CN112287232A - Method and device for generating recommendation information - Google Patents

Method and device for generating recommendation information Download PDF

Info

Publication number
CN112287232A
CN112287232A CN202011239590.5A CN202011239590A CN112287232A CN 112287232 A CN112287232 A CN 112287232A CN 202011239590 A CN202011239590 A CN 202011239590A CN 112287232 A CN112287232 A CN 112287232A
Authority
CN
China
Prior art keywords
information
feature
user
recommendation
current
Prior art date
Legal status (The legal status 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 status listed.)
Granted
Application number
CN202011239590.5A
Other languages
Chinese (zh)
Other versions
CN112287232B (en
Inventor
刘宗节
王永鹏
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Jingdong Tuoxian Technology Co Ltd
Original Assignee
Beijing Jingdong Tuoxian Technology Co Ltd
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 Beijing Jingdong Tuoxian Technology Co Ltd filed Critical Beijing Jingdong Tuoxian Technology Co Ltd
Priority to CN202011239590.5A priority Critical patent/CN112287232B/en
Publication of CN112287232A publication Critical patent/CN112287232A/en
Priority to PCT/CN2021/128399 priority patent/WO2022095893A1/en
Application granted granted Critical
Publication of CN112287232B publication Critical patent/CN112287232B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33—Querying
    • G06F16/332—Query formulation
    • G06F16/3329—Natural language query formulation or dialogue systems
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33—Querying
    • G06F16/3331—Query processing
    • G06F16/334—Query execution
    • G06F16/3344—Query execution using natural language analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • 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
    • G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Computational Linguistics (AREA)
  • Mathematical Physics (AREA)
  • Biomedical Technology (AREA)
  • General Health & Medical Sciences (AREA)
  • Biophysics (AREA)
  • Computing Systems (AREA)
  • Molecular Biology (AREA)
  • Software Systems (AREA)
  • Evolutionary Computation (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Medical Informatics (AREA)
  • Public Health (AREA)
  • Pathology (AREA)
  • Epidemiology (AREA)
  • Primary Health Care (AREA)
  • Human Computer Interaction (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The application discloses a method and a device for generating recommendation information, electronic equipment and a computer readable storage medium, and relates to the field of artificial intelligence, the technical field of natural language processing, the technical field of knowledge maps and the technical field of big data. The specific implementation scheme is as follows: the method comprises the steps of responding to a situation that a user does not visit the user for the first time, obtaining current description information input by the user and historical description information of the user, inputting the historical description information into a historical feature matching model for processing to obtain historical feature information, inquiring current feature information corresponding to the current description information, splicing the historical feature information and the current feature information to generate recommended feature information, then adopting a predetermined recommended feature recognition neural network to recognize the recommended feature information, obtaining first recommended information, sending the first recommended information to the user, determining the recommended information by combining the historical description information of the user, providing the recommended information for the user more accurately, and improving the quality of the recommended information.

