CN110417635A - Automatic information reply method, device and equipment - Google Patents
Automatic information reply method, device and equipment Download PDFInfo
- Publication number
- CN110417635A CN110417635A CN201810404909.1A CN201810404909A CN110417635A CN 110417635 A CN110417635 A CN 110417635A CN 201810404909 A CN201810404909 A CN 201810404909A CN 110417635 A CN110417635 A CN 110417635A
- Authority
- CN
- China
- Prior art keywords
- title
- information
- communication object
- reply
- application program
- 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
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L51/00—User-to-user messaging in packet-switching networks, transmitted according to store-and-forward or real-time protocols, e.g. e-mail
- H04L51/02—User-to-user messaging in packet-switching networks, transmitted according to store-and-forward or real-time protocols, e.g. e-mail using automatic reactions or user delegation, e.g. automatic replies or chatbot-generated messages
Abstract
The invention discloses an automatic information reply method, device and equipment, wherein the method comprises the following steps: analyzing context information of a current user and a communication object through an application program according to a natural language processing model, and automatically generating one or more reply contents; analyzing and identifying a first name of a communication object according to the name corpus of the context information, and reading a second name of the communication object according to the application setting information of the application program; comparing whether the first name is the same as the second name or not, if the first name is different from the second name, respectively inserting the first name and the second name into the calling position of each reply content, and generating a plurality of candidate reply messages; and acquiring target reply information selected by the current user from the candidate reply information, and sending the target reply information to the communication object through the application program. Therefore, the call in the quick reply content is more in line with the user expectation, and the accuracy and the individuation degree of quick reply are improved.
Description
Technical field
The present invention relates to technical field of information processing more particularly to a kind of information automatic reply methods, device and equipment.
Background technique
The function that quickly revert is added in input method can effectively save the input cost of user, improve user and link up effect
Rate.
In the related technology, quickly revert model is trained according to the corpus data of great amount of samples user, when
When reply content is related to address or nickname information to communication object, the address or the pet name obtained by quickly revert model is believed
Breath does not often have specific aim, such as the word general using wider range such as " babe ", is not able to satisfy the individual demand of user.
Summary of the invention
The present invention is directed to solve at least some of the technical problems in related technologies.
For this purpose, the first purpose of this invention is to propose a kind of information automatic reply method, contact person is utilized to realize
Data and context semantic information obtain the salutation information to communication object and are added in quickly revert sentence, so that fast
Address in reply more meets user's expectation, improves the accuracy of quickly revert, makes fast replying function more intelligent and individual character
Change.
Second object of the present invention is to propose that a kind of information automatically replies device.
Third object of the present invention is to propose a kind of terminal device.
Fourth object of the present invention is to propose a kind of computer program product.
5th purpose of the invention is to propose a kind of non-transitorycomputer readable storage medium.
In order to achieve the above object, first aspect present invention embodiment proposes a kind of information automatic reply method, comprising:
The contextual information that active user passes through application program and communication object is analyzed according to Natural Language Processing Models, from
It is dynamic to generate one or more reply contents；
The first title of the communication object is identified according to the title Concordance to the contextual information, and according to institute
State the second title that the communication object is read using setting information of application program；
Compare first title and whether second title is identical, knows first title and described if comparing
Two titles are not identical, then first title and second title are inserted respectively into the address position of each reply content,
Generate multiple candidate return informations；
The target retro information that the active user selects from the candidate return information is obtained, by the target retro
Information is sent to the communication object by the application program.
In addition, information automatic reply method according to the above embodiment of the present invention can also have following supplementary technology special
Sign:
Optionally, second title that the communication object is read using setting information according to the application program,
Include:
According to the application page setting information of the application program, the communication object is read from contact name control
The second title.
Optionally, second title that the communication object is read using setting information according to the application program,
Include:
According to the application registration setting information of the application program, the communication object is read from register name control
Second title.
Optionally, after whether first title and second title are identical, further includes: if comparing
Know that first title is identical with second title, then first title and second title is carried out at duplicate removal
Reason；
First title or second title be inserted into the address position of each reply content, generate one or
The multiple candidate return informations of person.
Optionally, the method, further includes:
Obtain the parameter letter of the target retro information training network neural model of active user's selection in preset period of time
Breath generates title scoring model.
Optionally, in the address position that first title and second title are inserted respectively into each reply content
It sets, after the multiple candidate return informations of generation, further includes:
According to the title scoring model pair candidate return information corresponding with first title, and with described second
The corresponding candidate return information of title is given a mark；
The low candidate return information of score is deleted according to marking result, provides a user the high candidate return information of score.
Optionally, the method, further includes:
Judge that the active user automatically replies in function whether be provided with corresponding with the communication object the in information
Three titles simultaneously store, wherein the priority of the third title is higher than first title and second title；
If knowing that the active user does not automatically reply in information is arranged the third title in function, according to institute
The title Concordance for stating contextual information identifies the first title of the communication object；
If knowing that the active user automatically replies in information is provided with the third title in function, with the third
Title is inserted into the address position of each reply content, generates one or more candidate return information.
