CN109637651A

CN109637651A - More doctor's recommended methods and device, online consultation system

Info

Publication number: CN109637651A
Application number: CN201811285097.XA
Authority: CN
Inventors: 曾柏毅; 王建国; 周岩; 杨三喜; 黄鹏; 李蕊; 雷宇
Original assignee: Beijing Chun Yu World Software Co Ltd
Current assignee: Beijing Chun Yu World Software Co Ltd
Priority date: 2018-10-31
Filing date: 2018-10-31
Publication date: 2019-04-16

Abstract

The application provides a kind of more doctor's recommended methods and device, online consultation system, method includes: to obtain target user's question information and candidate doctor's list corresponding with target user's question information, it wherein, include multiple information about doctor in candidate doctor's list；Target user's question information is combined with each information about doctor respectively, forms multiple forecast samples；And each forecast sample is inputted into preset prediction model respectively, and multiple target information about doctor for recommending to target user are chosen according to each output of the prediction model.The multidigit doctor that the application efficiently and accurately can recommend matching degree high and strongly professional to user allows users to obtain the information such as more accurate and high matching degree disease symptoms explanation and treatment recommendations.

Description

More doctor's recommended methods and device, online consultation system

Technical field

This application involves technical field of data processing, and in particular to a kind of more doctor's recommended methods and device, the online consultation of doctors System.

Background technique

With the development of science and technology, people carry out disease symptoms and the consultation way for the treatment of is no longer only limitted to enter hospital's progress It sees a doctor on line, the suggestion for its disease symptoms and therapeutic modality can also be obtained by internet.Especially for remoteness Patient, and a certain hospital carried out diagnosis it is desired that obtain for patient second suggest crowd, The professional advice that doctor is obtained by internet is clearly a kind of more convenient and efficient consultation way.

In the prior art, user on carrying out line when seeking advice from, it usually needs the answer of doctor is waited after issuing problem, so And user can not but know the doctor to answer a question, if domain expert corresponding to the disease or symptom proposed for it, i.e., Domain expert corresponding to the disease or symptom for proposing user to it in advance has gained some understanding, but can obtain the neck without approach The answer of domain expert, therefore, user in the prior art are in passive state when seeking advice from carrying out line always, and can not obtain It obtains matching degree and the second of strongly professional doctor suggests.

Therefore, a kind of method for how designing doctor for recommending matching degree high and strongly professional to user, is urgently It solves the problems, such as.

Summary of the invention

For the problems of the prior art, the application provides a kind of more doctor's recommended methods and device, online consultation system, The multidigit doctor that efficiently and accurately matching degree can be recommended high and strongly professional to user allows users to obtain subject to more The information such as true and high matching degree disease symptoms explanation and treatment recommendations.

In order to solve the above technical problems, the application the following technical schemes are provided:

In a first aspect, the application provides a kind of more doctor's recommended methods, comprising:

Obtain target user's question information and candidate doctor's list corresponding with target user's question information, wherein It include multiple information about doctor in candidate doctor's list；

Target user's question information is combined with each information about doctor respectively, forms multiple pre- test samples This；

And each forecast sample is inputted into preset prediction model respectively, and according to each of the prediction model Multiple target information about doctor for recommending to target user are chosen in a output.

Further, further includes:

Log is recommended to generate multiple training samples according to history doctor, wherein the history doctor recommends includes in log The multiple target information about doctor and each mesh for having multiple historical user's question informations, recommending respectively to each historical user Mark the adopted rate of information about doctor；

The prediction model is trained using the training sample set.

Further, further includes:

Generate the target doctor's recommendation list being made of multiple target information about doctor；

The target doctor recommendation list is sent to target user, so that the target user recommends in the target doctor At least one described target user's question information of the corresponding target doctor answer of target information about doctor is selected in list, with reality The present line consultation of doctors.

Further, the acquisition target user question information and candidate doctor corresponding with target user's question information Raw list, comprising:

Receive target user's question information that the target user sends；

And the candidate doctor's list being made of multiple information about doctor is determined according to target user's question information.

It is further, described that corresponding candidate doctor's list is determined according to target user's question information, comprising:

According to target user's question information, the corresponding target disease department of target user's question information is determined And/or extension department；

In the corresponding each information about doctor of the target disease department and/or extension department, it is default that selection meets first Multiple information about doctor of rule, form doctor's list to be selected；

Doctor's list to be selected is screened using the second preset rules, obtains candidate doctor's list.

It is further, described to choose the first multiple information about doctor for meeting preset rules, comprising:

If present period is in the period in the daytime, it is chosen at each doctor couple that on-line operation is carried out in the first preset period of time The information about doctor answered；

If present period is in night-time hours, it is chosen at each doctor couple that on-line operation is carried out in the second preset period of time The information about doctor answered；

Wherein, first preset period of time is less than second preset period of time, the period in the daytime and the night-time hours The sum of be 24 hours.

Further, the second preset rules of the application screen doctor's list to be selected, comprising:

From in doctor's list to be selected, the corresponding doctor's letter of doctor for having received target user's question information is deleted Breath, and/or；

From in doctor's list to be selected, the doctor couple for having answered the relevant issues of target user's question information is deleted The information about doctor answered.

From in doctor's list to be selected, deletion has received and has not answered historical user's question information quantity in advance greater than first If the corresponding information about doctor of doctor of value, and/or；

From in doctor's list to be selected, the information about doctor in default blacklist is deleted.

According to target user's question information, determining target user's question information, corresponding to prestore hospital corresponding Target disease department and/or extension department；

In the corresponding each information about doctor of the target disease department and/or extension department, choose default profession degree with Positive rating meets the corresponding information about doctor of each test doctor of corresponding preset requirement, forms candidate doctor's column Table.

It further, also include the corresponding user information of target user and problem letter in target user's question information Breath；

It is described to be combined target user's question information with each information about doctor respectively, form multiple predictions Sample, comprising:

Feature extraction is carried out respectively to the user information and problem information, the user for respectively obtaining the user information is special Vector is levied, and, the corresponding problem characteristic vector of described problem information；

And obtain the corresponding doctor's feature vector of each information about doctor in candidate doctor's list；

By the corresponding problem characteristic vector of the target user and user characteristics vector respectively with each doctor's feature Vector is combined, and obtains multiple forecast samples.

Further, the user information includes: user's gender and/or age of user information.

Further, the user information further include: user and patient relationship information, user region information, user Free enquirement number, user charges enquirement number, user contact details, user seeks advice from number, user extends appreciation doctor visits and each Any one of a History Order payment amount or any combination.

Further, described problem information includes: disease description information and/or symptom description information.

Further, described problem information further include: the affiliated department's information of disease, problem source identification, question text are long Spend any one of information, problem creation time information and problem creation area information or any combination.

Further, the information about doctor includes: the target disease department belonging to doctor and/or extension department letter Breath, and, doctor is good at disease information.

Further, the information about doctor further include: doctor gender information, doctor's age information, doctor academic title's information, doctor Price is answered a question in city where hospital mark, doctor where raw, physician ratings, doctor and doctor services appointing in person-time information One or any combination.

Further, described that each forecast sample is inputted into preset prediction model respectively, and according to the prediction Multiple target information about doctor for recommending to target user are chosen in each output of model, comprising:

Each forecast sample is inputted into preset prediction model respectively, using each output of the prediction model as The adopted rate of each information about doctor in candidate's doctor's list；

It chooses and is greater than information about doctor corresponding to the adopted rate of the second preset value, as being pushed away to target user The target information about doctor recommended.