Description

Method and device for generating recommendation information
Technical Field
The present application relates to the field of artificial intelligence, and in particular, to the field of natural language processing and big data processing technologies, and in particular, to a method and an apparatus for generating recommendation information, an electronic device, and a computer-readable storage medium.
Background
When the patient asks for a doctor through the internet hospital, the triage system can simply interact with the patient so as to acquire the case information of the patient in advance, then the triage system can carry out natural language processing according to the case information of the patient, and intelligent triage departments are carried out so as to replace the traditional rule matching, thereby bringing about multi-aspect benefits.
In the prior art, a language identification neural network is usually adopted to identify the current description of a user, and department assignment and recommendation are completed according to the obtained identification result and department label matching.
Disclosure of Invention
The application provides a method and a device for generating recommendation information, electronic equipment and a storage medium.
In a first aspect, an embodiment of the present application provides a method for generating recommendation information, including: in response to the fact that the user is not the first-time access user, current description information input by the user and historical description information of the user are obtained; inputting the historical description information into a historical feature matching model for processing to obtain historical feature information; inquiring current characteristic information corresponding to the current description information; splicing the historical characteristic information and the current characteristic information to generate recommended characteristic information; and identifying the recommendation characteristic information by adopting a predetermined recommendation characteristic identification neural network to obtain first recommendation information, and sending the first recommendation information to the user.
In some embodiments, further comprising: responding to the user as a first access user, and acquiring current description information input by the user; inquiring current characteristic information corresponding to the current description information; and identifying the current characteristic information by adopting the recommended characteristic identification neural network to obtain second recommended information, and sending the second recommended information to the user.
In some embodiments, inputting the history description information into the history feature matching model for processing, and obtaining the history feature information includes: acquiring user description information and recommendation result description information in the historical description information of the user; after the user description information is normalized, inputting the user description information into a first deep learning neural network layer forming the historical feature matching model for processing to generate first feature information; inputting the recommended result description information to a multilayer bidirectional converter encoder layer forming the historical characteristic matching model to generate second characteristic information; and inputting the first characteristic information and the second characteristic information into a second deep learning network layer forming the historical characteristic matching model for splicing to obtain the historical characteristic information.
In some embodiments, querying the current feature information corresponding to the current description information includes: and inputting the current description information into a multi-layer bidirectional converter encoder layer for outputting the current feature information, and inquiring the current feature information corresponding to the current description information.
In some embodiments, the obtaining of the historical feature information and the current feature information, and the generating of the recommended feature information by splicing includes: respectively acquiring historical characteristic vectors and current characteristic vectors corresponding to the historical characteristic information and the current characteristic information; inputting the historical feature vector and the current feature vector into a splicing deep learning neural network layer for splicing, and splicing to generate a recommended feature vector as recommended feature information; the dimension number of the recommended feature vector is obtained based on the sum of the dimension number of the historical feature vector and the dimension number of the current feature vector.
In some embodiments, identifying recommendation feature information using a predetermined recommendation feature identification neural network to obtain first recommendation information, and sending the first recommendation information to the user includes: adopting a predetermined recommendation characteristic recognition neural network to recognize the recommendation characteristic information, and iteratively generating a plurality of recommendation information; wherein the loss function of the recommended feature recognition neural network is a cross entropy function; and stopping iterative identification after the loss function is determined to be converged, determining the finally obtained recommendation information as the first recommendation information, and sending the first recommendation information to the user.
In a second aspect, an embodiment of the present application provides an apparatus for generating recommendation information, including: a description information acquisition unit configured to acquire current description information input by a user and historical description information of the user in response to determining that the user does not visit the user for the first time; the history feature matching unit is configured to input the history description information into a history feature matching model for processing to obtain history feature information; a current feature query unit configured to query current feature information corresponding to the current description information; a recommended feature generation unit configured to splice the historical feature information and the current feature information to generate recommended feature information; and the recommendation information sending unit is configured to identify the recommendation characteristic information by adopting a predetermined recommendation characteristic identification neural network, obtain first recommendation information and send the first recommendation information to the user.
In some embodiments, the description information obtaining unit is further configured to, in response to the user being a first access user, obtain that the current description information input by the user includes; the recommended feature generation unit is further configured to query current feature information corresponding to the current description information; the recommendation information sending unit is further configured to identify the current feature information by using the recommendation feature recognition neural network, obtain second recommendation information, and send the second recommendation information to the user.
In some embodiments, the history feature matching unit includes: a history information obtaining subunit configured to obtain user description information and recommendation result description information in the history description information of the user; the first characteristic information generating subunit is configured to normalize the user description information, input the normalized user description information to a first deep learning neural network layer forming the historical characteristic matching model, and process the normalized user description information to generate first characteristic information; a second feature information generation subunit configured to input the recommendation result description information to a multilayer bidirectional converter encoder layer constituting the historical feature matching model, and generate second feature information; and the historical characteristic information generating subunit is configured to input the first characteristic information and the second characteristic information into a second deep learning network layer forming the historical characteristic matching model for splicing to obtain the historical characteristic information.
In some embodiments, the current feature query unit is further configured to: and inputting the current description information into a multi-layer bidirectional converter encoder layer for outputting the current feature information, and inquiring the current feature information corresponding to the current description information.
In some embodiments, the recommendation feature generation unit comprises: a feature vector generation subunit configured to obtain a history feature vector and a current feature vector corresponding to the history feature information and the current feature information, respectively; the recommended vector splicing subunit is configured to input the historical feature vector and the current feature vector to a splicing deep learning neural network layer for splicing, and generate a recommended feature vector as recommended feature information through splicing; the dimension number of the recommended feature vector is obtained based on the sum of the dimension number of the historical feature vector and the dimension number of the current feature vector.
In some embodiments, the recommendation information sending unit includes: the recommendation information generation subunit is configured to identify the recommendation characteristic information by adopting a predetermined recommendation characteristic identification neural network, and iteratively generate a plurality of recommendation information; wherein the loss function of the recommended feature recognition neural network is a cross entropy function; a recommendation information determination subunit configured to stop iterative identification in response to determining that the loss function converges, and determine finally obtained recommendation information as the first recommendation information; and the recommendation information sending subunit is configured to send the first recommendation information to the user.
In a third aspect, an embodiment of the present application provides an electronic device, including: at least one processor; and a memory communicatively coupled to the at least one processor; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method for generating recommendation information described in any implementation manner.
In a fourth aspect, embodiments of the present application provide a non-transitory computer readable storage medium having computer instructions stored thereon, comprising: the computer instructions are used for causing the computer to execute the method for generating recommendation information described in any implementation mode.