In order to achieve the above object, second aspect of the present invention embodiment, which proposes a kind of information, automatically replies device, comprising:
First generation module, for according to Natural Language Processing Models analyze active user by application program with communicate pair
The contextual information of elephant automatically generates one or more reply contents；
Module is obtained, the first of the communication object is identified to the title Concordance of the contextual information for basis
Title, and according to the second title for reading the communication object using setting information of the application program；
Whether processing module is identical for first title and second title；Second generation module, is used for
Know that first title and second title be not identical if comparing, first title and second title are distinguished
It is inserted into the address position of each reply content, generates multiple candidate return informations；
Sending module, the target retro information selected from the candidate return information for obtaining the active user,
The target retro information is sent to the communication object by the application program.
Optionally, the acquisition module is specifically used for: according to the application page setting information of the application program, from connection
The second title of the communication object is read in people's name control.
Optionally, the module that obtains is specifically used for: according to the application registration setting information of the application program, from registration
The second title of the communication object is read in name control.
Optionally, second generation module is also used to: knowing first title and the second title phase if comparing
Together, then duplicate removal processing is carried out to first title and second title；By first title or second title
It is inserted into the address position of each reply content, generates one or more candidate return information.
Optionally, the device, further includes:
Third generation module, for obtaining the target retro information training network of active user's selection in preset period of time
The parameter information of neural model generates title scoring model.
Optionally, the device, further includes:
Screening module is used for according to the title scoring model pair candidate return information corresponding with first title,
And candidate return information corresponding with second title is given a mark；The low candidate reply of score is deleted according to marking result
Information provides a user the high candidate return information of score.
Optionally, the device, further includes:
4th generation module, for judge the active user information automatically reply in function whether be provided with it is described
The corresponding third title of communication object simultaneously stores, wherein the priority of the third title is higher than first title and described
Second title；
If knowing that the active user does not automatically reply in information is arranged the third title in function, according to institute
The title Concordance for stating contextual information identifies the first title of the communication object；
If knowing that the active user automatically replies in information is provided with the third title in function, with the third
Title is inserted into the address position of each reply content, generates one or more candidate return information.
In order to achieve the above object, third aspect present invention embodiment proposes a kind of terminal device, including processor and storage
Device；Wherein, the processor is run by reading the executable program code stored in the memory can be performed with described
The corresponding program of program code, for realizing the information automatic reply method as described in first aspect embodiment.
In order to achieve the above object, fourth aspect present invention embodiment proposes a kind of computer program product, which is characterized in that
Realize that the information as described in first aspect embodiment is returned automatically when the instruction processing unit in the computer program product executes
Compound method.
In order to achieve the above object, fifth aspect present invention embodiment proposes a kind of non-transitory computer-readable storage medium
Matter is stored thereon with computer program, which is characterized in that realizes when the program is executed by processor such as first aspect embodiment institute
The information automatic reply method stated.
Technical solution provided in an embodiment of the present invention can include the following benefits:
The contextual information that active user passes through application program and communication object is analyzed according to Natural Language Processing Models, from
It is dynamic to generate one or more reply contents, and then according to first of the title Concordance identification communication object to contextual information
Title, and the second title of communication object is read according to application program using setting information, further relatively first title and
Whether the second title is identical, knows that the first title and the second title be not identical if comparing, by the first title and the second title point
It is not inserted into the address position of each reply content, generates multiple candidate return informations, further obtains active user from candidate
Target retro information is sent to communication object by application program by the target retro information selected in return information.Lead to as a result,
It crosses using contact data and context semantic information, obtain the salutation information to communication object and is added to quickly revert sentence
In so that the address in quickly revert more meet user expectation, improve the accuracy of quickly revert, allow fast replying function more
Intelligence and personalization.
The additional aspect of the present invention and advantage will be set forth in part in the description, and will partially become from the following description
Obviously, or practice through the invention is recognized.
Detailed description of the invention
Fig. 1 is a kind of flow diagram of information automatic reply method provided by the embodiment of the present invention；
Fig. 2 a- Fig. 2 b is a kind of schematic diagram of candidate return information provided by the embodiment of the present invention；
Fig. 3 is the flow diagram of another kind information automatic reply method provided by the embodiment of the present invention；
Fig. 4 is the schematic diagram of the candidate return information of another kind provided by the embodiment of the present invention；
Fig. 5 is the flow diagram of another information automatic reply method provided by the embodiment of the present invention；
Fig. 6 automatically replies the structural schematic diagram of device for a kind of information provided by the embodiment of the present invention；
Fig. 7 is the structural schematic diagram that another kind information provided by the embodiment of the present invention automatically replies device；
Fig. 8 is a kind of structural schematic diagram of terminal device provided by the embodiment of the present invention.