Second aspect, the application provide a kind of more doctor's recommendation apparatus, comprising:

Data obtaining module, for obtaining target user's question information and time corresponding with target user's question information Select doctor's list, wherein include multiple information about doctor in candidate doctor's list；

Forecast sample determining module, for carrying out target user's question information with each information about doctor respectively Combination, forms multiple forecast samples；

Multiple target doctor's recommending module, for each forecast sample to be inputted preset prediction model, and root respectively Multiple target information about doctor for recommending to target user are chosen according to each output of the prediction model.

The third aspect, the application provide a kind of online consultation system, comprising: user terminal and more doctors recommend dress It sets；

The user terminal is used to send target user's question informations to more doctor's recommendation apparatus, and, it receives simultaneously Multiple target information about doctor that more doctor's recommendation apparatus are sent are shown, so that target user selects described at least one Target information about doctor corresponding target doctor answer target user's question information, to realize the online consultation of doctors.

Fourth aspect, the application provides a kind of electronic equipment, including memory, processor and storage are on a memory and can The computer program run on a processor, the processor realize the step of more doctor's recommended methods when executing described program Suddenly.

5th aspect, the application provide a kind of computer readable storage medium, are stored thereon with computer program, the calculating The step of more doctor's recommended methods are realized when machine program is executed by processor.

As shown from the above technical solution, the application provides a kind of more doctor's recommended methods and device, online consultation system, institute More doctor's recommended methods are stated by obtaining target user's question information and candidate doctor corresponding with target user's question information Raw list, wherein include multiple information about doctor in candidate doctor's list, the data basis of forecast sample can be effectively improved Specific aim, and then the accuracy of model prediction can be effectively improved；By by target user's question information respectively and respectively A information about doctor is combined, and forms multiple forecast samples, can effectively improve the comprehensive of forecast sample, and can have The selection diversity for imitating the model prediction obtained, by the way that each forecast sample is inputted preset prediction model respectively, and Choose multiple target information about doctor for recommending to target user according to each output of the prediction model, can efficiently and The multidigit doctor for accurately recommending matching degree high and strongly professional to user, allows users to obtain more accurate and matching degree The information such as high disease symptoms explanation and treatment recommendations, and then user experience can be effectively improved.

Detailed description of the invention

In order to illustrate the technical solutions in the embodiments of the present application or in the prior art more clearly, to embodiment or will show below There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is the application Some embodiments for those of ordinary skill in the art without creative efforts, can also basis These attached drawings obtain other attached drawings.

Fig. 1 is a kind of configuration diagram of the online consultation system in the embodiment of the present application.

Fig. 2 be the embodiment of the present application in include doctor terminal online consultation system configuration diagram.

Fig. 3 be the embodiment of the present application in include back-end server and big data server online consultation system frame Structure schematic diagram.

Fig. 4 is the process structure schematic diagram of more doctor's recommended methods in the embodiment of the present application.

Fig. 5 be the embodiment of the present application in include step 001 and 002 more doctor's recommended methods process structure signal Figure.

Fig. 6 be the embodiment of the present application in include step 400 and 500 more doctor's recommended methods process structure signal Figure.

Fig. 7 be the embodiment of the present application in more doctor's recommended methods in step 100 process structure schematic diagram.

Fig. 8 is the first process with embodiment of step 102 in more doctor's recommended methods in the embodiment of the present application Structure schematic diagram.

Fig. 9 is second of step 102 process with embodiment in more doctor's recommended methods in the embodiment of the present application Structure schematic diagram.

Figure 10 be the embodiment of the present application in more doctor's recommended methods in step 200 process structure schematic diagram.

Figure 11 be the embodiment of the present application in more doctor's recommended methods in step 300 process structure schematic diagram.

Figure 12 is that the data of more doctor's recommended methods in the application application example transmit example schematic.

Figure 13 is the structural schematic diagram of more doctor's recommendation apparatus in the embodiment of the present application.

Figure 14 is the structural schematic diagram of the electronic equipment in the embodiment of the present application.

Specific embodiment

To keep the purposes, technical schemes and advantages of the embodiment of the present application clearer, below in conjunction with the embodiment of the present application In attached drawing, technical solutions in the embodiments of the present application carries out clear, complete description, it is clear that described embodiment is Some embodiments of the present application, instead of all the embodiments.Based on the embodiment in the application, those of ordinary skill in the art Every other embodiment obtained without creative efforts, shall fall in the protection scope of this application.

In view of existing user carry out line on seek advice from when be in passive state always, and can not obtain matching degree and specially The situation that the second of the strong doctor of industry is suggested, the application provide a kind of more doctor's recommended methods, more doctor's recommendation apparatus, electronics Consultation system on equipment, computer readable storage medium and a kind of line, by obtain target user's question information and with the mesh Mark the corresponding candidate doctor's list of user's question information, wherein it include multiple information about doctor in candidate doctor's list, it can The specific aim of the data basis of forecast sample is effectively improved, and then the accuracy of model prediction can be effectively improved；By by institute It states target user's question information to be combined with each information about doctor respectively, forms multiple forecast samples, can effectively mention Comprehensive, and the selection diversity for the model prediction that can effectively obtain of high forecast sample, by by each pre- test sample This inputs preset prediction model respectively, and is chosen according to each output of the prediction model for recommending to target user Multiple target information about doctor, the multidigit doctor that efficiently and accurately matching degree can be recommended high and strongly professional to user, so that User can obtain the information such as more accurate and high matching degree disease symptoms explanation and treatment recommendations, and then can effectively improve User experience.

In one or more embodiments of the application, referring to Fig. 1, the online consultation system includes more doctors Recommendation apparatus A1 and user terminal B1；Target user's question information that target user edits is sent to institute by the user terminal B1 More doctor's recommendation apparatus A1 are stated, more doctor's recommendation apparatus A1 obtain target user's question information and mention with the target user Ask information corresponding candidate doctor's list, wherein to include multiple information about doctor in candidate doctor's list, and by the target User's question information is combined with each information about doctor respectively, forms multiple forecast samples, and will be each described pre- Test sample originally inputs preset prediction model respectively, and is chosen according to each output of the prediction model for pushing away to target user Multiple target information about doctor are then sent to the user terminal B1, the user terminal by the multiple target information about doctor recommended B1 have display screen B2, multiple target information about doctor are shown, user by display screen B2 in multiple target doctors Information chooses corresponding one or more doctors, carries out the consulting such as disease symptoms or therapeutic scheme to it.

It can also include doctor terminal C1 in the online consultation system referring to fig. 2, target is used based on above content Family by display screen B2 after multiple target information about doctor choose corresponding one or more doctors, the user terminal B1 Target user's question information is sent to user and chooses corresponding one or more doctors institute in multiple target information about doctor The doctor terminal C1 held, the doctor are directed to target user's question information to target by its corresponding doctor terminal C1 User provides the disease symptoms explanation and treatment recommendations of profession.

It is understood that more doctor's recommendation apparatus A1 can be a server, it can also be by back-end server A2 And big data server A 3 collectively constitutes, referring to Fig. 3, the target user that the user terminal B1 edits target user puts question to letter Breath is sent to the back-end server A2, the back-end server A2 obtain target user's question information and with the target user The corresponding candidate doctor's list of question information, wherein it include multiple information about doctor in candidate doctor's list, it is then, described Target user's question information and candidate doctor's list are sent to the big data server A 3 by back-end server A2, described Target user's question information is combined with each information about doctor by big data server A 3 respectively, is formed multiple pre- Test sample sheet, and each forecast sample is inputted into preset prediction model respectively, and according to each of the prediction model Multiple target information about doctor for recommending to target user are chosen in output, are then sent to multiple target information about doctor described Back-end server A2, and corresponding user terminal B1 is forwarded to by the back-end server A2.