According to the method and the device, after the user is determined not to access the user for the first time, current description information input by the user and historical description information of the user are obtained, the historical description information is input into a historical feature matching model to be processed, historical feature information is obtained, current feature information corresponding to the current description information is inquired, the historical feature information and the current feature information are spliced to generate recommended feature information, then the recommended feature information is identified by adopting a predetermined recommended feature identification neural network to obtain first recommended information, the first recommended information is sent to the user, the recommended information is determined by combining the historical description information of the user, the recommended information is provided for the user more accurately, and the quality of the recommended information is improved.
It should be understood that the statements in this section do not necessarily identify key or critical features of the embodiments of the present application, nor do they limit the scope of the present application. Other features of the present application will become apparent from the following description.
Drawings
The drawings are included to provide a better understanding of the present solution and are not intended to limit the present application. Wherein:
FIG. 1 is an exemplary system architecture to which embodiments of the present application may be applied;
FIG. 2 is a flow diagram of one embodiment of a method of generating recommendation information according to the present application;
FIG. 3 is a flow diagram of one implementation of obtaining historical feature information in a method of generating recommendation information according to the present application;
FIG. 4 is a schematic block diagram illustrating one embodiment of an apparatus for generating recommendation information according to the present application;
FIG. 5 is a block diagram of an electronic device suitable for use in implementing a method of generating recommendation information according to an embodiment of the present application.
Detailed Description
The following description of the exemplary embodiments of the present application, taken in conjunction with the accompanying drawings, includes various details of the embodiments of the application for the understanding of the same, which are to be considered exemplary only. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the present application. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.
It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict. The present application will be described in detail below with reference to the embodiments with reference to the attached drawings.
Fig. 1 illustrates an exemplary system architecture 100 to which embodiments of the method, apparatus, electronic device, and computer-readable storage medium for generating recommendation information of the present application may be applied.
As shown in fig. 1, the system architecture 100 may include terminal devices 101, 102, 103, a network 104, and a server 105. The network 104 serves as a medium for providing communication links between the terminal devices 101, 102, 103 and the server 105. Network 104 may include various connection types, such as wired, wireless communication links, or fiber optic cables, to name a few.
The user may use the terminal devices 101, 102, 103 to interact with the server 105 via the network 104 for the purpose of sending user input information, etc. The terminal devices 101, 102, 103 may have applications related to receiving scheduling instructions and recommendation information installed thereon, such as a service reservation application, a scene query application, a distribution guidance application, and the like.
The terminal apparatuses 101, 102, and 103 may be hardware or software. Hardware, various electronic devices with display screens are possible, including but not limited to smart phones, tablet computers, laptop portable computers, desktop computers, and the like. When the terminal apparatuses 101, 102, 103 are software, they can be installed in the electronic apparatuses listed above. It may be implemented as a plurality of software or software modules (for example, to transmit input information of a user, etc.), or may be implemented as a single software or software module. And is not particularly limited herein.
The server 105 may be a server providing various services, such as a server providing push information for the terminal devices 101, 102, 103. For example, after a user is determined not to access the user for the first time, current description information input by the user and historical description information of the user are obtained, the historical description information is input into a historical feature matching model to be processed to obtain historical feature information, current feature information corresponding to the current description information is inquired, the historical feature information and the current feature information are spliced, and recommended feature information is generated; and identifying the recommendation characteristic information by adopting a predetermined recommendation characteristic identification neural network to obtain first recommendation information and sending the first recommendation information to the user.
It should be noted that the method for generating recommendation information provided in the embodiments of the present application is generally executed by the server 105, and accordingly, the apparatus for generating recommendation information is generally disposed in the server 105.
The server may be hardware or software. When the server is hardware, it may be implemented as a distributed server cluster formed by multiple servers, or may be implemented as a single server. When the server is software, it may be implemented as multiple pieces of software or software modules, for example, to provide distributed services, or as a single piece of software or software module. And is not particularly limited herein.
Further, the method of generating the recommendation information may be executed by the terminal apparatuses 101, 102, 103, and accordingly, the apparatus for generating the recommendation information may be provided in the terminal apparatuses 101, 102, 103. At this point, the exemplary system architecture 100 may also not include the server 105 and the network 104.
It should be understood that the number of terminal devices, networks, and servers in fig. 1 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation.
With continued reference to FIG. 2, a flow 200 of one embodiment of a method of generating recommendation information in accordance with the present application is shown. The method for generating the recommendation information comprises the following steps:
step 201, in response to determining that the user is not the first access user, acquiring current description information input by the user and historical description information of the user.
In this embodiment, an executing entity (for example, the server 105 shown in fig. 1) of the method for generating recommendation information may first determine whether a user receiving push information is a first-time access user, and may generally determine according to whether a history access record of the user is stored in a local area of the executing entity, or may prove the user in a human-computer interaction manner to assist the determining of the executing entity.
When the user is determined to be a non-first-time-access user, current description information input by the user is acquired, the current description information refers to description information which is provided by the user at this time and is expected to acquire corresponding matched recommendation information, and generally, the current description information is input by the user through a used human-computer interaction device (for example, terminal devices 101, 102, 103 shown in fig. 1), and then is communicated with the execution main body (for example, through a network 104) to complete transmission of the current description information, so that the user can complete acquisition of the recommendation information through the used human-computer interaction device.
For example, when the recommendation information is classification and referral information of a department, the current description information may be information such as disease information and symptom information which the user desires to diagnose this time, and when the recommendation information is dining recommendation information, the current description information may be information such as taste of dishes and diet information which the user desires to taste.
After the current description information input by the user is obtained, the history description information of the user can be obtained simultaneously or sequentially, the history description information is description information existing relative to the description information of the user at this time, and when the recommendation information is classification and diagnosis guide information of departments, the history description information can allocate result conditions, symptom conditions, case information, diagnosis results and the like to the history departments of the user, it should be understood that the history description information is generally description information stored in the local of the execution main body, the execution main body collects and stores the information of the user for subsequent use, the same history description information can be input by the user or stored in the local of the human-computer interaction device used by the user, the execution main body can obtain the corresponding history description information from the human-computer interaction device in a communication interaction manner, this is not limited in this application.
Step 202, inputting the history description information into a history feature matching model for processing to obtain history feature information.
In this embodiment, after obtaining history description information of a user, the history description information is input into a history feature matching model for processing, the history feature matching model is generally a model for collecting and storing histories of a large number of different users, and after inputting all or part of the history description information of the user, the history feature matching model can obtain more comprehensive history description information of the user, perform corresponding connection and compression, generate history feature information containing as much history information of the user as possible, so that the history feature information is subsequently used to represent the history description information and history recommendation information of the user, and complete behavior analysis of the user based on history data.
It should be understood that, in a general application scenario, the history description information and the history recommendation information generation result of the user are both stored in the execution subject, and the execution subject may generate a history database correspondingly to record the contents so as to generate the history feature matching model.
Step 203, inquiring the current characteristic information corresponding to the current description information.
In this embodiment, after obtaining the current description information of the user, corresponding feature extraction may be performed on the current description information by using a feature extraction method, a feature recognition neural network, and the like, so as to obtain current feature information representing a behavior that the user desires to recommend this time, and taking department classification and diagnosis guidance application scenarios as examples, the current description information of the user may be information such as specific symptom information that the user desires to diagnose and solve this time, department information that the user subjectively desires to go to, and special requirement information that the user selects for a department.