Specific embodiment
The embodiment of the present invention is described below in detail, examples of the embodiments are shown in the accompanying drawings, wherein from beginning to end
Same or similar label indicates same or similar element or element with the same or similar functions.Below with reference to attached
The embodiment of figure description is exemplary, it is intended to is used to explain the present invention, and is not considered as limiting the invention.
Below with reference to the accompanying drawings information automatic reply method, device and the equipment of the embodiment of the present invention are described.
Fig. 1 is a kind of flow diagram of information automatic reply method provided by the embodiment of the present invention, as shown in Figure 1,
The information automatic reply method includes:
Step 101, according to Natural Language Processing Models analyze active user by application program and communication object up and down
Literary information automatically generates one or more reply contents.
It is alternatively possible to which the natural language sentence inputted by collecting sample user, forms training data, and then utilize instruction
Practice data to be trained machine learning model, to obtain the Natural Language Processing Models that reply content can be generated.
Further, when active user is communicated by application program with communication object, according to trained natural language
Speech processing model identifies and analyzes the contextual information of active user and communication object, to automatically generate one or more time
Multiple content.For example, after carrying out discriminance analysis to contextual information " Happy birthday " according to Natural Language Processing Models,
The reply contents such as " Thank you ", " Thanks babe ", " Thanks for thinking of me " can be automatically generated.
Step 102, according to the first title of the title Concordance identification communication object to contextual information, and according to answering
With the second title for reading communication object using setting information of program.
It is understood that Natural Language Processing Models in the related technology be with the corpus data of great amount of samples user into
Row training obtains, when being related to the address of communication object in reply content, often without specific aim, such as using
The general word such as " babe " wider range, these are called words usually and are not met user's expectation for a user, therefore, in order to
The salutation information in reply content is set more to meet user's expectation, it, can also be from upper after generating one or more reply contents
The name information of communication object is obtained in the title corpus and contact name control of context information.
As an example, the title corpus in contextual information can be known by natural language processing technique
It not and analyzes, to obtain user to the salutation information of communication object, and the first title as communication object.For example, to upper
" bye mom " in context information is identified and analyzed, and can obtain first title of " mom " as communication object.
As an example, hypertext markup language (Hypertext Markup Language, HTML) 5 can be passed through
Technology reads the contact name of communication object according to the application page setting information of application program from contact name control
Claim information, and the second title as communication object.For example, when application program is SMS, it can be by HTML5 to short message
The contact name control of chat interface is parsed, to read the contact name information of communication object, and as logical
Believe the second title of object.
It as an example, can also be by HTML5 technology, according to the application registration setting information of application program, from note
The register name information of communication object, and the second title as communication object are read in volume name control.For example, application program
When for wechat, if wechat is registered by cell-phone number, the register name control that register interface can also be arranged by HTML5
It is parsed, to read the register name information of communication object, and as the second title of communication object, to realize
The name information for the communication object being arranged in mobile phone is synchronized in wechat.
Step 103, compare the first title and whether the second title is identical, know the first title and the second title not if comparing
It is identical, then the first title and the second title are inserted respectively into the address position of each reply content, generate multiple candidate replies
Information.
In one embodiment of the invention, know that the first title and the second title be not identical if comparing, by first place
Claim the address position for being inserted respectively into each reply content with the second title, generate multiple candidate return informations, to be user
Diversified candidate return information is provided.For example, as shown in Figure 2 a, analyzing contextual information according to Natural Language Processing Models
" Morning beauty " automatically generates reply content " Hey ", " Hi " and " Hello ", and then according to contextual information
Title Concordance identifies first entitled " mom " of communication object, and according to the page setup information of application program from connection
Second entitled " Ellie " that communication object is read in people's name control, knows the first title and the second title not by comparing
Together, the first title and the second title are inserted respectively into the address position of each reply content, generate candidate return information " Hey
Mom ", " Hi mom ", " Hello mom ", " Hey Ellie ", " Hi Ellie " and " Hello Ellie ".
In another embodiment of the present invention, when compare know that the first title is identical as the second title when, in order to avoid
It repeats to show, promotes user experience, duplicate removal processing can also be carried out to the first title and the second title.It in turn, will be after duplicate removal
First title or the second title are inserted into the address position of each reply content, generate one or more candidate and reply letter
Breath.For example, as shown in Figure 2 b, contextual information " Morning beauty " is analyzed according to Natural Language Processing Models, it is automatic raw
At reply content " Hey ", " Hi " and " Hello ", and then according to the title Concordance identification communication pair to contextual information
The first of elephant is entitled " mom ", and reads communication pair from contact name control according to the page setup information of application program
The second of elephant is entitled " mom ", knows that the first title is identical as the second title by comparing, to the first title and the second title into
After row duplicate removal processing, " mom " is inserted into the address position of each reply content, generates candidate return information " Heymom ", " Hi
Mom " and " Hello mom ".