It is understood that the user terminal and doctor terminal all can be a kind of terminal device, and the terminal device It can be communicated to connect with more doctor's recommendation apparatus A1.Specifically, the terminal device may include smart phone, plate electricity Sub- equipment, network machine top box, portable computer, desktop computer, personal digital assistant (PDA), mobile unit, intelligence wearing are set It is standby etc..Wherein, the intelligent wearable device may include smart glasses, smart watches, Intelligent bracelet etc..

In practical applications, the part for carrying out more doctor's recommendations can be in more doctor's recommendation apparatus as described in above content (i.e. server device) side A1 executes, that is, framework as shown in Figure 1, operation that can also be all are complete all in client device At.It can specifically be selected according to the processing capacity of equipment and limitation of user's usage scenario etc..The application does not make this It limits.If all operations are all completed in client device, client device can also include processor, for carrying out cure more The raw specific processing recommended.

Above-mentioned client device can have communication module (i.e. communication unit), can be led to long-range server Letter connection, realizes and transmits with the data of the server.For example, communication unit can be by the above-mentioned target for editing target user User's question information is sent to server device, pushes away so that server device carries out more doctors according to target user's question information It recommends.Communication unit can also receive the multiple target information about doctor of server device return.The server may include The server of task schedule center side also may include the server of halfpace in other implement scenes, for example, with appoint Business control centre's server has the server of the third-party server platform of communication linkage.The server may include separate unit Computer equipment also may include the server cluster of multiple server compositions or the server architecture of distributed devices.

Any suitable network protocol can be used between the server and the terminal device to be communicated, be included in The network protocol that the application submitting day is not yet developed.The network protocol for example may include ICP/IP protocol, UDP/IP association View, http protocol, HTTPS agreement etc..Certainly, the network protocol for example can also include using on above-mentioned agreement RPC agreement (Remote Procedure Call Protocol, remote procedure call protocol), REST agreement (Representational State Transfer, declarative state transfer protocol) etc..

In one or more embodiments of the application, history doctor can be stored in advance and recommend log, history doctor Recommend log to be used to store historical user's question information, multiple target information about doctor for recommending respectively to each historical user and Corresponding relationship between the adopted rate of each target information about doctor.

In addition, the history doctor recommends log to be specifically as follows a distributed data base, illustrate referring to table 1.

Table 1

Based on table 1, the distributed data base be used for store user's question information, recommendation multiple target information about doctor and Corresponding relationship between the adopted rate of each target information about doctor.Wherein, each recommended target information about doctor with It is one-to-one relationship between its adopted rate, and user's question information is then between recommended target information about doctor One-to-many relationship.It is illustrated with user's question information Q2, is recommended to the target user for issuing user's question information Q2 Target information about doctor respectively there are five, be followed successively by Y21, Y22, Y23, Y24 and Y25, and the corresponding quilt of five target information about doctor Adopting rate is respectively 15%, 29%, 37%, 10% and 9%.

It is understood that user's question information is that the subject of question issued for target user carries out target critical User's question information after word extraction.For example, if the target user has issued the original inquiry content " cough of its editor Three days or more and with fever phenomenon, tongue fur jaundice, does is I virus flu? ", then according to preset keywords database pair The content is matched, identification obtain keyword " cough ", generate heat ", " tongue fur jaundice " and " virus flu ", therefore, the mesh The practical corresponding user's question information of mark user is " cough, fever, tongue fur jaundice and virus flu ", and non-user editor Original inquiry information.

Based on this, recommend in log in history doctor above-mentioned, user's question information is actually not directed to one Target user, but the user for proposing similar problems and being extracted same keyword and combinations thereof, therefore, the target doctor The adopted rate of raw information illustrates to illustrate are as follows: if proposing similar problems and being extracted the target of same subscriber question information Q1 User has 100, and wherein has 75 people that target information about doctor Y2 has finally been selected to seek advice from Y1, Y2 and Y3, then described The adopted rate of target information about doctor Y2 is 75%.

In one or more embodiments of the application, target user's question information is according to current target user The original inquiry content of its editor issued carries out user's question information after feature extraction.Wherein, the user issue its The original inquiry content of editor can also include picture in addition to above-mentioned word content.

That is, being based on preset picture feature if including picture in the original inquiry content received Database carries out feature identification to the picture, if unidentified arrive corresponding feature, which is determined as invalid picture, if Corresponding picture feature is recognized, then in the use for obtain after feature extraction according to the word content in the original inquiry content In the question information of family, judge whether to include the picture feature, if not having, the corresponding character features of the picture feature are added In user's question information, target user's question information corresponding to the target user is obtained.

In a kind of citing, if the original inquiry content of target user input includes word content: " cough three days Above and with fever phenomenon, I is virus flu? " and image content: the picture of tongue fur jaundice.It is then right The original inquiry content carries out feature extraction, obtains initial user's question information " cough, fever and virus flu ", and Feature extraction is carried out to the picture, obtains the character features of corresponding " tongue fur jaundice ", then the initial user mentions again It asks and does not retrieve feature " tongue fur jaundice " in information, then initial user's question information is added in feature " tongue fur jaundice " In, form the target user's question information " cough, fever, tongue fur jaundice and virus flu " for corresponding to the target user.

In one or more embodiments of the application, the user information includes: user's gender and/or age of user Information.The user information can also include: user freely mentions with patient relationship information, user region information, user Ask that number, user charges put question to number, user contact details, user's consulting number, user to extend appreciation doctor visits and each history Any one of order payment amount or any combination.

In one or more embodiments of the application, described problem packet contains: disease description information and/or symptom Description information.Described problem information can also include: the affiliated department's information of disease, problem source identification, question text length Any one of information, problem creation time information and problem creation area information or any combination.

In one or more embodiments of the application, the information about doctor includes: the target disease belonging to doctor Sick department and/or extension department's information, and, doctor is good at disease information.The information about doctor can also include: Yi Shengxing City, physician ratings, doctor where hospital mark, doctor where other information, doctor's age information, doctor academic title's information, doctor Price of answering a question and doctor service any one of person-time information or any combination.

In order to the multidigit doctor for efficiently and accurately recommending matching degree high and strongly professional to user, enable a user to The information such as more accurate and high matching degree disease symptoms explanation and treatment recommendations are enough obtained, it is more that the embodiment of the present application provides one kind The specific embodiment of doctor's recommended method, referring to fig. 4, it includes following content that more doctor's recommended methods, which have:

Step 100: obtaining target user's question information and candidate doctor's column corresponding with target user's question information Table, wherein include multiple information about doctor in candidate doctor's list.

In step 100, more doctor's recommendation apparatus obtain target user's question information online and mention with the target user Ask information corresponding candidate doctor's list.It is understood that candidate's doctor's list is used to store several doctors to this Corresponding relationship between mark and information about doctor.

For example, the doctor identification can be physician names or default unique number.Therefore, referring to shown in table 2 Candidate's doctor's list, the corresponding information about doctor of doctor D1 are as follows: D1, orthopedics and traumatology, orthopedics and traumatology common disease, pain in the loins, numbness, joint Scorching and fracture.The corresponding information about doctor of doctor D2 are as follows: D2, orthopedics and traumatology, Waist disc bulging, pain in waist and lower extremities and osteoproliferation.

Table 2

Step 200: target user's question information being combined with each information about doctor respectively, is formed multiple Forecast sample.

For candidate doctor's list in table 2, more doctor's recommendation apparatus by target user's question information, such as Target user question information Q4, respectively information about doctor corresponding with the doctor D1 and the corresponding information about doctor of the doctor D2 into Row combination, obtain two forecast samples as shown in table 3, respectively forecast sample 1: " Q4, D1, orthopedics and traumatology, orthopedics and traumatology are common Disease, pain in the loins, numbness, arthritis and fracture " and forecast sample 2: " D2, orthopedics and traumatology, Waist disc bulging, pain in waist and lower extremities and hyperostosis It is raw ".