And 204, splicing the historical characteristic information and the current characteristic information to generate recommended characteristic information.
In this embodiment, after the historical feature information and the current feature information are obtained in step 202 and step 203, respectively, an appropriate splicing form may be selected in combination with the expression form of the feature information to splice the historical feature information and the current feature information, for example, when the expression forms of the historical feature information and the current feature information are both information sets, a quantity set may be merged to obtain a recommended feature information set, or when the expression forms of the historical feature information and the current feature information are both vector forms, overlay splicing may be performed based on the dimension quantity information of the historical feature information and the current feature information (for example, when the dimension quantity of the historical feature information is a, and the dimension quantity of the current feature information is B, the dimension quantity of the spliced recommended feature information is C + a) or overlay splicing (for example, the dimension quantity of the historical feature information is a, and if the dimension number of the current feature information is also A, overlapping the contents of the corresponding dimensions of the historical feature information and the current feature information) to realize splicing of the historical feature information and the current feature information so as to obtain the recommended feature information.
The method includes the steps that different constraint rules can be predetermined according to different forms of historical feature information and current feature information, so that the form of obtained recommendation information is adjusted, for example, when the expression forms of the historical feature information and the current feature information are vector forms, the form of the recommendation information can be predetermined to be feature vectors with preset dimensionality, at the moment, after the splicing result of the historical feature information and the current feature information is obtained, the splicing result is processed in a compression or amplification mode, and the feature vectors with the expression forms being the preset dimensionality are obtained and serve as the recommendation information.
Step 205, identifying the recommendation characteristic information by using a predetermined recommendation characteristic identification neural network to obtain first recommendation information, and sending the first recommendation information to the user.
In this embodiment, the predetermined recommendation feature recognition neural network is a recommendation feature neural network determined in advance according to the form of the obtained recommendation feature information, and the recommendation feature information in the history data may be used as an input in advance, and the obtained first recommendation information is used as an output to train the recommendation feature neural network, so that after the recommendation feature information is obtained in step 204, the recommendation feature recognition neural network may be used to recognize the recommendation feature information, and obtain final recommendation information.
And after the first recommendation information is obtained, sending the first recommendation information to the user to complete the generation work of the recommendation information and provide the recommendation information for the user.
According to the method for generating the recommendation information, after the fact that the user does not access the user for the first time is responded, the current description information input by the user and the historical description information of the user are obtained, the historical description information is input into a historical feature matching model to be processed, historical feature information is obtained, the current feature information corresponding to the current description information is inquired, the historical feature information and the current feature information are spliced to generate the recommendation feature information, then the recommendation feature information is identified through a predetermined recommendation feature identification neural network, the first recommendation information is obtained, the first recommendation information is sent to the user, the recommendation information is determined according to the historical description information of the user, the recommendation information is accurately provided for the user, and the quality of the recommendation information is improved.
In some optional implementations of this embodiment, the method further includes: responding to the user as a first access user, and acquiring current description information input by the user; inquiring current characteristic information corresponding to the current description information; and identifying the current characteristic information by adopting the recommended characteristic identification neural network to obtain second recommended information, and sending the second recommended information to the user.
Specifically, when it is determined in step 201 that the user is the first access user, since there is no history description information of the user, then the current description input by the user is directly obtained, and the method in step 203 is also adopted to obtain the current feature information corresponding to the user, because the recommended feature information is obtained by splicing the current feature information and the historical feature information, if there is no history feature information, the content in the current feature information is the same as that of the recommended feature information, and the recommended feature is identified by using the recommended feature identification neural network in step 205, to obtain the second recommendation information and send the second recommendation information to the user to complete the information pushing work, so as to ensure that when the history description information does not exist, the recommendation method and the recommendation device have the advantages that the recommendation work when the user accesses the user for the first time is achieved, and the problem that the recommendation information cannot be generated due to lack of historical data is avoided.
In some optional implementations of the present embodiment, referring to fig. 3, a flow 300 for processing history description information by using a history feature matching model to obtain history feature information is shown, which specifically includes:
step 301, obtaining user description information and recommendation result description information in the history description information of the user.
Specifically, the history description information of the user usually has a large amount of information content, and therefore, for convenience of identification, the information may be divided into two types, that is, user description information and recommendation result description information, where the recommendation information is an application scene of department classification and diagnosis guidance, the user description information is information such as gender, age, last visit day, last visit department of the user, department to which a doctor who has last asked a doctor belongs, evaluation of the doctor who has last asked a doctor, and prescription of purchasing a medicine, and the recommendation result description information may be case description information, symptom information, diagnosis information of the user last time, and prescription information of the medicine.
Step 302, normalizing the user description information, and inputting the normalized user description information to a first deep learning neural network layer forming the historical feature matching model for processing to generate first feature information.
Specifically, since the user description information has less content of entities with clear descriptions and recognizable standards, the user description information obtained in step 301 is normalized in a processing manner such as maximum and minimum normalization and semantic normalization to obtain a standard processing result that can be recognized and extracted by a first Deep learning Neural network (Deep learning Neural network, DNN for short), and then the result is input into the first Deep learning Neural network forming a historical feature matching model, and the Deep learning Neural network (Deep learning Neural network, DNN for short) can recognize and extract the processing result to obtain first feature information used for representing the user description information.
Illustratively, based on the sex, age, number of days visited last time, characteristics of a department visited last time (binary code, 4-bit representation), a department to which a doctor in the last inquiry belongs (binary code, 4-bit representation), evaluation of the doctor in the last inquiry, prescription of purchasing medicine or not, and the like of a patient, the maximum and minimum normalization is performed, and the obtained normalization processing result has 15 dimensions in total, the DNN layer (including two hidden layers: a first hidden layer comprising 60 neurons, and using a Linear rectification activation function (Recui Unit), a second hidden layer comprising 30 neurons, and the activation function being Relu, and the number of dimensions of the output layer being 5) is input, and the output layer outputs first characteristic information with the number of dimensions being 5.
And step 303, inputting the recommendation result description information to a multilayer bidirectional converter encoder layer forming the historical feature matching model to generate second feature information.
Specifically, because the content described in the recommendation result description information is usually standard and accurate result content, the recommendation result description information is input to a multilayer Bidirectional transformer Encoder layer (Bert) that constitutes the historical feature matching model and is processed to obtain second feature information for representing the recommendation result description information.
Illustratively, the description of the latest case of the user, symptoms and medicine details are spliced to be used as the input of a Bert layer, so as to learn the characteristics of the last case of the user, and the user has a certain correlation in the state of illness in a short time according to the analysis of actual conditions. And the final output dimension number of the Bert layers is 10 second characteristic information.
And step 304, inputting the first characteristic information and the second characteristic information into a second deep learning network layer forming the historical characteristic matching model for splicing to obtain the historical characteristic information.
Specifically, the first feature information and the second feature information obtained in the above steps 302 and 303 are input into the second deep learning network layer to be spliced, so as to obtain final historical feature information, where the historical feature information includes related content of the first feature information used for representing user description information and the second feature information used for representing recommendation result description information.
