于人工智能的医疗问答服务方法、系统及设备

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112768091A (zh) * 2021-01-26 2021-05-07 北京搜狗科技发展有限公司 一种问诊信息处理方法、装置和介质
CN112948599A (zh) * 2021-05-13 2021-06-11 北京富通东方科技有限公司 一种基于医学知识图谱的路径推理的疾病诊断方法
CN113539520B (zh) * 2021-07-23 2024-10-15 平安科技（深圳）有限公司 实现问诊会话的方法、装置、计算机设备及存储介质
CN114300160B (zh) * 2021-11-16 2022-10-18 北京左医科技有限公司 问诊对话方法及系统

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109448838A (zh) * 2018-09-28 2019-03-08 小伍健康科技（上海）有限责任公司 一种基于深度神经网络的症状诊断方法及设备
US20200356586A1 (en) * 2017-11-27 2020-11-12 Zte Corporation Intelligent question and answer method and device
CN111984771A (zh) * 2020-07-17 2020-11-24 北京欧应信息技术有限公司 一种基于智能对话的自动问诊系统
CN112035636A (zh) * 2020-08-28 2020-12-04 康键信息技术（深圳）有限公司 医疗问诊系统的问答管理方法、装置、设备及存储介质
CN112768091A (zh) * 2021-01-26 2021-05-07 北京搜狗科技发展有限公司 一种问诊信息处理方法、装置和介质

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111274365B (zh) * 2020-02-25 2023-09-19 广州七乐康药业连锁有限公司 基于语义理解的智能问诊方法、装置、存储介质及服务器

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20200356586A1 (en) * 2017-11-27 2020-11-12 Zte Corporation Intelligent question and answer method and device
CN109448838A (zh) * 2018-09-28 2019-03-08 小伍健康科技（上海）有限责任公司 一种基于深度神经网络的症状诊断方法及设备
CN111984771A (zh) * 2020-07-17 2020-11-24 北京欧应信息技术有限公司 一种基于智能对话的自动问诊系统
CN112035636A (zh) * 2020-08-28 2020-12-04 康键信息技术（深圳）有限公司 医疗问诊系统的问答管理方法、装置、设备及存储介质
CN112768091A (zh) * 2021-01-26 2021-05-07 北京搜狗科技发展有限公司 一种问诊信息处理方法、装置和介质

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116612906A (zh) * 2023-07-20 2023-08-18 北方健康医疗大数据科技有限公司 一种基于人工智能的医疗问答服务方法、系统及设备
CN116612906B (zh) * 2023-07-20 2023-11-10 北方健康医疗大数据科技有限公司 一种基于人工智能的医疗问答服务方法、系统及设备

Also Published As

Publication number Publication date
US20230268073A1 (en) 2023-08-24
CN112768091A (zh) 2021-05-07

Similar Documents

Publication Publication Date Title
WO2022160596A1 (zh) 一种问诊信息处理方法、装置和介质
WO2021018154A1 (zh) 信息表示方法及装置
US12100398B2 (en) Generating IoT-based notification(s) and provisioning of command(s) to cause automatic rendering of the IoT-based notification(s) by automated assistant client(s) of client device(s)
US9280610B2 (en) Crowd sourcing information to fulfill user requests
US9363551B2 (en) TV program identification method, apparatus, terminal, server and system
WO2017084541A1 (zh) 会话中实现表情图像发送的方法和装置
US12008318B2 (en) Automatic personalized story generation for visual media
CN106874308B (zh) 一种推荐方法和装置、一种用于推荐的装置
US20190197315A1 (en) Automatic story generation for live media
US11729470B2 (en) Predictive media routing based on interrupt criteria
CN105701254A (zh) 一种信息处理方法和装置、一种用于信息处理的装置
WO2015196987A1 (zh) 支持自然语言的数据查询方法、开放平台及用户终端
CN107784045B (zh) 一种快捷回复方法和装置、一种用于快捷回复的装置
CN111708943A (zh) 一种搜索结果展示方法、装置和用于搜索结果展示的装置
US20140280299A1 (en) Simplified Collaborative Searching Through Pattern Recognition
KR102065029B1 (ko) 전자 장치의 데이터 공유 방법 및 그 전자 장치
US20210406736A1 (en) System and method of content recommendation
CN109213942A (zh) 一种搜索结果展示方法及装置
JP2020004410A (ja) メディアベースのコンテンツシェアを容易にする方法、コンピュータプログラム及びコンピューティングデバイス
WO2022116527A1 (zh) 一种数据处理方法及装置
CN110929122B (zh) 一种数据处理方法、装置和用于数据处理的装置
CN113239183B (zh) 排序模型的训练方法、装置、电子设备和存储介质
CN112749287A (zh) 知识图谱的构建方法和使用方法、装置和介质
KR20160059980A (ko) 제공된 콘텐트를 위한 참조 정보 관리 장치 및 방법
CN112749286A (zh) 知识图谱的构建和使用方法、装置和介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21922193

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21922193

Country of ref document: EP

Kind code of ref document: A1