Table 3

Step 300: each forecast sample being inputted into preset prediction model respectively, and according to the prediction model Multiple target information about doctor for recommending to target user are chosen in each output.

It is understood that each forecast sample is inputted preset prediction mould by more doctor's recommendation apparatus respectively Type, and export the adopted rate predicted value of each information about doctor.Then predicted according to the adopted rate of each information about doctor Value chooses several information about doctor as target information about doctor in candidate doctor's list.

Wherein, the prediction model refers to the quantity for prediction, between the things described in mathematical linguistics or formula Relationship.It discloses the inherent law between things to a certain extent, when prediction using it as calculate predicted value it is direct according to According to.Specifically, the prediction model can for regressive prediction model, Kalman prediction model, combination forecasting with And BP (Back-Propagation Network) neural network prediction model etc. is realized, specific choosing of the application to prediction model With with no restrictions.

As can be seen from the above description, more doctor's recommended methods provided by the embodiments of the present application, are putd question to by obtaining target user Information and candidate doctor's list corresponding with target user's question information, wherein include more in candidate doctor's list A information about doctor can effectively improve the specific aim of the data basis of forecast sample, and then can effectively improve model prediction Accuracy；By the way that target user's question information to be combined with each information about doctor respectively, multiple predictions are formed Sample can effectively improve the comprehensive of forecast sample, and the selection diversity for the model prediction that can effectively obtain, pass through by Each forecast sample inputs preset prediction model respectively, and according to each output of the prediction model choose for Multiple target information about doctor that target user recommends efficiently and accurately can recommend matching degree high and strongly professional to user Multidigit doctor allows users to obtain the information such as more accurate and high matching degree disease symptoms explanation and treatment recommendations, into And user experience can be effectively improved.

In order to effectively improve the accuracy of more doctor's recommendation results, in one embodiment, referring to Fig. 5, originally More doctor's recommended methods of application also specifically include following content:

Step 001: recommending log to generate multiple training samples according to history doctor, wherein the history doctor recommends day It include multiple historical user's question informations, multiple target information about doctor recommended respectively to each historical user and each in will The mark whether a target information about doctor is adopted by historical user.

Step 002: the prediction model being trained using the training sample set.

In the foregoing description, more doctor's recommendation apparatus can recommend log according to history doctor offline or online Multiple training samples are generated, and the prediction model is trained using the training sample set, so that the prediction model The forecast sample being made of the information about doctor in target user's question information and candidate doctor's list can received as defeated It is fashionable, the adopted rate predicted value of corresponding information about doctor can be exported.

In order to further increase the efficiency and reliability of the multidigit doctor for recommending matching degree high and strongly professional to user, In a kind of embodiment, referring to Fig. 6, more doctor's recommended methods of the application also specifically include following content:

Step 400: generating the target doctor's recommendation list being made of multiple target information about doctor.

It is understood that the target doctor recommendation list is for storing between each target doctor and information about doctor Corresponding relationship.

Step 500: the target doctor recommendation list being sent to target user, so that the target user is in the target At least one described target information about doctor corresponding target doctor answer target user is selected to put question in doctor's recommendation list Information, to realize the online consultation of doctors.

It is understood that the target doctor recommendation list is sent to target user couple by more doctor's recommendation apparatus The user terminal answered, so that the user terminal shows target doctor's recommendation list, so that the target is used Family is selected in the target doctor recommendation list described in the corresponding target doctor of one or more described target information about doctor answers Target user's question information, to realize the online consultation of doctors.

In order to further increase the acquisition accuracy of candidate doctor's list, in one embodiment, referring to Fig. 7, the application More doctor's recommended methods in step 100 specifically include following content:

Step 101: receiving target user's question information that the target user sends.

Step 102: the candidate doctor's list being made of multiple information about doctor is determined according to target user's question information.

Specifically, the step 102, which is divided, obtains candidate doctor's list or the candidate doctor of non real-time acquisition for real-time online List.

First, the first of the step 102 is as follows with embodiment referring to Fig. 8:

Step 1021A: according to target user's question information, the corresponding target of target user's question information is determined Disease department and/or extension department.

Step 1022A: in the corresponding each information about doctor of the target disease department and/or extension department, symbol is chosen The multiple information about doctor for closing the first preset rules form doctor's list to be selected.

Step 1023A: screening doctor's list to be selected using the second preset rules, obtains the candidate doctor List.

It is understood that the selection first in the step 1022A meets multiple information about doctor of preset rules Include specifically following content:

(1) if present period is in the period in the daytime, it is chosen at each doctor that on-line operation is carried out in the first preset period of time Raw corresponding information about doctor.

(2) if present period is in night-time hours, it is chosen at each doctor that on-line operation is carried out in the second preset period of time Raw corresponding information about doctor.

It is understood that the period in the daytime is daytime period, refer generally in one day sunrise time to sunset time, The night-time hours generally extremely refer to the sunset time in one day to sunrise time.Or the period in the daytime is that can refer to General operations time, the night-time hours are that can refer to the general non-working time.For example, the period in the daytime can be 8 points early To 5 points of evening, the night-time hours can be the period between 5 points to early 8 points of evening.In addition, first preset period of time is less than Second preset period of time, the sum of the period in the daytime and the night-time hours are 24 hours.

It is understood that the second preset rules of the application in the step 1023A are to doctor's list to be selected The first specific embodiment screened includes following content:

(1) from doctor's list to be selected, the corresponding doctor of doctor for having received target user's question information is deleted Raw information, and/or,

(2) from doctor's list to be selected, the doctor for having answered the relevant issues of target user's question information is deleted Raw corresponding information about doctor.

It is understood that the second preset rules of the application in the step 1023A are to doctor's list to be selected Second of the specific embodiment screened includes following content:

(3) from doctor's list to be selected, deletion, which has received and do not answered historical user's question information quantity, is greater than the The corresponding information about doctor of the doctor of one preset value, and/or；

(4) from doctor's list to be selected, the information about doctor in default blacklist is deleted.

Second, second of the step 102 has embodiment as follows referring to Fig. 9:

Step 1021B: it according to target user's question information, determines that target user's question information is corresponding and prestores Corresponding target disease department, hospital and/or extension department.

Step 1022B: it in the corresponding each information about doctor of the target disease department and/or extension department, chooses pre- If profession degree and positive rating meet the corresponding information about doctor of each test doctor of corresponding preset requirement, described in composition Candidate doctor's list.

In order to further increase the accuracy that more doctors recommend, it is based on above content, the target user puts question to letter It also include the corresponding user information of target user and problem information in breath, it is corresponding, in one embodiment, referring to figure 10, the step 200 in more doctor's recommended methods of the application specifically includes following content:

Step 201: feature extraction being carried out to the user information and problem information respectively, respectively obtains the user information User characteristics vector, and, the corresponding problem characteristic vector of described problem information.

Step 202: obtain the corresponding doctor's feature of each information about doctor in candidate doctor's list to Amount.

Step 203: by the corresponding problem characteristic vector of the target user and user characteristics vector respectively with it is each described Doctor's feature vector is combined, and obtains multiple forecast samples.

In order to the multidigit doctor for further recommending matching degree high and strongly professional to user, in one embodiment, Referring to Figure 11, the step 300 in more doctor's recommended methods of the application specifically includes following content:

Step 301: each forecast sample being inputted into preset prediction model respectively, by each of the prediction model Export the adopted rate as each information about doctor in candidate doctor's list.