Illustratively, output of the DNN and output of the Bert model are spliced, namely, the first characteristic information and the second characteristic information are spliced, on one hand, joint training learning is carried out, and combined characteristics of the two models are abstractly extracted; on the other hand, the two parts of characteristics on the upper side are compressed, and historical characteristic information with dimension number of 5 is output, so that the purposes of reducing characteristic dimensions and improving running performance are achieved. The partial DNN model has three hidden layers: a first hidden layer of 200 neurons, this layer using the Relu activation function; a second hidden layer of 100 neurons, again using the Relu activation function; the third hidden layer, 60 neurons, also using the Relu activation function; outputting the historical characteristic information with the dimension number of 5 as the historical characteristic information of the user.
In the implementation mode, the historical characteristic information of the historical description information of the user is determined through the historical characteristic matching model based on the user description information and the recommendation result information, so that the condition of the user can be reflected more comprehensively by adopting the historical characteristic information subsequently, and the recommendation information is determined by combining the historical data of the user, so that the problems that in the prior art, because the input information of the user is too little, high-quality recommendation information cannot be generated, and the content is accurately recommended to the user are solved.
In some optional implementation manners of this embodiment, querying the current feature information corresponding to the current description information includes: and inputting the current description information into a multi-layer bidirectional converter encoder layer for outputting the current feature information, and inquiring the current feature information corresponding to the current description information.
Specifically, after the current description information input by the user is acquired, the recommendation result description information is input into another Bert for outputting the current feature information, and the current feature information corresponding to the current description information is obtained and processed, so that the current description information is processed into the current feature information in an expression form similar to the historical feature information, and the historical feature information and the current feature information are spliced to generate the recommendation feature information.
In some optional implementation manners of this embodiment, the obtaining of the historical feature information and the current feature information, and the generating of the recommended feature information by splicing includes: respectively acquiring historical characteristic vectors and current characteristic vectors corresponding to the historical characteristic information and the current characteristic information; inputting the historical feature vector and the current feature vector into a splicing deep learning neural network layer for splicing, and splicing to generate a recommended feature vector as recommended feature information; the dimension number of the recommended feature vector is obtained based on the sum of the dimension number of the historical feature vector and the dimension number of the current feature vector.
Specifically, expression forms with vectors as historical feature information and current feature information can be determined, then historical feature vectors and current feature vectors corresponding to the historical feature information and the current feature information are respectively obtained, and then the historical feature vectors and the current feature vectors are directly spliced, namely the dimension number of the obtained recommended feature vectors serving as the recommended feature information is the sum of the dimension number of the historical feature vectors and the dimension number of the current feature vectors.
Illustratively, the obtained first feature information represented by a 5-dimensional vector and the obtained second feature information represented by a 10-dimensional vector are spliced by using a DNN layer to obtain feature information represented by a 15-dimensional vector.
In some optional implementation manners of this embodiment, identifying the recommendation feature information by using a predetermined recommendation feature identification neural network to obtain first recommendation information, and sending the first recommendation information to the user includes: adopting a predetermined recommendation characteristic recognition neural network to recognize the recommendation characteristic information, and iteratively generating a plurality of recommendation information; wherein the loss function of the recommended feature recognition neural network is a cross entropy function; and stopping iterative identification after the loss function is determined to be converged, determining the finally obtained recommendation information as the first recommendation information, and sending the first recommendation information to the user.
Specifically, after the recommendation feature information is obtained, iteration identification is performed on the recommendation feature by adopting a predetermined recommendation feature identification network which takes a cross entropy function as a loss function in an iteration identification mode, after the loss function is determined to be converged, the iteration identification is stopped, the finally obtained recommendation information is used as first recommendation information and is sent to a user, and the quality of the first recommendation information output by the recommendation feature identification network is improved in the iteration identification mode.
In order to deepen understanding, the application also provides a specific implementation scheme by combining a specific application scene. In this specific application scenario, the current description information "continuously belly from yesternight to the morning" input by the user u1 who does not access for the first time through the human-computer interaction device used by the user u1 is sent to the execution main body of the method for generating the recommendation information (referred to as the recommendation execution main body for short).
The recommending subject determines that the user u1 is a non-first-time-visit user, obtains the current description information "continuously pulling the belly from yesternight to the morning" sent by the user u1 and locally obtains the historical description information "sex male has visited the medical department once, the last symptom description information is continuous diarrhea and stomach pain, and the diagnosis result is superficial gastritis and the prescription of stomach medicine".
Respectively inputting user description information 'sex male, having visited a department in medicine once' and recommendation result description information 'symptom description information coming last time in the historical description information is continuous diarrhea and gastrodynia, and the diagnosis result is superficial gastritis and a prescription' into a first deep learning neural network layer forming a historical characteristic matching model and a multilayer bidirectional transducer encoder layer forming the historical characteristic matching model to obtain first characteristic information and second characteristic information.
And inputting the first characteristic information and the second characteristic information into a second deep learning network layer forming the historical characteristic matching model for splicing to obtain a historical characteristic vector with the dimension number of 5 as historical characteristic information.
Inputting the current description information 'pulling the belly continuously from yesternight to the morning' into a multilayer bidirectional converter encoder layer for outputting current feature information, and inquiring the current feature information which is represented by the current feature vector with the dimension number of 10 and corresponds to the current description information.
And then splicing the historical characteristic information and the current characteristic information to generate recommended characteristic information represented by recommended characteristic vectors with the dimension number of 10, identifying the recommended characteristic information by adopting a predetermined recommended characteristic identification neural network to obtain first recommended information, and sending the first recommended information to the user.
According to the method for generating the recommendation information, the current description information input by the user and the historical description information of the user are input into the historical feature matching model to be processed to obtain the historical feature information, the current feature information corresponding to the current description information is inquired, the historical feature information and the current feature information are spliced to generate the recommendation feature information, then the recommendation feature information is identified by adopting the predetermined recommendation feature identification neural network to obtain the first recommendation information, the first recommendation information is sent to the user, the recommendation information is determined by combining the historical description information of the user, the recommendation information is more accurately provided for the user, and the quality of the recommendation information is improved.
As shown in fig. 4, the apparatus 400 for generating recommendation information of the present embodiment may include: a description information obtaining unit 401 configured to obtain current description information input by a user and history description information of the user in response to determining that the user does not access the user for the first time; a history feature matching unit 402 configured to input the history description information into a history feature matching model for processing, so as to obtain history feature information; a current feature query unit 403 configured to query current feature information corresponding to the current description information; a recommended feature generating unit 404 configured to splice the historical feature information and the current feature information to generate recommended feature information; a recommendation information sending unit 405 configured to identify the recommendation feature information by using a predetermined recommendation feature identification neural network, obtain first recommendation information, and send the first recommendation information to the user.
In some optional implementation manners of this embodiment, the apparatus for generating recommendation information further includes: the description information obtaining unit 401 is further configured to, in response to the user being a first access user, obtain that the current description information input by the user includes; the recommended feature generating unit 404 is further configured to query current feature information corresponding to the current description information; the recommendation information sending unit 405 is further configured to identify the current feature information using the recommended feature recognition neural network, obtain second recommendation information, and send the second recommendation information to the user.