It should be noted that if the title corpus of contextual information, can not identify the first place of communication object
Claim, then the second title of communication object is read using setting information according to application program, and then the second title is inserted into often
The address position of a reply content generates one or more candidate return informations.
Step 104, the target retro information that active user selects from candidate return information is obtained, by target retro information
Communication object is sent to by application program.
Optionally, active user is obtained by modes such as phonetic order or click commands, is selected from candidate return information
Target retro information, and then target retro information is sent to communication object by application program, so that user be made not need
Being manually entered reply content can easily be replied.
In conclusion the information automatic reply method of the embodiment of the present invention, is analyzed current according to Natural Language Processing Models
User passes through the contextual information of application program and communication object, automatically generates one or more reply contents, and then according to right
First title of the title Concordance identification communication object of contextual information, and read according to the application setting information of application program
The second title of communication object is taken, further relatively whether the first title and the second title are identical, know the first title if comparing
It is not identical with the second title, then the first title and the second title are inserted respectively into the address position of each reply content, generated
Multiple candidate's return informations further obtain the target retro information that active user selects from candidate return information, by target
Return information is sent to communication object by application program.As a result, by utilizing contact data and context semantic information, obtain
It takes the salutation information to communication object and is added in quickly revert sentence, so that the address in quickly revert more meets user's phase
It hopes, improves the accuracy of quickly revert, make fast replying function more intelligent and personalized, improve user and input experience.
Based on the above embodiment, title scoring model can also be trained, by title scoring model to candidate return information
It is screened, so that the candidate return information being presented to the user more meets user's expectation, further increases the accurate of quickly revert
Property and personalization level.
Fig. 3 is the flow diagram of another kind information automatic reply method provided by the embodiment of the present invention, such as Fig. 3 institute
Show, after step 103, further includes:
Step 201, according to title scoring model pair candidate return information corresponding with the first title, and with the second title
Corresponding candidate's return information is given a mark.
It is alternatively possible to obtain target retro information that active user in preset period of time selects as training data, to net
The parameter information of network neural model is trained, to generate title scoring model.In turn, pass through trained title scoring model
It gives a mark to candidate return information, the high candidate return information of score is supplied to user.
As an example, the target retro information that active user in preset period of time (such as in one week) can be selected is made
For positive sample, other nonoptional candidate return informations are input to net as negative sample, and then by positive sample and negative sample
Parameter information is trained in network neural model, to generate title scoring model.
In the present embodiment, since title scoring model is that precondition is good, corresponding with the first title getting
Candidate return information will corresponding candidate return information input and after candidate return information corresponding with the second title
Into title scoring model, so that it may obtain the marking result of candidate return information.
Step 202, the low candidate return information of score is deleted according to marking result, provides a user the high candidate of score and returns
Complex information.
It is understood that the score of marking result is higher, illustrate that candidate's return information more meets user's expectation, instead
It, the score for result of giving a mark is lower, illustrates that candidate's return information does not meet user's expectation more.
In one embodiment of the invention, can to each candidate return information according to marking result score by height to
It is low to be ranked up, and then the candidate return information of sequence rearward is deleted, and choose the forward candidate return information that sorts and be supplied to
User.Alternatively, the score for result of giving a mark is compared with preset threshold value, score is deleted lower than the candidate of threshold value and replys letter
Breath, and the candidate return information that score is higher by threshold value is chosen, as the candidate return information provided a user.Further, it examines
Consider interface shape size, the number of the candidate return information provided a user can also be limited.
For the clearer explanation present invention, citing is described in detail below:
As shown in figure 4, first entitled " mom ", second entitled " Ellie ", corresponding candidate's return information are respectively
" Thanks mom " and " Thanks Ellie ", candidate return information is input in title scoring model, knows candidate return
The score of complex information " Thanks mom " is high, and the score of candidate return information " Thanks Ellie " is low, and then deletes " Thanks
Ellie ", and candidate return information " Thanks mom " is supplied to user.
The information automatic reply method of the embodiment of the present invention, the target by obtaining active user's selection in preset period of time are returned
The parameter information of complex information training network neural model, generates title scoring model, and then according to title scoring model to candidate
Return information is given a mark, and deletes the low candidate return information of score, and the high candidate return information of score is supplied to user,
So that the candidate return information being presented to the user more meets user's expectation, the accuracy and personalization of quickly revert are further increased
Degree.
Based on the above embodiment, further, in practical applications, user may wish to destinations traffic object into
When row communication, destinations traffic object is called using the specific pet name of oneself setting always.Fig. 5 is provided by the embodiment of the present invention
Another information automatic reply method flow diagram, as shown in figure 5, before step 102, the information side of automatically replying
Method includes:
Step 301, judge that active user automatically replies in function whether be provided with corresponding with communication object the in information
Three titles simultaneously store, wherein the priority of third title is higher than the first title and the second title.
Wherein, third title is by user's self-setting, with realize user when use information automatically replies function, always with
The specific pet name of oneself setting calls destinations traffic object, and the user experience is improved.
In one embodiment of the invention, the corresponding relationship for being stored with communication object and third title can be set
Mapping table, and then between can determining that communication object and third are referred to as by inquiry mapping table in application
Corresponding relationship.For example, when user communicates with communication object, if between communication object and third can be inquired referred to as
Corresponding relationship then judges and obtains user setting third title corresponding with communication object.
Step 302, if knowing that active user automatically replies in information is provided with third title in function, third title is used
It is inserted into the address position of each reply content, generates one or more candidate return information.
It should be noted that the priority due to third title is higher than the first title and the second title, judging simultaneously
Active user is obtained when information automatically replies and is provided with third title in function, no longer obtains the first title and the second title,
But third title is directly inserted into the address position of each reply content, to generate candidate return information, and then obtains and work as
Target retro information is sent to communication by application program by the target retro information that preceding user selects from candidate return information
Object.Treatment effeciency is improved as a result, allows to believe the address of communication object in user's sets itself quickly revert content
Breath.
In one embodiment of the invention, if knowing that active user does not automatically reply in information is arranged third in function
Title, then according to the first title of the title Concordance identification communication object to contextual information.
The information automatic reply method of the embodiment of the present invention is by judging that active user automatically replies in function in information
It is no to be provided with third title corresponding with communication object and store, and know that active user automatically replies function in information in judgement
In be provided with third title when, the address position of each reply content is inserted into third title, generates one or more time
Select return information；Active user is known not when information automatically replies and third title is arranged in function in judgement, according to upper
First title of the title Concordance identification communication object of context information.As a result, user can fast be returned with sets itself
The salutation information of communication object is further improved so that the address in quickly revert more meets user's expectation in multiple content
The personalization level of quickly revert, the user experience is improved.
In order to realize above-described embodiment, the invention also provides a kind of information to automatically reply device, and Fig. 6 is that the present invention is implemented
A kind of information provided by example automatically replies the structural schematic diagram of device, as shown in fig. 6, the information automatically replies device includes:
First generation module 100 obtains module 200, processing module 300, the second generation module 400, sending module 500.
Wherein, the first generation module 100 passes through application program for analyzing active user according to Natural Language Processing Models
With the contextual information of communication object, one or more reply contents are automatically generated；
Module 200 is obtained, for identifying the first title of communication object according to the title Concordance to contextual information,
And according to the second title for reading communication object using setting information of application program；
Whether processing module 300 is identical for comparing the first title and the second title；
Second generation module 400, if knowing that the first title and the second title be not identical for comparing, by the first title and
Second title is inserted respectively into the address position of each reply content, generates multiple candidate return informations；
Sending module 500, the target retro information selected from candidate return information for obtaining active user, by target
Return information is sent to communication object by application program.
On the basis of Fig. 6, the information that Fig. 7 is provided automatically replies device further include: third generation module 600 screens mould
Block 700, the 4th generation module 800.
Wherein, third generation module 600, for obtaining the target retro information of active user's selection in preset period of time
The parameter information of training network neural model, generates title scoring model.
Screening module 700, is used for
According to title scoring model pair candidate return information corresponding with the first title, and time corresponding with the second title
Return information is selected to give a mark；
The low candidate return information of score is deleted according to marking result, provides a user the high candidate return information of score.
4th generation module 800, is used for
Judge that active user automatically replies in function whether be provided with third title corresponding with communication object simultaneously in information
Storage, wherein the priority of third title is higher than the first title and the second title；
If knowing, active user does not automatically reply setting third title in function in information, according to contextual information
Title Concordance identification communication object the first title；
If knowing that active user automatically replies in information is provided with third title in function, it is inserted into often with third title
The address position of a reply content generates one or more candidate return information.
Further, module 200 is obtained, specifically for the application page setting information according to application program, from contact person
The second title of communication object is read in name control.
Further, module 200 is obtained, specifically for the application registration setting information according to application program, from login name
Claim the second title that communication object is read in control.
It should be noted that previous embodiment is equally applicable to the present embodiment to the explanation of information automatic reply method
Information automatically reply device, details are not described herein again.
In conclusion the information of the embodiment of the present invention automatically replies device, analyzed according to Natural Language Processing Models current
User passes through the contextual information of application program and communication object, automatically generates one or more reply contents, and then according to right
First title of the title Concordance identification communication object of contextual information, and read according to the application setting information of application program
The second title of communication object is taken, further relatively whether the first title and the second title are identical, know the first title if comparing
It is not identical with the second title, then the first title and the second title are inserted respectively into the address position of each reply content, generated
Multiple candidate's return informations further obtain the target retro information that active user selects from candidate return information, by target
Return information is sent to communication object by application program.As a result, by utilizing contact data and context semantic information, obtain
It takes the salutation information to communication object and is added in quickly revert sentence, so that the address in quickly revert more meets user's phase
It hopes, improves the accuracy of quickly revert, make fast replying function more intelligent and personalized.
In order to realize above-described embodiment, the present invention also proposes that a kind of terminal device, Fig. 8 are provided by the embodiment of the present invention
A kind of structural schematic diagram of terminal device.As shown in figure 8, the terminal device includes: memory 21, processor 22 and is stored in
On reservoir 21 and the computer program that can be run on processor 22.
Processor 22 realizes the information automatic reply method provided in above-described embodiment when executing described program, the information is certainly
Dynamic answering method includes:
Step 101 ', according to Natural Language Processing Models analyze active user by application program and communication object up and down
Literary information automatically generates one or more reply contents.
Step 102 ', according to the first title of the title Concordance identification communication object to contextual information, and according to
The second title that communication object is read using setting information of application program.
Step 103 ', compare the first title and whether the second title is identical, knows the first title and the second title if comparing
It is not identical, then the first title and the second title are inserted respectively into the address position of each reply content, generated multiple candidate times
Complex information.
Step 104 ', the target retro information that active user selects from candidate return information is obtained, target retro is believed
Breath is sent to communication object by application program.
Further, terminal device further include:
Communication interface 23, for the communication between memory 21 and processor 22.
Memory 21, for storing the computer program that can be run on processor 22.
Memory 21 may include high speed RAM memory, it is also possible to further include nonvolatile memory (non-volatile
Memory), a for example, at least magnetic disk storage.
Processor 22 realizes information automatic reply method described in above-described embodiment when for executing described program.
If memory 21, processor 22 and the independent realization of communication interface 23, communication interface 21, memory 21 and processing
Device 22 can be connected with each other by bus and complete mutual communication.The bus can be industry standard architecture
(Industry Standard Architecture, referred to as ISA) bus, external equipment interconnection (Peripheral
Component, referred to as PCI) bus or extended industry-standard architecture (Extended Industry Standard
Architecture, referred to as EISA) bus etc..The bus can be divided into address bus, data/address bus, control bus etc..
Only to be indicated with a thick line in Fig. 8, it is not intended that an only bus or a type of bus convenient for indicating.
Optionally, in specific implementation, if memory 21, processor 22 and communication interface 23, are integrated in chip piece
Upper realization, then memory 21, processor 22 and communication interface 23 can complete mutual communication by internal interface.
Processor 22 may be a central processing unit (Central Processing Unit, referred to as CPU), or
It is specific integrated circuit (Application Specific Integrated Circuit, referred to as ASIC), or is matched
It is set to the one or more integrated circuits for implementing the embodiment of the present invention.
In order to realize above-described embodiment, the present invention also proposes a kind of computer program product, when in computer program product
Instruction processing unit realize information automatic reply method as described in aforementioned any embodiment when executing.
In order to realize above-described embodiment, the present invention also proposes a kind of non-transitorycomputer readable storage medium, deposits thereon
Computer program is contained, the information side of automatically replying as described in aforementioned any embodiment is realized when which is executed by processor
Method.
In the description of the present invention, it is to be understood that, term " first ", " second " are used for description purposes only, and cannot
It is interpreted as indication or suggestion relative importance or implicitly indicates the quantity of indicated technical characteristic.Define as a result, " the
One ", the feature of " second " can explicitly or implicitly include at least one of the features.In the description of the present invention, " multiple "
It is meant that at least two, such as two, three etc., unless otherwise specifically defined.
In the description of this specification, reference term " one embodiment ", " some embodiments ", " example ", " specifically show
The description of example " or " some examples " etc. means specific features, structure, material or spy described in conjunction with this embodiment or example
Point is included at least one embodiment or example of the invention.In the present specification, schematic expression of the above terms are not
It must be directed to identical embodiment or example.Moreover, particular features, structures, materials, or characteristics described can be in office
It can be combined in any suitable manner in one or more embodiment or examples.In addition, without conflicting with each other, the skill of this field
Art personnel can tie the feature of different embodiments or examples described in this specification and different embodiments or examples
It closes and combines.
Although the embodiments of the present invention has been shown and described above, it is to be understood that above-described embodiment is example
Property, it is not considered as limiting the invention, those skilled in the art within the scope of the invention can be to above-mentioned
Embodiment is changed, modifies, replacement and variant.
Claims (10)
1. a kind of information automatic reply method, which comprises the following steps:
The contextual information that active user passes through application program and communication object is analyzed according to Natural Language Processing Models, it is automatic raw
At one or more reply contents；
The first title of the communication object is identified according to the title Concordance to the contextual information, and is answered according to described
With the second title for reading the communication object using setting information of program；
Compare first title and whether second title is identical, knows first title and the second place if comparing
Title is not identical, then first title and second title are inserted respectively into the address position of each reply content, generates
Multiple candidate's return informations；
The target retro information that the active user selects from the candidate return information is obtained, by the target retro information
The communication object is sent to by the application program.
2. the method as described in claim 1, which is characterized in that described to be read according to the application setting information of the application program
Second title of the communication object, comprising:
According to the application page setting information of the application program, the of the communication object is read from contact name control
Two titles.
3. the method as described in claim 1, which is characterized in that described to be read according to the application setting information of the application program
Second title of the communication object, comprising:
According to the application registration setting information of the application program, the second of the communication object is read from register name control
Title.
4. the method as described in claim 1, which is characterized in that be in first title and second title
It is no it is identical after, further includes:
Know that first title is identical with second title if comparing, to first title and second title into
Row duplicate removal processing；
First title address position that perhaps second title is inserted into each reply content is generated one or more
A candidate's return information.
5. the method as described in claim 1, which is characterized in that further include:
The parameter information of the target retro information training network neural model of active user's selection in preset period of time is obtained, it is raw
At title scoring model.
6. method as claimed in claim 5, which is characterized in that be inserted into respectively by first title and second title
To the address position of each reply content, after generating multiple candidate return informations, further includes:
According to the title scoring model pair candidate return information corresponding with first title, and with second title
Corresponding candidate's return information is given a mark；
The low candidate return information of score is deleted according to marking result, provides a user the high candidate return information of score.
7. the method as described in claim 1-6 is any, which is characterized in that in the basis to the title of the contextual information
Before Concordance identifies that the first place of the communication object is referred to as, further includes:
Judge that the active user automatically replies in function whether be provided with third corresponding with the communication object in information
Claim and store, wherein the priority of the third title is higher than first title and second title；
If knowing that the active user does not automatically reply in information is arranged the third title in function, according to on described
The title Concordance of context information identifies the first title of the communication object；
If knowing that the active user automatically replies in information is provided with the third title in function, with the third title
It is inserted into the address position of each reply content, generates one or more candidate return information.
8. a kind of information automatically replies device characterized by comprising
First generation module passes through application program and communication object for analyzing active user according to Natural Language Processing Models
Contextual information automatically generates one or more reply contents；
Module is obtained, for according to the first place for identifying the communication object to the title Concordance of the contextual information
Claim, and according to the second title for reading the communication object using setting information of the application program；
Whether processing module is identical for first title and second title；
Second generation module, if knowing that first title and second title be not identical for comparing, by described first
Title and second title are inserted respectively into the address position of each reply content, generate multiple candidate return informations；
Sending module, the target retro information selected from the candidate return information for obtaining the active user, by institute
It states target retro information and the communication object is sent to by the application program.
9. device as claimed in claim 8, which is characterized in that the acquisition module is specifically used for:
According to the application page setting information of the application program, the of the communication object is read from contact name control
Two titles.
10. device as claimed in claim 8, which is characterized in that the acquisition module is specifically used for:
According to the application registration setting information of the application program, the second of the communication object is read from register name control
Title.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810404909.1A CN110417635B (en) | 2018-04-28 | 2018-04-28 | Automatic information reply method, device and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810404909.1A CN110417635B (en) | 2018-04-28 | 2018-04-28 | Automatic information reply method, device and equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110417635A true CN110417635A (en) | 2019-11-05 |
| CN110417635B CN110417635B (en) | 2021-08-24 |
Family
ID=68357571
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810404909.1A Active CN110417635B (en) | 2018-04-28 | 2018-04-28 | Automatic information reply method, device and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110417635B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111443973A (en) * | 2020-03-26 | 2020-07-24 | 腾讯科技（深圳）有限公司 | Filling method, device and equipment of remark information and storage medium |
| CN111935496A (en) * | 2020-08-21 | 2020-11-13 | 腾讯科技（深圳）有限公司 | Information display method and related equipment |
| CN112527126A (en) * | 2020-12-23 | 2021-03-19 | 北京百度网讯科技有限公司 | Information acquisition method and device and electronic equipment |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7085806B1 (en) * | 2000-11-03 | 2006-08-01 | Matchnet Plc | Method and apparatus for recommending a match to another |
| CN103051769A (en) * | 2012-12-14 | 2013-04-17 | 广东欧珀移动通信有限公司 | Method and system of mobile terminal for adding contact name to message |
| CN104144392A (en) * | 2013-05-08 | 2014-11-12 | 华为技术有限公司 | Message processing method, device and system |
| US20160306800A1 (en) * | 2015-04-16 | 2016-10-20 | Fluenty Korea Inc. | Reply recommendation apparatus and system and method for text construction |
| CN106789595A (en) * | 2017-01-17 | 2017-05-31 | 北京诸葛找房信息技术有限公司 | Information-pushing method and device |
| CN107577655A (en) * | 2016-07-05 | 2018-01-12 | 北京国双科技有限公司 | Name acquiring method and apparatus |
| CN107636568A (en) * | 2015-06-07 | 2018-01-26 | 苹果公司 | Response of the personalization prediction to instant message |
| CN107690781A (en) * | 2015-04-16 | 2018-02-13 | 三星电子株式会社 | Method and apparatus for recommending answer message |
-
2018
- 2018-04-28 CN CN201810404909.1A patent/CN110417635B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7085806B1 (en) * | 2000-11-03 | 2006-08-01 | Matchnet Plc | Method and apparatus for recommending a match to another |
| CN103051769A (en) * | 2012-12-14 | 2013-04-17 | 广东欧珀移动通信有限公司 | Method and system of mobile terminal for adding contact name to message |
| CN104144392A (en) * | 2013-05-08 | 2014-11-12 | 华为技术有限公司 | Message processing method, device and system |
| US20160306800A1 (en) * | 2015-04-16 | 2016-10-20 | Fluenty Korea Inc. | Reply recommendation apparatus and system and method for text construction |
| CN107690781A (en) * | 2015-04-16 | 2018-02-13 | 三星电子株式会社 | Method and apparatus for recommending answer message |
| CN107636568A (en) * | 2015-06-07 | 2018-01-26 | 苹果公司 | Response of the personalization prediction to instant message |
| CN107577655A (en) * | 2016-07-05 | 2018-01-12 | 北京国双科技有限公司 | Name acquiring method and apparatus |
| CN106789595A (en) * | 2017-01-17 | 2017-05-31 | 北京诸葛找房信息技术有限公司 | Information-pushing method and device |
Non-Patent Citations (3)
| Title |
|---|
| C小姐漫谈: ""微信公众号怎么设置关键词回复"", 《HTTPS://JINGYAN.BAIDU.COM/ARTICLE/B907E62780098846E7891CA3.HTML》 * |
| PATILL DIPAK: ""intelligent automated message for computer-implemented devices"", 《FREEPATENTSONLINE》 * |
| 梁建增: ""用于客服辅助的对话模型研究"", 《万方数据库》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111443973A (en) * | 2020-03-26 | 2020-07-24 | 腾讯科技（深圳）有限公司 | Filling method, device and equipment of remark information and storage medium |
| CN111443973B (en) * | 2020-03-26 | 2021-10-08 | 腾讯科技（深圳）有限公司 | Filling method, device and equipment of remark information and storage medium |
| CN111935496A (en) * | 2020-08-21 | 2020-11-13 | 腾讯科技（深圳）有限公司 | Information display method and related equipment |
| CN112527126A (en) * | 2020-12-23 | 2021-03-19 | 北京百度网讯科技有限公司 | Information acquisition method and device and electronic equipment |
| CN112527126B (en) * | 2020-12-23 | 2023-11-07 | 北京百度网讯科技有限公司 | Information acquisition method and device and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110417635B (en) | 2021-08-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR100695392B1 (en) | A method for converting SMS message to multimedia message and sending the multimedia message and text-image converting server | |
| CN110417635A (en) | Automatic information reply method, device and equipment | |
| CN106776936A (en) | intelligent interactive method and system | |
| CN109254669A (en) | A kind of expression picture input method, device, electronic equipment and system | |
| CN107797984A (en) | Intelligent interactive method, equipment and storage medium | |
| CN112291423B (en) | Communication call intelligent response processing method and device, electronic equipment and storage medium | |
| CN112507139B (en) | Knowledge graph-based question and answer method, system, equipment and storage medium | |
| CN113254815B (en) | Document processing method and device and page processing method and device | |
| CN104915345B (en) | A kind of method and terminal of recommendation service information | |
| CN111222837A (en) | Intelligent interviewing method, system, equipment and computer storage medium | |
| CN111488186A (en) | Data processing method and device, electronic equipment and computer storage medium | |
| CN113806503A (en) | Dialog fusion method, device and equipment | |
| CN116168119A (en) | Image editing method, image editing device, electronic device, storage medium, and program product | |
| CN110209768B (en) | Question processing method and device for automatic question answering | |
| CN110187780B (en) | Long text prediction method, long text prediction device, long text prediction equipment and storage medium | |
| US20230126821A1 (en) | Systems, devices and methods for the dynamic generation of dialog-based interactive content | |
| CN110008445A (en) | Event extraction method and device, electronic equipment | |
| CN107423441B (en) | Picture association method and device and electronic equipment | |
| CN105956593A (en) | Image processing method and device | |
| CN108595141A (en) | Pronunciation inputting method and device, computer installation and computer readable storage medium | |
| CN107357444A (en) | A kind of input pattern intelligent switch method and device based on usage scenario | |
| CN112052316A (en) | Model evaluation method, model evaluation device, storage medium and electronic equipment | |
| KR100627853B1 (en) | A method for converting sms message to multimedia message and sending the multimedia message and text-image converting server | |
| KR101786823B1 (en) | Method for providing photo in sns | |
| CN108804652A (en) | Generation method, device, storage medium and the electronic device of cover picture |
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 |