Step 302: information about doctor corresponding to the adopted rate for being greater than the second preset value is chosen, as to mesh Mark the target information about doctor that user recommends.

For example, can will be adopted if presently described prediction model exports the adopted rate of 10 information about doctor Information about doctor of the rate greater than 30% is as the target information about doctor for recommending to target user.

Furthermore it is also possible to according to the value of the adopted rate of the information about doctor, it is right according to descending sequence is worth The adopted rate of the information about doctor is ranked up, and chooses the corresponding doctor of adopted rate of information about doctor described in top n later Information is as the target information about doctor for recommending to target user.Wherein, N is the integer greater than 1.

As can be seen from the above description, more doctor's recommended methods in the embodiment of the present application, put question to letter by obtaining target user Breath and candidate doctor's list corresponding with target user's question information, wherein include multiple in candidate doctor's list Information about doctor can effectively improve the specific aim of the data basis of forecast sample, and then can effectively improve the standard of model prediction True property；By the way that target user's question information to be combined with each information about doctor respectively, multiple pre- test samples are formed This, can effectively improve the comprehensive of forecast sample, and the selection diversity for the model prediction that can effectively obtain, by will be each A forecast sample inputs preset prediction model respectively, and is chosen according to each output of the prediction model for mesh Multiple target information about doctor that user recommends are marked, efficiently and accurately can recommend high and strongly professional more of matching degree to user Position doctor allows users to obtain the information such as more accurate and high matching degree disease symptoms explanation and treatment recommendations, in turn User experience can be effectively improved.

In order to further illustrate this programme, the application also provides a kind of specific application example of more doctor's recommended methods, In this application example, more doctor's recommended method applications online consultation system as shown in Figure 3 realizes that more doctors recommend Method specifically includes following content:

Wherein, the big data server A 3 carries out model training offline, and particular content is as follows:

S11: big data server A 3 recommends log to generate multiple training samples according to history doctor, wherein the history Multiple target doctors that it includes multiple historical user's question informations that doctor, which recommends in log, is recommended respectively to each historical user The adopted rate of information and each target information about doctor；

S12: big data server A 3 is trained the prediction model using the training sample set.

Wherein, more doctor's recommendation apparatus A1 carry out the particular content of the first way of more doctor's recommendations such as online Under:

S21: back-end server A2 receives target user's question information that the target user sends.

S22: back-end server A2 according to target user's question information, determines that target user's question information is corresponding Target disease department and/or extension department.

S23: back-end server A2 in the corresponding each information about doctor of the target disease department and/or extension department, The multiple information about doctor for meeting the first preset rules are chosen, doctor's list to be selected is formed.It is understood that if at present period In the period in the daytime, then the corresponding information about doctor of each doctor that on-line operation is carried out in the first preset period of time is chosen at；If current Period is in night-time hours, then is chosen at the corresponding information about doctor of each doctor that on-line operation is carried out in the second preset period of time.

S24: back-end server A2 screens doctor's list to be selected using the second preset rules, obtains the time Select doctor's list.It is understood that deletion has received target user's question information from doctor's list to be selected The corresponding information about doctor of doctor, and/or, from doctor's list to be selected, target user's question information has been answered in deletion Relevant issues the corresponding information about doctor of doctor.And from doctor's list to be selected, deletion, which has been received and do not answered, is gone through History user's question information quantity is greater than the corresponding information about doctor of doctor of the first preset value, and/or, from doctor's list to be selected In, delete the information about doctor in default blacklist.

S25: big data server A 3 in target user's question information user information and problem information respectively into Row feature extraction respectively obtains the user characteristics vector of the user information, and, the corresponding problem characteristic of described problem information Vector.

S26: each information about doctor that big data server A 3 obtains in candidate doctor's list is corresponding Doctor's feature vector.

S27: big data server A 3 distinguishes the corresponding problem characteristic vector of the target user and user characteristics vector It is combined with each doctor's feature vector, obtains multiple forecast samples.

S28: each forecast sample is inputted preset prediction model by big data server A 3 respectively, by the prediction Each output of model is putd question to the target user respectively as each information about doctor in candidate doctor's list The matching degree of information.

S29: big data server A 3 chooses the corresponding information about doctor of each matching degree for being greater than the second preset value, makees For multiple target information about doctor for recommending to target user.

S210: big data server A 3 generates the target doctor's recommendation list being made of multiple target information about doctor, And target doctor's recommendation list is sent to back-end server A2.

The target doctor recommendation list is sent to target user by S211: back-end server A2, so that the target user At least one target information about doctor is chosen in the target doctor recommendation list to carry out inline diagnosis.

Wherein, more doctor's recommendation apparatus A1 carry out the particular content of the second way of more doctor's recommendations such as online Under:

S31: back-end server A2 receives target user's question information that the target user sends.

S32: back-end server A2 according to target user's question information, determines that target user's question information is corresponding Prestore corresponding target disease department, hospital and/or extension department；

S33: back-end server A2 in the corresponding each information about doctor of the target disease department and/or extension department, It chooses default profession degree and positive rating meets the corresponding information about doctor of each test doctor of corresponding preset requirement, group At candidate doctor's list.

S34: big data server A 3 carries out feature extraction to the user information and problem information respectively, respectively obtains institute The user characteristics vector of user information is stated, and, the corresponding problem characteristic vector of described problem information.

S35: each information about doctor that big data server A 3 obtains in candidate doctor's list is corresponding Doctor's feature vector.

S36: big data server A 3 distinguishes the corresponding problem characteristic vector of the target user and user characteristics vector It is combined with each doctor's feature vector, obtains multiple forecast samples.

S37: each forecast sample is inputted preset prediction model by big data server A 3 respectively, by the prediction Each output of model is putd question to the target user respectively as each information about doctor in candidate doctor's list The matching degree of information.

S38: big data server A 3 chooses the corresponding information about doctor of each matching degree for being greater than the second preset value, makees For multiple target information about doctor for recommending to target user.

S39: big data server A 3 generates the target doctor's recommendation list being made of multiple target information about doctor, and Target doctor's recommendation list is sent to back-end server A2.

The target doctor recommendation list is sent to target user by S310: back-end server A2, so that the target user At least one target information about doctor is chosen in the target doctor recommendation list to carry out inline diagnosis.

In a kind of citing, referring to Figure 12, by a given user U and it mention the problem of P, recommend this use The N number of doctor D in family.Method is that the set S of a candidate doctor is recalled by certain method, to all doctor D in S, is calculated The feature vector Dv for obtaining each doctor, to user U obtain feature vector Uv, to problem P obtain feature vector Pv, to it is each by User, problem, the vector V=Uv+Pv+Dv of doctor's composition calculate this combination purchased Probability p roba, E=proba* Log (1+price) takes top n according to E value descending sort to all doctors in S, obtains doctor's list of the secondary recommendation.

(1) user terminal B1 is used for:

S1: patient puts question to.

S2: problem content, picture (optional).

S3: creation archives (gender, age).

S11: show ranking results.

(2) back-end server A2 is used for:

S4: creation problem.

S5: candidate doctor d is obtained.

It is understood that expanding department using the department of problem and by the department, filters out and work as in corresponding section room Online doctor of preceding moment.Wherein online condition is any one below meeting: a. doctor in 30min has effectively in doctor terminal Operation (refer to some refreshings problem lists etc. operate), b. crowdsourcing problem hair topic is online, if current time between [7,23] point, Take out a condition expand to 24 hours top hospital and top department doctor.