In some optional implementations of this embodiment, the history feature matching unit includes: a history information obtaining subunit configured to obtain user description information and recommendation result description information in the history description information of the user; the first characteristic information generating subunit is configured to normalize the user description information, input the normalized user description information to a first deep learning neural network layer forming the historical characteristic matching model, and process the normalized user description information to generate first characteristic information; a second feature information generation subunit configured to input the recommendation result description information to a multilayer bidirectional converter encoder layer constituting the historical feature matching model, and generate second feature information; and the historical characteristic information generating subunit is configured to input the first characteristic information and the second characteristic information into a second deep learning network layer forming the historical characteristic matching model for splicing to obtain the historical characteristic information.
In some optional implementations of this embodiment, the push information sending unit is further configured to: sorting the medical state entities by adopting a probability graph model, and selecting a preset number of medical state entities according to a sorting result to generate a push information set; and sending the push set to the user.
In some optional implementations of this embodiment, the current feature query unit 403 is further configured to: and inputting the current description information into a multi-layer bidirectional converter encoder layer for outputting the current feature information, and inquiring the current feature information corresponding to the current description information.
In some optional implementations of this embodiment, the recommended feature generating unit 404 includes: a feature vector generation subunit configured to obtain a history feature vector and a current feature vector corresponding to the history feature information and the current feature information, respectively; the recommended vector splicing subunit is configured to input the historical feature vector and the current feature vector to a splicing deep learning neural network layer for splicing, and generate a recommended feature vector as recommended feature information through splicing; the dimension number of the recommended feature vector is obtained based on the sum of the dimension number of the historical feature vector and the dimension number of the current feature vector.
In some optional implementations of this embodiment, the recommendation information sending unit 405 includes: the recommendation information generation subunit is configured to identify the recommendation characteristic information by adopting a predetermined recommendation characteristic identification neural network, and iteratively generate a plurality of recommendation information; wherein the loss function of the recommended feature recognition neural network is a cross entropy function; a recommendation information determination subunit configured to stop iterative identification in response to determining that the loss function converges, and determine finally obtained recommendation information as the first recommendation information; and the recommendation information sending subunit is configured to send the first recommendation information to the user.
The present embodiment exists as an apparatus embodiment corresponding to the above method embodiment, and the same contents refer to the description of the above method embodiment, which is not repeated herein. According to the device for generating the recommendation information, the recommendation information is determined by combining the historical description information of the user, the recommendation information is provided for the user more accurately, and the quality of the recommendation information is improved.
As shown in fig. 5, the electronic device is a block diagram of an electronic device according to an embodiment of the present disclosure. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be examples only, and are not meant to limit implementations of the present application that are described and/or claimed herein.
As shown in fig. 5, the electronic apparatus includes: one or more processors 501, memory 502, and interfaces for connecting the various components, including high-speed interfaces and low-speed interfaces. The various components are interconnected using different buses and may be mounted on a common motherboard or in other manners as desired. The processor may process instructions for execution within the electronic device, including instructions stored in or on the memory to display graphical information of a GUI on an external input/output apparatus (such as a display device coupled to the interface). In other embodiments, multiple processors and/or multiple buses may be used, along with multiple memories and multiple memories, as desired. Also, multiple electronic devices may be connected, with each device providing portions of the necessary operations (e.g., as a server array, a group of blade servers, or a multi-processor system). In fig. 5, one processor 501 is taken as an example.
Memory 502 is a non-transitory computer readable storage medium as provided herein. The memory stores instructions executable by at least one processor to cause the at least one processor to perform the method for generating recommendation information provided herein. The non-transitory computer readable storage medium of the present application stores computer instructions for causing a computer to perform the method of generating recommendation information provided herein.
The memory 502, which is a non-transitory computer-readable storage medium, may be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as program instructions/modules corresponding to the method of generating recommendation information in the embodiments of the present application (for example, the description information acquiring unit 401, the historical feature matching unit 402, the current feature querying unit 403, the recommendation feature generating unit 404, and the recommendation information transmitting unit 405 shown in fig. 4). The processor 501 executes various functional applications of the server and data processing by running non-transitory software programs, instructions, and modules stored in the memory 502, that is, implements the method of generating recommendation information in the above method embodiments.
The memory 502 may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store data created according to use of the electronic device that generates the recommendation information, and the like. Further, the memory 502 may include high speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, memory 502 may optionally include memory located remotely from processor 501, which may be connected to an electronic device that generates recommendation information via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The electronic device for performing the method of generating recommendation information may further include: an input device 503 and an output device 504. The processor 501, the memory 502, the input device 503 and the output device 504 may be connected by a bus or other means, and fig. 5 illustrates the connection by a bus as an example.
The input device 503 may receive input numeric or character information and generate key signal inputs related to user settings and function control of the electronic apparatus generating the recommendation information, such as an input device such as a touch screen, a keypad, a mouse, a track pad, a touch pad, a pointer, one or more mouse buttons, a track ball, a joystick, or the like. The output devices 504 may include a display device, auxiliary lighting devices (e.g., LEDs), and haptic feedback devices (e.g., vibrating motors), among others. The display device may include, but is not limited to, a Liquid Crystal Display (LCD), a Light Emitting Diode (LED) display, and a plasma display. In some implementations, the display device can be a touch screen.
Various implementations of the systems and techniques described here can be realized in digital electronic circuitry, integrated circuitry, application specific ASICs (application specific integrated circuits), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs that are executable and/or interpretable on a programmable system including at least one programmable processor, which may be special or general purpose, receiving data and instructions from, and transmitting data and instructions to, a storage system, at least one input device, and at least one output device.
These computer programs (also known as programs, software applications, or code) include machine instructions for a programmable processor, and may be implemented using high-level procedural and/or object-oriented programming languages, and/or assembly/machine languages. As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, apparatus, and/or device (e.g., magnetic discs, optical disks, memory, Programmable Logic Devices (PLDs)) used to provide machine instructions and/or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term "machine-readable signal" refers to any signal used to provide machine instructions and/or data to a programmable processor.
To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic, speech, or tactile input.
The systems and techniques described here can be implemented in a computing system that includes a back-end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), Wide Area Networks (WANs), and the Internet.
The computer system may include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other.
According to the technical scheme of the embodiment of the application, after the user is determined not to access the user for the first time, the current description information input by the user and the historical description information of the user are obtained, the historical description information is input into a historical feature matching model to be processed to obtain historical feature information, the current feature information corresponding to the current description information is inquired, the historical feature information and the current feature information are spliced to generate recommended feature information, then the recommended feature information is identified by adopting a predetermined recommended feature identification neural network to obtain first recommended information, the first recommended information is sent to the user, the recommended information is determined by combining the historical description information of the user, the recommended information is more accurately provided for the user, and the quality of the recommended information is improved.
It should be understood that various forms of the flows shown above may be used, with steps reordered, added, or deleted. For example, the steps described in the present application may be executed in parallel, sequentially, or in different orders, as long as the desired results of the technical solutions disclosed in the present application can be achieved, and the present invention is not limited herein.
The above-described embodiments should not be construed as limiting the scope of the present application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may be made in accordance with design requirements and other factors. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims (14)

1. A method of generating recommendation information, comprising:
in response to determining that a user does not access the user for the first time, acquiring current description information input by the user and historical description information of the user;
inputting the historical description information into a historical feature matching model for processing to obtain historical feature information;
inquiring current characteristic information corresponding to the current description information;
splicing the historical characteristic information and the current characteristic information to generate recommended characteristic information;
and identifying the recommendation characteristic information by adopting a predetermined recommendation characteristic identification neural network to obtain first recommendation information, and sending the first recommendation information to the user.
2. The method of claim 1, further comprising:
responding to the user as a first-time access user, and acquiring current description information input by the user;
inquiring current characteristic information corresponding to the current description information;
and identifying the current characteristic information by adopting the recommended characteristic identification neural network to obtain second recommended information, and sending the second recommended information to the user.
3. The method of claim 1, wherein inputting the historical description information into a historical feature matching model for processing to obtain historical feature information comprises:
acquiring user description information and recommendation result description information in the historical description information of the user;
after the user description information is normalized, inputting the user description information into a first deep learning neural network layer forming the historical feature matching model for processing to generate first feature information;
inputting the recommended result description information to a multilayer bidirectional converter encoder layer forming the historical feature matching model to generate second feature information;
and inputting the first characteristic information and the second characteristic information into a second deep learning network layer forming the historical characteristic matching model for splicing to obtain the historical characteristic information.
4. The method of claim 1, wherein the querying the current feature information corresponding to the current description information comprises:
and inputting the current description information into a multi-layer bidirectional converter encoder layer for outputting the current feature information, and inquiring the current feature information corresponding to the current description information.
5. The method of claim 1, wherein the obtaining the historical feature information and the current feature information and the concatenating to generate the recommended feature information comprises:
respectively acquiring historical feature vectors and current feature vectors corresponding to the historical feature information and the current feature information;
inputting the historical feature vector and the current feature vector into a splicing deep learning neural network layer for splicing, and splicing to generate a recommended feature vector as recommended feature information; and the dimension number of the recommended feature vector is obtained based on the sum of the dimension number of the historical feature vector and the dimension number of the current feature vector.
6. The method of claim 1, wherein the identifying the recommendation feature information using a predetermined recommendation feature identification neural network, obtaining first recommendation information, and sending the first recommendation information to the user comprises:
adopting a predetermined recommendation characteristic recognition neural network to recognize the recommendation characteristic information, and iteratively generating a plurality of recommendation information; wherein the loss function of the recommended feature recognition neural network is a cross entropy function;
and stopping iterative identification after the loss function is determined to be converged, determining the finally obtained recommendation information as the first recommendation information, and sending the first recommendation information to the user.
7. An apparatus to generate recommendation information, comprising:
a description information acquisition unit configured to acquire current description information input by a user and historical description information of the user in response to a determination that the user is not a first-time access user;
the history feature matching unit is configured to input the history description information into a history feature matching model for processing to obtain history feature information;
a current feature query unit configured to query current feature information corresponding to the current description information;
the recommendation characteristic generation unit is configured to splice the historical characteristic information and the current characteristic information to generate recommendation characteristic information;
and the recommendation information sending unit is configured to identify the recommendation characteristic information by adopting a predetermined recommendation characteristic identification neural network, obtain first recommendation information and send the first recommendation information to the user.
8. The apparatus of claim 7, further comprising:
the description information acquisition unit is further configured to, in response to the user being a first access user, acquire current description information input by the user including;
the recommended feature generation unit is further configured to query current feature information corresponding to the current description information;
the recommendation information sending unit is further configured to identify the current feature information by using the recommendation feature identification neural network, obtain second recommendation information, and send the second recommendation information to the user.
9. The apparatus of claim 7, wherein the historical feature matching unit comprises:
a history information obtaining subunit configured to obtain user description information and recommendation result description information in history description information of the user;
the first characteristic information generation subunit is configured to normalize the user description information, input the normalized user description information to a first deep learning neural network layer forming the historical characteristic matching model, and process the normalized user description information to generate first characteristic information;
a second feature information generation subunit configured to input the recommendation result description information to a multilayer bidirectional converter encoder layer constituting the historical feature matching model, and generate second feature information;
a historical feature information generating subunit configured to input the first feature information and the second feature information to a second deep learning network layer constituting the historical feature matching model for stitching, so as to obtain the historical feature information.
10. The apparatus of claim 7, wherein the current feature querying unit is further configured to:
and inputting the current description information into a multi-layer bidirectional converter encoder layer for outputting the current feature information, and inquiring the current feature information corresponding to the current description information.
11. The apparatus of claim 7, wherein the recommendation feature generation unit comprises:
a feature vector generation subunit configured to acquire a history feature vector and a current feature vector corresponding to the history feature information and the current feature information, respectively;
a recommended vector splicing subunit, configured to input the historical feature vector and the current feature vector to a splicing deep learning neural network layer for splicing, and generate a recommended feature vector as recommended feature information by splicing; and the dimension number of the recommended feature vector is obtained based on the sum of the dimension number of the historical feature vector and the dimension number of the current feature vector.
12. The apparatus of claim 7, wherein the recommendation information transmitting unit comprises:
the recommendation information generation subunit is configured to identify the recommendation characteristic information by adopting a predetermined recommendation characteristic identification neural network, and iteratively generate a plurality of recommendation information; wherein the loss function of the recommended feature recognition neural network is a cross entropy function;
a recommendation information determination subunit configured to stop iterative identification in response to determining that the loss function converges, and determine finally obtained recommendation information as the first recommendation information;
and the recommendation information sending subunit is configured to send the first recommendation information to the user.
13. An electronic device, comprising:
at least one processor; and
a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of generating recommendation information of any of claims 1-6.
14. A non-transitory computer readable storage medium storing computer instructions, comprising: the computer instructions are for causing the computer to perform the method of generating recommendation information of any of claims 1-6.
CN202011239590.5A 2020-11-09 2020-11-09 Method and device for generating recommendation information Active CN112287232B (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN202011239590.5A CN112287232B (en) 2020-11-09 2020-11-09 Method and device for generating recommendation information
PCT/CN2021/128399 WO2022095893A1 (en) 2020-11-09 2021-11-03 Method and apparatus for generating recommendation information