To the doctor that previous step generates, if current problem is in hair, the topic stage (refers to and is issued to doctor A, but doctor is not yet Get), filter out doctor A；Filter out current problem claims doctor B；It is more to filter out all topics caused by a problem (for example user buys since a problem p0 and asks a topic more the reply doctor asked, generates two new problems p1, p2, and from p1 Ask of one topic generates p3 more, then filters out p0, p1, p2, the corresponding doctor of p3).

It filters out and does not reply the doctor that problem is greater than 3, filter out a topic and ask blacklist doctor more.

Non real-time filter condition: doctor is firstly the need of meeting the following conditions: professional degree >=90, hospital id not [' Cygzyyzjt', ' bu4zfg76s1pel85d '] (spring rain internal indicator, test data) inner, non-test doctor, 1 positive rating >= 92 (90 are new doctors).

S10: ranking results are returned to.

(3) big data server A 3 is used for:

S01: delineation training set, data preparation a.

The topic data asked bought to a pile user and do not bought, are formatted into [V, label] set, wherein V more =Uv+Pv+Dv, label=0/1,1 indicates purchase, and 0 indicates not buy.

S02: having found and calculates assemblage characteristic b.

It is understood that combining to all [V, label], assemblage characteristic, such as Pekinese user are found and calculated Compare this rule of doctor for tending to buy Grade A hospital, can extract and be come out as individual feature extraction, for example claim For V1, V=V+V1 is corrected.

S03: fitting formula obtains optimal solution c.

It combines, calculates it is understood that owning [V, label] using a formula fitting similar to y=WV+b Optimal W and b, W are the weight of each feature, and b is the biasing of each feature.

S6 is as follows:

S6-1:(S4 problem characteristic vector Pv e) is obtained.

Problem characteristic includes: level-one department, second level department, put question to platform, text size, creation time, province, city, Whether there is or not pictures, etc..

S6-2:(S3 user characteristics vector Uv f) is obtained.

User characteristics include: gender, age, quizmaster and the relationship of patient put question to free problem number, put question to payment Problem number, consulting set meal number, send regard number, service total number, pay invoice sum, payment problem branch at phone number Pay number, phone-payment amount, consulting set meal payment amount, send that the total amount of regard, first order payment amount, the last one orders Single payment amount, total payoff amount, level payment amount, payment problem be averaged amount, phone be averaged amount, seek advice from set meal and be averaged Amount send regard level payment amount, etc..

S6-3:(S5 doctor's feature vector Dv g) is obtained.

Doctor's feature includes: gender, age, academic title, province, city, hospital ID, physician ratings, picture and text price, service people It is secondary, whether top hospital, vast city of whether going up north, whether provincial capital.

S7:(S6-1, S6-2 and S6-3) composition vector V=Uv+Pv+Dv.

S8:(S03 and S7) calculate purchase probability proba.

S9: sequence: proba*log (1+price).

As can be seen from the above description, more doctor's recommended methods that the application application example provides, are mentioned by obtaining target user Ask information and candidate doctor's list corresponding with target user's question information, wherein include in candidate doctor's list Multiple information about doctor can effectively improve the specific aim of the data basis of forecast sample, and then can effectively improve model prediction Accuracy；By the way that target user's question information to be combined with each information about doctor respectively, formed multiple pre- Test sample sheet can effectively improve the comprehensive of forecast sample, and the selection diversity for the model prediction that can effectively obtain, pass through Each forecast sample is inputted into preset prediction model respectively, and is used for according to each output of prediction model selection The multiple target information about doctor recommended to target user efficiently and accurately can recommend matching degree high and strongly professional to user Multidigit doctor, allow users to obtain the information such as more accurate and high matching degree disease symptoms explanation and treatment recommendations, And then user experience can be effectively improved.

In order to the multidigit doctor for efficiently and accurately recommending matching degree high and strongly professional to user, enable a user to It enough obtains information, the embodiment of the present application such as more accurate and high matching degree disease symptoms explanation and treatment recommendations and a kind of energy is provided The specific embodiment for enough realizing more doctor's recommendation apparatus of full content in more doctor's recommended methods, referring to Figure 13, institute Stating more doctor's recommendation apparatus to have includes following content:

Data obtaining module 10, for obtaining target user's question information and corresponding with target user's question information Candidate doctor's list, wherein include multiple information about doctor in candidate doctor's list.

Forecast sample determining module 20, for by target user's question information respectively with each information about doctor into Row combination, forms multiple forecast samples.

Multiple target doctor recommending module 30, for each forecast sample to be inputted preset prediction model respectively, and Multiple target information about doctor for recommending to target user are chosen according to each output of the prediction model.

The embodiment of more doctor's recommendation apparatus provided by the present application specifically can be used for executing more doctors in above-described embodiment The process flow of the embodiment of raw recommended method, details are not described herein for function, is referred to the detailed of above method embodiment Description.

As can be seen from the above description, more doctor's recommendation apparatus provided by the embodiments of the present application, are putd question to by obtaining target user Information and candidate doctor's list corresponding with target user's question information, wherein include more in candidate doctor's list A information about doctor can effectively improve the specific aim of the data basis of forecast sample, and then can effectively improve model prediction Accuracy；By the way that target user's question information to be combined with each information about doctor respectively, multiple predictions are formed Sample can effectively improve the comprehensive of forecast sample, and the selection diversity for the model prediction that can effectively obtain, pass through by Each forecast sample inputs preset prediction model respectively, and according to each output of the prediction model choose for Multiple target information about doctor that target user recommends efficiently and accurately can recommend matching degree high and strongly professional to user Multidigit doctor allows users to obtain the information such as more accurate and high matching degree disease symptoms explanation and treatment recommendations, into And user experience can be effectively improved.

In order to effectively improve the accuracy of more doctor's recommendation results, in one embodiment, the application's is more Doctor's recommendation apparatus recommends log to generate multiple training samples also particularly useful for according to history doctor, wherein the history doctor To recommend in log include multiple historical user's question informations, multiple target information about doctor for recommending respectively to each historical user And the mark whether each target information about doctor is adopted by historical user, using the training sample set to the prediction Model is trained.

In order to further increase the efficiency and reliability of the multidigit doctor for recommending matching degree high and strongly professional to user, In a kind of embodiment, more doctor's recommendation apparatus of the application are made of multiple target information about doctor also particularly useful for generating The target doctor recommendation list is sent to target user, so that the target user is in the mesh by target doctor's recommendation list At least one described target information about doctor corresponding target doctor answer target user is selected to mention in mark doctor's recommendation list Information is asked, to realize the online consultation of doctors.

In order to further increase the acquisition accuracy of candidate doctor's list, in one embodiment, more doctors of the application Data obtaining module 10 in recommendation apparatus is specifically used for receiving target user's question information that the target user sends, The candidate doctor's list being made of multiple information about doctor is determined according to target user's question information.

In order to further increase the accuracy that more doctors recommend, it is based on above content, the target user puts question to letter It also include the corresponding user information of target user and problem information in breath, it is corresponding, in one embodiment, the application's Forecast sample determining module 20 in more doctor's recommendation apparatus is specifically used for carrying out the user information and problem information respectively Feature extraction respectively obtains the user characteristics vector of the user information, and, the corresponding problem characteristic of described problem information to Amount obtains the corresponding doctor's feature vector of each information about doctor in candidate doctor's list, and will be described The corresponding problem characteristic vector of target user and user characteristics vector are combined with each doctor's feature vector respectively, are obtained To multiple forecast samples.

In order to the multidigit doctor for further recommending matching degree high and strongly professional to user, in one embodiment, Multiple target doctor recommending module 30 in more doctor's recommendation apparatus of the application is specifically used for distinguishing each forecast sample Preset prediction model is inputted, using each output of the prediction model as each doctor in candidate doctor's list The adopted rate of raw information chooses information about doctor corresponding to the adopted rate for being greater than the second preset value, as to The target information about doctor that target user recommends.

In order to the multidigit doctor for efficiently and accurately recommending matching degree high and strongly professional to user, enable a user to It enough obtains information, the embodiment of the present application such as more accurate and high matching degree disease symptoms explanation and treatment recommendations and a kind of energy is provided Enough realize the specific embodiment of the online consultation system of full content in more doctor's recommended methods, the online consultation of doctors system It includes following content that system, which has: user terminal and more doctor's recommendation apparatus, and the user terminal is used for described More doctor's recommendation apparatus send target user's question information, and, it receives and shows the more of more doctor's recommendation apparatus transmissions A target information about doctor, so that target user selects the corresponding target doctor of at least one described target information about doctor to answer Target user's question information, to realize the online consultation of doctors.

As can be seen from the above description, online consultation system provided by the embodiments of the present application, puts question to letter by obtaining target user Breath and candidate doctor's list corresponding with target user's question information, wherein include multiple in candidate doctor's list Information about doctor can effectively improve the specific aim of the data basis of forecast sample, and then can effectively improve the standard of model prediction True property；By the way that target user's question information to be combined with each information about doctor respectively, multiple pre- test samples are formed This, can effectively improve the comprehensive of forecast sample, and the selection diversity for the model prediction that can effectively obtain, by will be each A forecast sample inputs preset prediction model respectively, and is chosen according to each output of the prediction model for mesh Multiple target information about doctor that user recommends are marked, efficiently and accurately can recommend high and strongly professional more of matching degree to user Position doctor allows users to obtain the information such as more accurate and high matching degree disease symptoms explanation and treatment recommendations, in turn User experience can be effectively improved.

Embodiments herein, which also provides, can be realized Overall Steps in more doctor's recommended methods in above-described embodiment The specific embodiment of a kind of electronic equipment, referring to Figure 14, the electronic equipment specifically includes following content:

Processor (processor) 601, memory (memory) 602, communication interface (Communications Interface) 603 and bus 604；

Wherein, the processor 601, memory 602, communication interface 603 complete mutual lead to by the bus 604 Letter；The communication interface 603 is for realizing user terminal, doctor terminal, back-end server, big number server and other participations Information transmission between mechanism；

The processor 601 is used to call the computer program in the memory 602, and the processor executes the meter The Overall Steps in more doctor's recommended methods in above-described embodiment are realized when calculation machine program, for example, the processor executes institute Following step is realized when stating computer program:

As can be seen from the above description, electronic equipment provided by the embodiments of the present application, by obtain target user's question information with And candidate doctor's list corresponding with target user's question information, wherein include multiple doctors in candidate doctor's list Information can effectively improve the specific aim of the data basis of forecast sample, and then can effectively improve the accuracy of model prediction； By the way that target user's question information to be combined with each information about doctor respectively, multiple forecast samples, energy are formed The comprehensive of forecast sample, and the selection diversity for the model prediction that can effectively obtain enough are effectively improved, by by each institute It states forecast sample and inputs preset prediction model respectively, and chosen according to each output of the prediction model for being used to target Multiple target information about doctor that family is recommended, the multidigit doctor that efficiently and accurately matching degree can be recommended high and strongly professional to user It is raw, it allows users to obtain the information such as more accurate and high matching degree disease symptoms explanation and treatment recommendations, and then can Effectively improve user experience.

Embodiments herein, which also provides, can be realized Overall Steps in more doctor's recommended methods in above-described embodiment A kind of computer readable storage medium is stored with computer program on the computer readable storage medium, the computer program The Overall Steps of more doctor's recommended methods in above-described embodiment are realized when being executed by processor, for example, the processor executes Following step is realized when the computer program:

As can be seen from the above description, computer readable storage medium provided by the embodiments of the present application, by obtaining target user Question information and candidate doctor's list corresponding with target user's question information, wherein include in candidate doctor's list There are multiple information about doctor, the specific aim of the data basis of forecast sample can be effectively improved, and then it is pre- to effectively improve model The accuracy of survey；By the way that target user's question information to be combined with each information about doctor respectively, formed multiple Forecast sample can effectively improve the comprehensive of forecast sample, and the selection diversity for the model prediction that can effectively obtain, and lead to It crosses and each forecast sample is inputted into preset prediction model respectively, and chosen and used according to each output of the prediction model In the multiple target information about doctor recommended to target user, efficiently and accurately matching degree can be recommended high and professional to user Strong multidigit doctor allows users to obtain the letters such as more accurate and high matching degree disease symptoms explanation and treatment recommendations Breath, and then user experience can be effectively improved.

All the embodiments in this specification are described in a progressive manner, same and similar portion between each embodiment Dividing may refer to each other, and each embodiment focuses on the differences from other embodiments.Especially for hardware+ For program class embodiment, since it is substantially similar to the method embodiment, so being described relatively simple, related place is referring to side The part of method embodiment illustrates.

It is above-mentioned that this specification specific embodiment is described.Other embodiments are in the scope of the appended claims It is interior.In some cases, the movement recorded in detail in the claims or step can be come according to the sequence being different from embodiment It executes and desired result still may be implemented.In addition, process depicted in the drawing not necessarily require show it is specific suitable Sequence or consecutive order are just able to achieve desired result.In some embodiments, multitasking and parallel processing be also can With or may be advantageous.

Although this application provides the method operating procedure as described in embodiment or flow chart, based on conventional or noninvasive The labour for the property made may include more or less operating procedure.The step of enumerating in embodiment sequence is only numerous steps One of execution sequence mode, does not represent and unique executes sequence.It, can when device or client production in practice executes To execute or parallel execute (such as at parallel processor or multithreading according to embodiment or method shown in the drawings sequence The environment of reason).

System, device, module or the unit that above-described embodiment illustrates can specifically realize by computer chip or entity, Or it is realized by the product with certain function.It is a kind of typically to realize that equipment is computer.Specifically, computer for example may be used Think personal computer, laptop computer, vehicle-mounted human-computer interaction device, cellular phone, camera phone, smart phone, individual Digital assistants, media player, navigation equipment, electronic mail equipment, game console, tablet computer, wearable device or The combination of any equipment in these equipment of person.

These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates, Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or The function of being specified in multiple boxes.

These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one The step of function of being specified in a box or multiple boxes.

In a typical configuration, calculating equipment includes one or more processors (CPU), input/output interface, net Network interface and memory.

Memory may include the non-volatile memory in computer-readable medium, random access memory (RAM) and/or The forms such as Nonvolatile memory, such as read-only memory (ROM) or flash memory (flash RAM).Memory is computer-readable medium Example.

Computer-readable medium includes permanent and non-permanent, removable and non-removable media can be by any method Or technology come realize information store.Information can be computer readable instructions, data structure, the module of program or other data. The example of the storage medium of computer includes, but are not limited to phase change memory (PRAM), static random access memory (SRAM), moves State random access memory (DRAM), other kinds of random access memory (RAM), read-only memory (ROM), electric erasable Programmable read only memory (EEPROM), flash memory or other memory techniques, read-only disc read only memory (CD-ROM) (CD-ROM), Digital versatile disc (DVD) or other optical storage, magnetic cassettes, tape magnetic disk storage or other magnetic storage devices Or any other non-transmission medium, can be used for storage can be accessed by a computing device information.As defined in this article, it calculates Machine readable medium does not include temporary computer readable media (transitory media), such as the data-signal and carrier wave of modulation.

It will be understood by those skilled in the art that the embodiment of this specification can provide as the production of method, system or computer program Product.Therefore, in terms of this specification embodiment can be used complete hardware embodiment, complete software embodiment or combine software and hardware Embodiment form.

This specification embodiment can describe in the general context of computer-executable instructions executed by a computer, Such as program module.Generally, program module includes routines performing specific tasks or implementing specific abstract data types, journey Sequence, object, component, data structure etc..This specification embodiment can also be practiced in a distributed computing environment, in these points Cloth calculates in environment, by executing task by the connected remote processing devices of communication network.In distributed computing ring In border, program module can be located in the local and remote computer storage media including storage equipment.

All the embodiments in this specification are described in a progressive manner, same and similar portion between each embodiment Dividing may refer to each other, and each embodiment focuses on the differences from other embodiments.Especially for system reality For applying example, since it is substantially similar to the method embodiment, so being described relatively simple, related place is referring to embodiment of the method Part explanation.In the description of this specification, reference term " one embodiment ", " some embodiments ", " example ", The description of " specific example " or " some examples " etc. means specific features described in conjunction with this embodiment or example, structure, material Or feature is contained at least one embodiment or example of this specification embodiment.In the present specification, to above-mentioned term Schematic representation be necessarily directed to identical embodiment or example.Moreover, description specific features, structure, material or Person's feature may be combined in any suitable manner in any one or more of the embodiments or examples.In addition, in not conflicting feelings Under condition, those skilled in the art by different embodiments or examples described in this specification and different embodiment or can show The feature of example is combined.

The foregoing is merely the embodiments of this specification, are not limited to this specification embodiment.For ability For field technique personnel, this specification embodiment can have various modifications and variations.It is all this specification embodiment spirit and Any modification, equivalent replacement, improvement and so within principle should be included in the scope of the claims of this specification embodiment Within.

Claims

1. a kind of more doctor's recommended methods characterized by comprising

Obtain target user's question information and candidate doctor's list corresponding with target user's question information, wherein the time Selecting includes multiple information about doctor in doctor's list；

Target user's question information is combined with each information about doctor respectively, forms multiple forecast samples；

And each forecast sample is inputted into preset prediction model respectively, and according to each defeated of the prediction model Multiple target information about doctor for recommending to target user are chosen out.

2. more doctor's recommended methods according to claim 1, which is characterized in that further include:

Log is recommended to generate multiple training samples according to history doctor, wherein the history doctor recommends in log to include more A historical user's question information, the multiple target information about doctor recommended respectively to each historical user and each target doctor The adopted rate of raw information；

The prediction model is trained using the training sample set.

3. more doctor's recommended methods according to claim 1, which is characterized in that further include:

The target doctor recommendation list is sent to target user, so that the target user is in the target doctor recommendation list At least one described target user's question information of the corresponding target doctor answer of target information about doctor of middle selection, is existed with realizing The line consultation of doctors.

4. more doctor's recommended methods according to claim 1, which is characterized in that the acquisition target user question information with And candidate doctor's list corresponding with target user's question information, comprising:

Receive target user's question information that the target user sends；

5. more doctor's recommended methods according to claim 4, which is characterized in that described putd question to according to the target user is believed Breath determines corresponding candidate doctor's list, comprising:

According to target user's question information, determine the corresponding target disease department of target user's question information and/or Extend department；

In the corresponding each information about doctor of the target disease department and/or extension department, selection meets the first preset rules Multiple information about doctor, form doctor's list to be selected；

6. more doctor's recommended methods according to claim 5, which is characterized in that the selection first meets preset rules Multiple information about doctor, comprising:

If present period is in the period in the daytime, each doctor for being chosen at progress on-line operation in the first preset period of time is corresponding Information about doctor；

If present period is in night-time hours, each doctor for being chosen at progress on-line operation in the second preset period of time is corresponding Information about doctor；

Wherein, first preset period of time is less than second preset period of time, the sum of the period in the daytime and the night-time hours It is 24 hours.

7. more doctor's recommended methods according to claim 5, which is characterized in that the second preset rules of the application are to described Doctor's list to be selected is screened, comprising:

From in doctor's list to be selected, the corresponding information about doctor of doctor for having received target user's question information is deleted, And/or；

From in doctor's list to be selected, the doctor for deleting the relevant issues for having answered target user's question information is corresponding Information about doctor.

8. more doctor's recommended methods according to claim 5, which is characterized in that the second preset rules of the application are to described Doctor's list to be selected is screened, comprising:

From in doctor's list to be selected, deletion has received and has not answered historical user's question information quantity greater than the first preset value The corresponding information about doctor of doctor, and/or；

9. more doctor's recommended methods according to claim 4, which is characterized in that described putd question to according to the target user is believed Breath determines corresponding candidate doctor's list, comprising:

According to target user's question information, determines that target user's question information is corresponding and prestore the corresponding target of hospital Disease department and/or extension department；

In the corresponding each information about doctor of the target disease department and/or extension department, default profession degree and favorable comment are chosen Rate meets the corresponding information about doctor of each test doctor of corresponding preset requirement, forms candidate doctor's list.

10. more doctor's recommended methods according to claim 1, which is characterized in that in target user's question information also It include the corresponding user information of target user and problem information；

It is described to be combined target user's question information with each information about doctor respectively, form multiple pre- test samples This, comprising:

Feature extraction is carried out respectively to the user information and problem information, respectively obtain the user characteristics of the user information to Amount, and, the corresponding problem characteristic vector of described problem information；

By the corresponding problem characteristic vector of the target user and user characteristics vector respectively with each doctor's feature vector It is combined, obtains multiple forecast samples.

11. more doctor's recommended methods according to claim 10, which is characterized in that the user information includes: user's property Other and/or age of user information.

12. more doctor's recommended methods according to claim 10, which is characterized in that described problem information includes: that disease is retouched State information and/or symptom description information.

13. more doctor's recommended methods according to any of claims 1 to 10, which is characterized in that the information about doctor packet It includes: the target disease department belonging to doctor and/or extension department's information, and, doctor is good at disease information.

14. more doctor's recommended methods according to claim 1, which is characterized in that described by each forecast sample point Preset prediction model is not inputted, and multiple for what is recommended to target user according to each output of prediction model selection Target information about doctor, comprising:

Each forecast sample is inputted into preset prediction model respectively, using each output of the prediction model as described in The adopted rate of each information about doctor in candidate doctor's list；

It chooses and is greater than information about doctor corresponding to the adopted rate of the second preset value, as recommending to target user Target information about doctor.

15. a kind of more doctor's recommendation apparatus characterized by comprising

Data obtaining module, for obtaining target user's question information and candidate doctor corresponding with target user's question information Raw list, wherein include multiple information about doctor in candidate doctor's list；

Forecast sample determining module, for target user's question information to be carried out group with each information about doctor respectively It closes, forms multiple forecast samples；

Multiple target doctor's recommending module, for each forecast sample to be inputted preset prediction model respectively, and according to institute State multiple target information about doctor of each output selection of prediction model for recommending to target user.

16. a kind of online consultation system characterized by comprising user terminal and more doctors as claimed in claim 15 Recommendation apparatus；

The user terminal is used to send target user's question informations to more doctor's recommendation apparatus, and, it receives and shows Multiple target information about doctor that more doctor's recommendation apparatus are sent, so that target user selects at least one described target Information about doctor corresponding target doctor answer target user's question information, to realize the online consultation of doctors.

17. a kind of electronic equipment including memory, processor and stores the calculating that can be run on a memory and on a processor Machine program, which is characterized in that realize that any one of claim 1 to 14 more doctors push away when the processor executes described program The step of recommending method.

18. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program quilt The step of claim 1 to 14 described in any item more doctor's recommended methods are realized when processor executes.