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011239590.5A CN112287232B (en) 2020-11-09 2020-11-09 Method and device for generating recommendation information

Publications (2)

Publication Number Publication Date
CN112287232A true CN112287232A (en) 2021-01-29
CN112287232B CN112287232B (en) 2024-04-16

Family

ID=74351793

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011239590.5A Active CN112287232B (en) 2020-11-09 2020-11-09 Method and device for generating recommendation information

Country Status (2)

Country Link
CN (1) CN112287232B (en)
WO (1) WO2022095893A1 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113821731A (en) * 2021-11-23 2021-12-21 湖北亿咖通科技有限公司 Information push method, device and medium
WO2022095893A1 (en) * 2020-11-09 2022-05-12 北京京东拓先科技有限公司 Method and apparatus for generating recommendation information

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115274086B (en) * 2022-09-27 2023-01-13 无码科技（杭州）有限公司 Intelligent diagnosis guiding method and system
CN115687674A (en) * 2022-12-20 2023-02-03 昆明勤砖晟信息科技有限公司 Big data demand analysis method and system serving smart cloud service platform
CN116450925B (en) * 2022-12-27 2023-12-15 深圳市网新新思软件有限公司 User relationship analysis method and system based on artificial intelligence
CN116629810B (en) * 2023-07-24 2023-10-20 中建五局第三建设有限公司 Operation recommendation method, device, equipment and medium based on building office system

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108733703A (en) * 2017-04-20 2018-11-02 北京京东尚科信息技术有限公司 The answer prediction technique and device of question answering system, electronic equipment, storage medium
CN108766545A (en) * 2018-05-31 2018-11-06 平安医疗科技有限公司 Online interrogation section office distribution method, device, computer equipment and storage medium
US20180365828A1 (en) * 2017-06-19 2018-12-20 Viz.ai, Inc. Method and system for computer-aided triage
CN111523976A (en) * 2020-04-23 2020-08-11 京东数字科技控股有限公司 Commodity recommendation method and device, electronic equipment and storage medium
CN111782774A (en) * 2019-04-03 2020-10-16 北京嘀嘀无限科技发展有限公司 Question recommendation method and device

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112287232B (en) * 2020-11-09 2024-04-16 北京京东拓先科技有限公司 Method and device for generating recommendation information

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108733703A (en) * 2017-04-20 2018-11-02 北京京东尚科信息技术有限公司 The answer prediction technique and device of question answering system, electronic equipment, storage medium
US20180365828A1 (en) * 2017-06-19 2018-12-20 Viz.ai, Inc. Method and system for computer-aided triage
CN108766545A (en) * 2018-05-31 2018-11-06 平安医疗科技有限公司 Online interrogation section office distribution method, device, computer equipment and storage medium
CN111782774A (en) * 2019-04-03 2020-10-16 北京嘀嘀无限科技发展有限公司 Question recommendation method and device
CN111523976A (en) * 2020-04-23 2020-08-11 京东数字科技控股有限公司 Commodity recommendation method and device, electronic equipment and storage medium

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2022095893A1 (en) * 2020-11-09 2022-05-12 北京京东拓先科技有限公司 Method and apparatus for generating recommendation information
CN113821731A (en) * 2021-11-23 2021-12-21 湖北亿咖通科技有限公司 Information push method, device and medium

Also Published As

Publication number Publication date
WO2022095893A1 (en) 2022-05-12
CN112287232B (en) 2024-04-16

Similar Documents

Publication Publication Date Title
CN112287232A (en) Method and device for generating recommendation information
US11226673B2 (en) Affective interaction systems, devices, and methods based on affective computing user interface
US20230351102A1 (en) Machine content generation
CN111710412B (en) Diagnostic result verification method and device and electronic equipment
CA3137096A1 (en) Computer-implemented natural language understanding of medical reports
Altuwairqi et al. Student behavior analysis to measure engagement levels in online learning environments
US20180314960A1 (en) Utilizing artificial intelligence for data extraction
CN112530576A (en) Online doctor-patient matching method and device, electronic equipment and storage medium
US11847411B2 (en) Obtaining supported decision trees from text for medical health applications
CN109273101A (en) Man-machine interaction method, device, computer equipment and medium based on decision tree
CN112509690B (en) Method, apparatus, device and storage medium for controlling quality
CN111326251A (en) Method and device for outputting inquiry questions and electronic equipment
US11947578B2 (en) Method for retrieving multi-turn dialogue, storage medium, and electronic device
CN112287121A (en) Push information generation method and device
CN111967599B (en) Method, apparatus, electronic device and readable storage medium for training model
Khilji et al. Healfavor: Dataset and a prototype system for healthcare chatbot
US20240143605A1 (en) System And Method For Improved State Identification And Prediction In Computerized Queries
Schueller et al. Automated pattern analysis in gesture research: Similarity measuring in 3D motion capture models of communicative action
Noor et al. Depression Detection In Social Media Using Bagging Classifier
Zhu et al. Emotion recognition based on brain-like multimodal hierarchical perception
CN113160914A (en) Online inquiry method and device, electronic equipment and storage medium
CN116894498A (en) Training method, strategy identification method, device and equipment of network model
Mehta et al. Artificial intelligence powered chatbot for mental healthcare based on sentiment analysis
CN114242267A (en) Neural network-based inquiry reply method, device, equipment and storage medium
CN113836284A (en) Method and device for constructing knowledge base and generating response statement

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant