CN110211001A - A kind of hotel assistant customer service system, data processing method and relevant device - Google Patents
A kind of hotel assistant customer service system, data processing method and relevant device Download PDFInfo
- Publication number
- CN110211001A CN110211001A CN201910410960.8A CN201910410960A CN110211001A CN 110211001 A CN110211001 A CN 110211001A CN 201910410960 A CN201910410960 A CN 201910410960A CN 110211001 A CN110211001 A CN 110211001A
- Authority
- CN
- China
- Prior art keywords
- user
- information
- order
- template
- text
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0633—Lists, e.g. purchase orders, compilation or processing
- G06Q30/0635—Processing of requisition or of purchase orders
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/12—Hotels or restaurants
Landscapes
- Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Engineering & Computer Science (AREA)
- Finance (AREA)
- Economics (AREA)
- Accounting & Taxation (AREA)
- Marketing (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Human Resources & Organizations (AREA)
- General Health & Medical Sciences (AREA)
- Development Economics (AREA)
- Telephonic Communication Services (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
This application provides a kind of hotel assistant customer service system, data processing method and relevant devices, which comprises obtains the first voice messaging of user, wherein the first voice messaging of the user includes that the room number of user identifies;The first text information is converted by first voice messaging;First text information is matched with the keyword of order template, obtains matched first order template, wherein the order template is used to generate order information according to text information;According to first text information and the first order template, order information is generated;The room number mark and the order information are sent to the terminal of attendant.
Description
Technical field
This application involves field of artificial intelligence more particularly to a kind of hotel assistant customer service systems, data processing method
And relevant device.
Background technique
With the continuous promotion of people's living standard, client is also higher and higher for the service request in hotel, more visitors
Family wishes that the demand of oneself can be met at the first time in hotel accommodations, so that the energy in hotel, manpower, consumer goods etc.
The management of articles also becomes increasingly complex.
Currently, the housekeeping service in hotel is also in the manual service stage, client needs through phone in room and foreground
Personnel carry out telephonic communication, propose the housekeeping service oneself needed, for example need replacing sheet, need replacing hair dryer, maintenance
Remote controler etc. after the personnel record's user demand of foreground, is reinformed and is handled to corresponding logistics service personnel.It is this to pass through
The mode for sending a telegraph foreground request housekeeping service is made a phone call there are many drawbacks, is mainly shown as: manually being passed by reception counter of hotel
Housekeeping service information is passed, the time is wasted, reduces working efficiency;For scale hotel, contain much information, manual record efficiency
It is low, directly affect hotel service quality;Tenant, reception counter of hotel, logistics are extremely difficult to efficient synchronizing information etc..
Summary of the invention
This application provides a kind of hotel assistant customer service system, data processing method and relevant devices, current for solving
The low efficiency of housekeeping service, the problem that time-consuming, customer experience is bad.
In a first aspect, a kind of data processing method is provided, the method includes
Obtain the first voice messaging of user, wherein the first voice messaging of the user includes the room number mark of user
Know;
The first text information is converted by first voice messaging;
First text information is matched with the keyword of order template, obtains matched first order template,
Wherein, the order template is used to generate order information according to text information;
According to first text information and the first order template, order information is generated;
The room number mark and the order information are sent to the terminal of attendant.
Optionally, it is described convert the first text information for first voice messaging after, the method also includes: according to
First text information determines that the first interaction of the user is intended to Consultant or Task;The first of the user
In the case that interaction is intended to Consultant, first text information is matched with consulting template, first is obtained and replies text
This;Described first is exported to the user by output equipment and replies text, wherein the output equipment includes at least loudspeaker
And one or more of display screen;It is described to match first text information with the keyword of order template,
Obtaining matched first order template includes: in the case where the first interaction of the user is intended to Task, by described the
One text information is matched with the keyword of order template, obtains matched first order template.
Optionally, after the acquisition order information, the method also includes: determine whether the order information is complete;?
In the incomplete situation of order information, accordingly art template is obtained according to the missing content of the order information, is generated
First inquiry text;The first inquiry text is exported to the user by output equipment;Obtain the second language of the user
Message breath, converts the second text information for second voice messaging;According to second text information, the user is determined
Second interaction be intended to Consultant or Task;In the case where the second interaction of the user is intended to Consultant, by institute
It states the second text information to be matched with consulting template, obtains second and reply text;It is exported by output equipment to the user
Described second replies text;In the case where the second interaction of the user is intended to Task, according to second text envelope
Breath improves the missing content of the order information;It is described to send room number mark to the terminal of attendant and described order
Single information include: in the complete situation of the order information, to the terminal of attendant send room number mark and
The order information.
Optionally, it is described by output equipment to the user export it is described second reply text after, the method is also
It include: that the first inquiry text is exported to the user by output equipment;The third voice messaging of the user is obtained, it will
The third voice messaging is converted into third text information;According to the third text information, determine that the third of the user is handed over
Mutually it is intended to Consultant or Task;In the case where the interaction of the third of the user is intended to Consultant, by the third text
This information is matched with consulting template, is obtained third and is replied text;The third is exported to the user by output equipment
After replying text, text is inquired to user output described first again by output equipment;It is handed in the third of the user
In the case where being mutually intended to Task, the missing content of the order information is improved according to the third text information.
Optionally, before first voice messaging for obtaining user, the method also includes: obtain the attendant
Available information on services;According to the information on services, the keyword of the order template and the order template is obtained.
Optionally, the order template is generated by specification, logic reasoning or machine learning in advance, the order mould
Version includes multiple slot positions, described according to first text information and the first order template, and generating order information includes: basis
First text information fills multiple slot positions of the first order template;The slot position includes that must fill out slot position and select to fill out
Slot position, what whether the determination order information completely comprised determining that the first order template must fill out whether slot position fills
Completely;It is described that according to the acquisition of the missing content of the order information, accordingly art template includes: to be lacked according to the order information
The slot position of mistake obtains art template if corresponding to, wherein the words art template passes through specification, logic reasoning or machine in advance
Study generates, and each slot position of the order information corresponds to what one or more said art template, and the words art template is used for basis
The slot position of the missing of the order information generates corresponding inquiry text;It is described that described order is improved according to second text information
The missing content of single information includes: to fill the slot position of the order template according to second text information.
Second aspect provides a kind of hotel assistant customer service system, which is characterized in that including acquiring unit, conversion unit,
Matching unit, generation unit and transmission unit, wherein
The acquiring unit is used to obtain the first voice messaging of user, and the first voice messaging of the user includes user
Room number mark;
The conversion unit is used to convert the first text information for first voice messaging;
The matching unit is matched for matching first text information with the keyword of order template
The first order template;
The generation unit is used to generate order information according to first text information and the first order template;
The transmission unit is used to send the address mark and order information to the terminal of attendant.
Optionally, the system also includes determination unit, the determination unit is used to believe first voice described
After breath is converted into the first text information: according to first text information, determining that the first interaction of the user is intended to seek advice from
Type or Task;In the case where the first interaction of the user is intended to Consultant, by first text information and consulting
Template is matched, and is obtained first and is replied text;Described first, which is exported, to the user by output equipment replies text,
In, the output equipment includes at least one or more of loudspeaker and display screen;The matching unit is specifically used for:
In the case where the first interaction of the user is intended to Task, by the keyword of first text information and order template
It is matched, obtains matched first order template.
Optionally, the determination unit is also used to after the acquisition order information: determining whether the order information is complete
It is whole;In the incomplete situation of the order information, accordingly art template is obtained according to the missing content of the order information,
Generate the first inquiry text;The first inquiry text is exported to the user by output equipment;Obtain the of the user
Second voice messaging is converted the second text information by two voice messagings;According to second text information, determine described in
The second interaction of user is intended to Consultant or Task;In the case where the second interaction of the user is intended to Consultant,
Second text information is matched with consulting template, second is obtained and replies text;By output equipment to the user
It exports described second and replies text;In the case where the second interaction of the user is intended to Task, according to second text
This information improves the missing content of the order information;The transmission unit is specifically used for: in the complete feelings of the order information
Under condition, the room number mark and the order information are sent to the terminal of attendant.
Optionally, the determination unit is also used to export second answer to the user by output equipment described
After text: exporting the first inquiry text to the user by output equipment;Obtain the third voice letter of the user
Breath, converts third text information for the third voice messaging;According to the third text information, the of the user is determined
Three interactions are intended to Consultant or Task;In the case where the interaction of the third of the user is intended to Consultant, by described the
Three text informations are matched with consulting template, are obtained third and are replied text;By output equipment to described in user output
After third replies text, text is inquired to user output described first again by output equipment;The of the user
In the case that three interactions are intended to Task, the missing content of the order information is improved according to the third text information.
Optionally, the acquiring unit is also used to before first voice messaging for obtaining user, obtains the clothes
The available information on services of business personnel;The acquiring unit is also used to according to the information on services, obtain the order template with
And the keyword of the order template.
Optionally, the order template is generated by specification, logic reasoning or machine learning in advance, the order mould
Version includes multiple slot positions, and the generation unit is specifically used for: according to first text information, filling the first order template
Multiple slot positions;The slot position includes that must fill out slot position and select to fill out slot position, and the determination unit is specifically used for: determining described first
Order template must fill out whether slot position fills completely;The determination unit is specifically used for: the slot lacked according to the order information
Position obtains art template if corresponding to, wherein the words art template is raw by specification, logic reasoning or machine learning in advance
At each slot position of the order information corresponds to what one or more said art template, and the words art template according to for ordering
The slot position of the missing of single information generates corresponding inquiry text;The determination unit is also used to: according to second text information,
Fill the slot position of the order template.
The third aspect provides a kind of server, including processor and memory;The memory refers to for storing
It enables, the processor realizes method described in first aspect when executing described instruction for executing described instruction, the processor.
Fourth aspect provides a kind of computer non-transitory storage media, the computer non-transitory storage media storage
There is computer program, which is characterized in that method described in first aspect is realized when the computer program is executed by calculating equipment.
5th aspect, provides a kind of computer program product, which is characterized in that when the computer program product is counted
It calculates machine-readable when taking and executing, realizes method described in first aspect.
Based on hotel assistant customer service system, data processing method and relevant device provided by the present application, by obtaining user
The first voice messaging, the first voice messaging of the user includes that the room number of user identifies, by first voice messaging
It is converted into the first text information, first text information is matched with the keyword of order template, obtains matched the
One order template generates order information, thus to attendant's according to first text information and the first order template
Terminal sends the room number mark and the order information.Use hotel assistant customer service system provided by the present application, data
Processing method and relevant device, when user needs housekeeping service, the hotel Zhi Xuyu assistant's customer service system is interacted, can be self-service
It completes housekeeping service order to cut operating costs to reduce the operating pressure of foreground personnel, promotes the work of hotel's logistics service
Make efficiency, and since system server can handle multiple user demands simultaneously, thus greatly reduce the waiting time of user,
Further promote the usage experience of user.
Detailed description of the invention
In order to illustrate the technical solutions in the embodiments of the present application or in the prior art more clearly, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of application without any creative labor, may be used also for those of ordinary skill in the art
To obtain other drawings based on these drawings.
Fig. 1 is a kind of running environment of data processing method provided by the present application and the system tray of hotel's assistant's customer service system
Structure schematic diagram;
Fig. 2 is a kind of flow diagram of data processing method provided by the present application;
Fig. 3 is a kind of hotel assistant customer service system configuration page schematic diagram provided by the present application;
Fig. 4 is the flow diagram that a kind of complete order provided by the present application generates;
Fig. 5 is that the user under an application scenarios provided by the present application and the record of the human-computer dialogue between hotel's customer service system show
It is intended to;
Fig. 6 is a kind of structural schematic diagram of hotel assistant customer service system provided by the present application;
Fig. 7 is a kind of structural schematic diagram of server provided by the present application.
Specific embodiment
The application is described in further detail below by specific embodiment combination attached drawing.In the following embodiments and the accompanying drawings
In, many datail descriptions are in order to enable the application can be better understood.However, those skilled in the art can be without lifting an eyebrow
Recognize, part of feature is dispensed in varied situations, or can be substituted by other methods.Certain
In the case of, the relevant some operations of the application are there is no display in the description or describe, this is the core in order to avoid the application
Center portion point is flooded by excessive description.To those skilled in the art, be described in detail these relevant operations be not must
It wants, they can completely understand relevant operation according to the general technology knowledge of description and this field in specification.
Below with reference to Fig. 1, running environment and hotel's assistant's customer service system to data processing method provided by the present application
System architecture is briefly described.As shown in Figure 1, hotel assistant customer service system provided by the present application include director server with
And multiple child servers, wherein the director server is connected with the child servers in multiple guest rooms and service department, each guest room
Child servers are connected in room for collecting the first terminal of user speech, the child servers of each service department and the service
Second terminal that the attendant of department is held is connected, here, the first terminal in room can be guestroom telephone, television set,
Intelligent sound box etc. can collect the terminal device of user speech information, and the second terminal that attendant is held can be mobile phone, right
Say that machine, tablet computer etc. can send the terminal device of order notification to attendant, the application is not especially limited.For example,
User A passes through phone A1 (first terminal) predetermined housekeeping service in room A: " I needs a fiber pillow ", phone A1 is obtained
After taking the voice messaging at family, which is sent to the child servers of room A, the child servers of room A are by the voice
Information feeds back to director server, after director server implements data processing method provided by the present application, the order information that will acquire
" room A needs a fiber pillow " is sent to the child servers of guest room department (the service department A of example as shown in figure 1), the guest room
The child servers of department notify the attendant for being responsible for room A to go to handle the order according to the order information, for example, notice room
Between the attendant A1 of floor where A take a fiber pillow to be sent into room A.It should be understood that the example above is merely to illustrate, it is described
Hotel's assistant's customer service system can also include more rooms, more service departments, more terminal devices etc., the application
It is not especially limited.
Below with reference to running environment shown in FIG. 1, data processing method provided by the present application is described in detail.Fig. 2
It is a kind of flow diagram of data processing method provided by the present application, as shown in Figure 2, data processing method provided by the present application
The following steps are included:
S101: the first voice messaging of user is obtained.
In the embodiment of the present application, the first voice messaging of the user includes that the room number of user identifies.With reference to aforementioned
Content is it is found that first voice messaging can be by obtaining for collecting the first terminal of user speech in guest room, and the
One terminal can be guestroom telephone, and user dials exchange number by dialing guestroom telephone, the call voice assistant in guestroom telephone
Obtain the first voice messaging of user;For another example, first terminal is the intelligent sound box in guest room, and user passes through specified speech phrase
Form wake up intelligent sound box after, intelligent letter box obtains the first voice messaging of user, it should be appreciated that the first terminal can be with
It is the terminal of other available user speech information, for example shape is the intelligent robot of a certain cartoon character, such as user
The application program installed on mobile phone, wechat small routine, public platform etc., the application is not especially limited.It should be noted that due to
User just may require that some housekeeping services after being generally into guest room, that is to say, that first voice messaging be with
It is obtained in the guest room at family, therefore corresponding room number can be set in the first terminal in guest room for obtaining user speech
Mark, so that the voice messaging that the first terminal obtains includes the room number mark in room locating for the first terminal, so as to
Attendant provides housekeeping service according to room number for user.
S102: the first text information is converted by first voice messaging.
In the specific implementation, after first terminal collects first voice messaging and is uploaded to director server, director server
Speech recognition module (Automatic Speech Recognition, ASR) can be by speech recognition algorithm, by a Duan Yin
Frequency becomes corresponding text, that is to say, that converts the first text information for first voice messaging, wherein the voice
Recognizer can be convolutional neural networks (Convolutional Neural Networks, CNN) algorithm, backpropagation mind
It is calculated through network (back propagation, BP) algorithm, radial base neural net (Radial Basis Function, RBF)
Method, fuzzy clustering (Fuzzy C-means, FCM) neural network algorithm, Recognition with Recurrent Neural Network (Recurrent Neural
Network, RNN) algorithm, wavelet neural network (Wavelet Neural Network, WNN) algorithm, Hidden Markov Model
(Hidden Markov Model, HMM) algorithm, orthogonal basis function opposite direction communication process neural network
(Counterpropagation Process Neural Network, CPNN) algorithm, support vector machines (Support
Vector Machine, SVM) Multiclass Classification etc., the application is not especially limited.
In the embodiment of the present application, it is described convert the first text information for first voice messaging after, the method
Further include: according to first text information, determine that the first interaction of the user is intended to Consultant or Task;Described
In the case that the first interaction of user is intended to Consultant, first text information is matched with consulting template, is obtained
First replies text;Described first is exported to the user by output equipment and replies text, wherein the output equipment is at least
Including one or more of loudspeaker and display screen;The keyword by first text information and order template
Matched, obtain matched first order template include: in the case where the first interaction of the user is intended to Task,
First text information is matched with the keyword of order template, obtains matched first order template.It is understood that
, the ASR module in foregoing teachings, although first voice messaging is converted to the first text information, ASR module is simultaneously
The meaning of first text information is not known, and therefore, first text information can be sent to semantic understanding mould by ASR module
Block (Natural Language Understanding, NLU), NLU module are understood that the intention (Intent) of user, specifically
In realization, NLU can determine that the first interaction of the user is intended to information type or Task first, then distributing them to can be with
Handle the program of the intention, wherein the program for handling the intention can be " robot (Bot) ".For example user says: " tomorrow day
How is gas? ", NLU judges that user's is intended that Consultant, it is intended that classification is weather, therefore calls out weather robot (Bot)
Tell user's weather condition of tomorrow, if user also wonders posteriori weather, weather Bot will continue as user service,
Until user is intended to change, when being no longer that weather is intended to, then switching to other Bot is user service.
Briefly, NLU module understands that this process of the intention of user can simply be interpreted as a classifier, the classification
Device belongs to Consultant or Task by the first text information that the first voice messaging is converted into for determining, so as to following
It performs corresponding processing.Wherein, the sentence of Consultant can be " may I ask hotel eatery in a few buildings? " " nearby there is any restaurant's ratio
It is relatively nice " " there is the service for the machine of picking in hotel? " etc., the sentence of Task can be " I needs a pillow " " my needs
Change the sheet " " my TV is broken, and is needed repairing " etc., the application is not especially limited.Wherein, for determining what user was intended to
Classifier can be the disaggregated model obtained by machine learning, and specifically can be is had and task using a large amount of labels
What the text information sample training disaggregated model of type obtained, the application is not especially limited.
It, can be by by described the it should be noted that in the case where the first interaction of the user is intended to Consultant
One text information is matched with consulting template, to obtain the first answer text asked questions, wherein the consulting template
It can be the template in knowledge base, when the first text information and very high knowledge base consulting stencil matching degree, system meeting
Directly reply user's answer (unique response);If a plurality of consulting template in the first text information and knowledge base match and
When very high with degree, system can ask in reply user " you are intended to ask XX problem or XX problem " by loudspeaker or screen, from
And further determine that the intention of user, improve communication efficiency and the usage experience of user;If the first text information and knowledge
Knowledge matching degree in library is very low, system can not answer or refuse answer the problem, such as reply user " this problem I
Do not know ".It should be understood that the example above is merely to illustrate, specific restriction can not be constituted.
S103: first text information is matched with the keyword of order template, obtains matched first order
Template, wherein the order template is used to generate order information according to text information.
In the embodiment of the present application, before first voice messaging for obtaining user, the method also includes: obtain institute
State the available information on services of attendant;According to the information on services, the order template and the order template are obtained
Keyword.It should be understood that the service that each hotel can provide is different, for example, some hotels can provide laundry service,
Some hotels do not provide laundry service.Therefore, for hotel's customer service system before for users to use, hotel Fang Huixian can by oneself
The keyword of the service of offer configures in system background.For example, certain hotel cannot provide laundry service, may only
The service of replacement pillow is provided, therefore only keyword " pillow " can be arranged in system background in the system of configuration for hotel.
When the first voice messaging for receiving user is " I thinks laundry clothes ", since system background does not have the keyword of this respect,
That is hotel, without this service, system will return to user " sorry, hotel is temporarily without this service " this letter
Breath specifically can broadcast voice by the loudspeaker of first terminal and be notified, also be may be displayed on the screen of first terminal,
The application is not especially limited.And when the first voice messaging for receiving user is " I wants to change pillow ", after at this time
Platform has this keyword, that is to say, that there is this service in hotel, so that it may be matched to the corresponding order mould of " pillow " keyword
Version, to generate corresponding order information.
In the specific implementation, the order template can by standardizing in advance, logic reasoning or machine learning generate, open
Hair personnel can carry out type division according to the available service in multiple hotels, and each type trains an order template, each
Order template corresponds to one or more service keyword, can according to oneself hotel when hotel side carries out system configuration
The housekeeping service of offer selects keyword, so that system background obtains order template workable for the hotel.For example, Fig. 3 is this
Apply for a kind of configuration interface of the hotel assistant customer service system provided, wherein selectable keyword includes six classifications, also
It is to say, developer has trained the order template of six kinds of housekeeping services, and hotel personnel are when carrying out system configuration, if hotel is only
There is the housekeeping service of replacement pillow, then hotel side can only choose " pillow " keyword, background system will only " pillow " be closed
Key word and corresponding order template can be matched, other templates that do not choose can not be then matched.It should be understood that this
Application determines order template by keyword match, and method is simple and matching speed is fast, improves the response time of system, also,
Also very simple is convenient for the configuration step of hotel user simultaneously, improves the usage experience of user.It should be understood that the example above is only used for
Illustrate, specific restriction can not be constituted.
It should be noted that if the housekeeping service in hotel can segment type, for example the pillow in hotel is divided into " buck wheat pillow
Head " and " fiber pillow ", then hotel side is when carrying out system configuration, it is acceptable further after having selected " pillow " keyword
The attributive classification keyword of pillow is set, and the attributive classification keyword is added into order template, that is to say, that when with
When family needs replacing pillow, system can further inquire user according to attributive classification keyword, specifically need which attribute point
The pillow of class improves user experience, similarly, the attributive classification of sheet connects so that more accurate provides a user housekeeping service
Sending the vehicle classification of machine can also further set, and the application is not especially limited.
S104: according to first text information and the first order template, order information is generated.
In the specific implementation, the order template includes multiple slot positions (Slot), it is described according to first text information with
And the first order template, generating order information includes: to fill the more of the first order template according to first text information
A slot position.Wherein, the slot position includes essential slot position and optionally slot position.It should be understood that in the first interaction of the user
It, just can be complete after system must know the Partial key information in first text information in the case where being intended to Task
Being issued at user for task.For example, user says that " the pillow inner core for being intended to oneself room is replaced with buckwheat, sleeps in this way
Feel pleasant " so system is at least resolved to after Item Information and syntactic information be " pillow increases, 1 ", phase could be generated
The order information answered, and " sleep is pleasant " this content, system, which does not parse, can also obtain order information.Based on this, open
Hair personnel can carry out slot position design in system development, and core information needed for each order is designed as essential slot position, non-
Core information is designed as non-to fill out slot position.
For example, user has bought the air ticket of 5:00 AM, and under this application scenarios, user, which may may require that, wakes clothes
Business, then must fill out slot position is that can be waken with the time if user does not express clearly for the order template of wake-up service
Time, it is what the specific wake-up time that system, which will question closely user, that is, time slot position must be filled to complete, Er Feibi
It fills out slot position and can be and wake reason, must be filled out for slot position for non-, if there is in this respect in the first text information of user
Hold, so that it may the slot position is filled, if the not content of this respect, system will not question closely user to fill the slot
Position.In the specific implementation, the slot position design of wake-up service can be as shown in table 1 below.It should be understood that the example above and table 1 are only used
In for example, specific restriction can not be constituted.
The design of 1 wake-up service slot position of table
Slot position description | Type | Field name | Citing |
Time | It must fill out | remind_time | Tomorrow 5:00 AM |
Repetition period | It is non-to fill out | repeat | Daily |
Content | It is non-to fill out | title | It seizes the opportunity |
Alerting pattern | It is non-to fill out | remind_type | Telephone bell |
In the embodiment of the present application, after the acquisition order information, the method also includes: determine that the order information is
It is no complete;In the incomplete situation of the order information, accordingly art is obtained according to the missing content of the order information
Template generates the first inquiry text;The first inquiry text is exported to the user by output equipment;Obtain the user
The second voice messaging, convert the second text information for second voice messaging;According to second text information, determine
The second interaction of the user is intended to Consultant or Task;The case where the second interaction of the user is intended to Consultant
Under, second text information is matched with consulting template, second is obtained and replies text;By output equipment to the use
Family output described second replies text;In the case where the second interaction of the user is intended to Task, according to described second
Text information improves the missing content of the order information;It is described to the terminal of attendant send room number mark and
The order information includes: to send the room number mark to the terminal of attendant in the complete situation of the order information
Knowledge and the order information.It should be understood that just completing the probability of order information only according to the first voice messaging of user
It is low-down, and incomplete order information is sent to hotel staff, attendant can not also understands the tool of user
Body demand.Therefore, after using the first text information filling order template, system will judge whether order template has been filled
It is whole, if imperfect, it will art template inquires user if being corresponded to according to the content of missing.For example, user says that " I thinks
Change pillow ", system analysis to Item Information and syntactic information are " pillow, increase, 1 ", and according to the setting on hotel backstage,
The attribute keyword of hotel's pillow further comprises " buckwheat pillow/fiber pillow ", art template if being determined according to missing content
Keyword can be " product name: pillow, Attributions selection: fiber pillow/buckwheat pillow ", therefore can be matched to missing content
Art template if correspondence: " we provide [product name] type have [Attributions selection], may I ask you need it is any? ", thus
Obtain first inquiry text " we provide pillow type have fiber pillow and buckwheat pillow, may I ask you need it is any ".?
After first terminal gets the answer of user, order is further improved according to the answer of user, is ultimately sent to attendant's
Order can be just it is accurate, completely express the wish of user, and then greatly improve the treatment effeciency of attendant, improve user
Usage experience.
In the specific implementation, the determination order information whether completely comprise determining that the first order template must
Fill out whether slot position fills completely;It is described that according to the acquisition of the missing content of the order information, accordingly art template includes: basis
The slot position of the order information missing, obtains art template if corresponding to, wherein the words art template passes through specification, logic in advance
Inference method or machine learning generate, and each slot position of the order information corresponds to one or more words art template, the words
Art template is used to generate corresponding inquiry text according to the slot position of the missing of the order information;It is described according to second text
The missing content that information improves the order information includes: to fill the slot of the order template according to second text information
Position.That is, whether order information completely determine according to must fill out slot position and fill to finish, when system can not basis
When slot position is filled complete by the first text information, it will questioned closely according to the slot position that must fill out of missing to user, until order mould
The slot position that must fill out of version is stuffed entirely with finishing.Wherein, must each fill out slot position can correspond to one or more inquiry template, system
After what determination lacked must fill out slot position, user is inquired according to corresponding inquiry template, specifically, can be broadcast by loudspeaker
Report the first inquiry text, can also be by showing that screen shows the first inquiry text.The application is not especially limited.In this way,
The method that must fill out slot position is filled by cross-question user, greatly improves the order information for being ultimately sent to attendant
Integrality improves the treatment effeciency of hotel guest room service.For example, it is still exemplified by Table 1, it is assumed that first text information is
" I will catch up with aircraft tomorrow, I needs wake-up service ", slot position at this time can only fill " content " this non-slot position that must be filled out, and must
Filling out slot position " time " can not be filled, and system will question closely user about the content that must fill out slot position " time ", for example, system can be with
By the way that loudspeaker plays that first inquires text " you need me to wake you in the several points? ", thus the second text answered according to user
The order information of information " 5 points of morning " generation " wake-up service ".It should be understood that the example above is merely to illustrate, tool can not be constituted
Body limits.
It should be noted that system receives the voice messaging of user for the first time, that is, after the first voice messaging, it will complete
The interactive process of whole record system and user, that is to say, that record user believes from whole voices after the first voice messaging
Breath, so that dialogue later can all associate with conversation content before, when system passes through the first inquiry of output equipment output
After text, if user does not answer the inquiry, but the problem of asked a Consultant, then system can be first according to knowing
The consulting template known in library finds corresponding answer, if being matched to corresponding consulting template, will answer the Consultant problem,
Then the intersection record before basis comes again the first inquiry text, to meet the various aspects need of user to the greatest extent
It asks.
In the embodiment of the present application, it is described by output equipment to the user export it is described second reply text after,
The method also includes: the first inquiry text is exported to the user by output equipment;Obtain the third of the user
The third voice messaging is converted third text information by voice messaging;According to the third text information, the use is determined
The third interaction at family is intended to Consultant or Task;It, will in the case where the interaction of the third of the user is intended to Consultant
The third text information is matched with consulting template, is obtained third and is replied text;It is defeated to the user by output equipment
After the third replies text out, text is inquired to user output described first again by output equipment;In the use
In the case that the third interaction at family is intended to Task, improved in the missing of the order information according to the third text information
Hold.With reference to foregoing teachings it is found that NLU module is after determining that first text information belongs to Consultant or Task, meeting
First text information is distributed to the Bot that can handle the information, that is to say, that the complete of order information is generated in step S104
Portion's process is all to be completed by handling the Bot of the category information, including fill template, determines template integrality, and generate and correspond to
Answer text etc., until generating last complete order information.Fig. 4 is the first of a kind of acquisition user provided by the present application
After text information, the Bot for handling first text information obtains the entire flow figure of order information, and in simple terms, first eventually
The voice messaging of the user obtained every time is held, system all can be intended to classify by user of the NLU module to this voice messaging,
It determines Consultant or Task, is performed corresponding processing further according to type, intert Consultant if it is Task, then first
Consultant is answered, Task is repeated, until order is completed, the specific processing mode of Task and Consultant are specific
Processing mode specifically refers to foregoing teachings, and which is not described herein again.
For example, still by taking the wake-up service template of table 1 as an example, Fig. 5 is a kind of user provided by the present application and system
Session log schematic diagram, system output first inquiry text " you need me to wake you in the several points? " later, system obtain the
Two text informations are " is the dining room in your hotels in a few buildings? ", system can first determine whether that second text information is to belong at this time
Consultant or Task are found according to the consulting template in knowledge base first after determining that second text information is Consultant
Corresponding answer, and the second answer text is exported, for example, saying that " our hotel eatery is two to user by loudspeaker
Building ", then again by loudspeaker repeat output first inquiry text " you need me to wake you in the several points? ", thus not delaying
While user subscribes housekeeping service, answer user wants the problem of asking as far as possible, promotes the usage experience of user.It should be understood that
The dialogue of system and user can be pure voice dialogue, be also possible to plain text dialogue as shown in Figure 5, can also be voice with
Text combination, conversational mode of the video in conjunction with voice, for example directly broadcasted in the dialogue in Fig. 5 to user speech by phone
Hold, the customer service image in a hotel is directly displayed by television set, bound fraction limb action is broadcasted in Fig. 5 to user speech
Conversation content, the application are not especially limited.
In the specific implementation, complete order information can also be corresponding with words art template, for example, complete order information can be with
Be " Item Title: pillow, type: fiber pillow, quantity: 1 ", can also be " Item Title: compromising, quantity: 2 ", it can be with
It is " service name: wake-up service, time: 5 points of morning ", art template can be that " good, you, which have clicked, [orders at this time
Single information], if confirmation? ", system can be in conjunction with words art template and order information, and generating final user, to reply words art " good
, you have clicked pillow, and type is fiber pillow, quantity 1, slippers, quantity 2, wake-up service, 5 points of morning, if really
Recognize? ", order situation is further finally improved according to the confirmation situation of user, promotes the accuracy rate of order to the greatest extent, is improved
Housekeeping service efficiency and the satisfaction of user.
S105: the room number mark and the order information are sent to the terminal of attendant.
In the embodiment of the present application, after attendant receives complete order information and room number mark, Ji Kegen
According to order contents complete user needed for housekeeping service, wherein the order information that attendant receives can be text information,
Voice broadcast order information etc., the application is not especially limited, and optionally, system can be indicated according to room number and order
The order information is sent to the attendant that can handle the order information by information, for example, order information be " Item Title:
Pillow, type: fiber pillow, quantity: 1 ", room identification is rooms 301, and system can send the order information to responsible wine
The attendant in the 3 buildings guest rooms in shop.It should be understood that the example above is merely to illustrate, specific restriction can not be constituted.
It should be noted that above method citing is Chinese order and Chinese human-computer dialogue process, and in practical application, step
After the voice messaging of user is converted text information by rapid S102, if text information is other category of language, such as English text
This information, then matched order information is also corresponding English order template, words art template also will be English words art template, ask
Asking template also will be English inquiry template, be not discussed here.
In the above method, by obtaining the first voice messaging of user, the first voice messaging of the user includes user
Room number mark, the first text information is converted by first voice messaging, by first text information and order mould
The keyword of version is matched, and matched first order template is obtained, according to first text information and the first order mould
Version generates order information, to send the room number mark and the order information to the terminal of attendant.By upper
Method is stated, when user needs housekeeping service, the hotel Zhi Xuyu assistant's customer service system is interacted, can self-service completion housekeeping service
Order cuts operating costs to reduce the operating pressure of foreground personnel, promotes the working efficiency of hotel's logistics service, and
Since system server can handle multiple user demands simultaneously, to greatly reduce the waiting time of user, further promoted
The usage experience of user.
Fig. 6 is a kind of structural schematic diagram of hotel's customer service system provided by the present application, it will be appreciated from fig. 6 that including acquiring unit
610, conversion unit 620, matching unit 630, generation unit 640, transmission unit 650.
The acquiring unit 610 is used to obtain the first voice messaging of user, and the first voice messaging of the user includes
The room number of user identifies.
In the embodiment of the present application, the first voice messaging of the user includes that the room number of user identifies.With reference to aforementioned
Content is it is found that first voice messaging can be by obtaining for collecting the first terminal of user speech in guest room, and the
One terminal can be guestroom telephone, and user dials exchange number by dialing guestroom telephone, the call voice assistant in guestroom telephone
Obtain the first voice messaging of user;For another example, first terminal is the intelligent sound box in guest room, and user passes through specified speech phrase
Form wake up intelligent sound box after, intelligent letter box obtains the first voice messaging of user, it should be appreciated that the first terminal can be with
It is the terminal of other available user speech information, for example shape is the intelligent robot of a certain cartoon character, such as user
The application program installed on mobile phone, wechat small routine, public platform etc., the application is not especially limited.It should be noted that due to
User just may require that some housekeeping services after being generally into guest room, that is to say, that first voice messaging be with
It is obtained in the guest room at family, therefore corresponding room number can be set in the first terminal in guest room for obtaining user speech
Mark, so that the voice messaging that the first terminal obtains includes the room number mark in room locating for the first terminal, so as to
Attendant provides housekeeping service according to room number for user.
The conversion unit 620 is used to convert the first text information for first voice messaging.
In the specific implementation, after first terminal collects first voice messaging and is uploaded to director server, director server
Speech recognition module (Automatic Speech Recognition, ASR) can be by speech recognition algorithm, by a Duan Yin
Frequency becomes corresponding text, that is to say, that converts the first text information for first voice messaging, wherein the voice
Recognizer can be convolutional neural networks (Convolutional Neural Networks, CNN) algorithm, backpropagation mind
It is calculated through network (back propagation, BP) algorithm, radial base neural net (Radial Basis Function, RBF)
Method, fuzzy clustering (Fuzzy C-means, FCM) neural network algorithm, Recognition with Recurrent Neural Network (Recurrent Neural
Network, RNN) algorithm, wavelet neural network (Wavelet Neural Network, WNN) algorithm, Hidden Markov Model
(Hidden Markov Model, HMM) algorithm, orthogonal basis function opposite direction communication process neural network
(Counterpropagation Process Neural Network, CPNN) algorithm, support vector machines (Support
Vector Machine, SVM) Multiclass Classification etc., the application is not especially limited.
In the embodiment of the present application, the system also includes determination unit 660, the determination unit 660 is used for described
After converting the first text information for first voice messaging: according to first text information, determining the of the user
One interaction is intended to Consultant or Task;In the case where the first interaction of the user is intended to Consultant, by described the
One text information is matched with consulting template, is obtained first and is replied text;By output equipment to described in user output
First replies text, wherein the output equipment includes at least one or more of loudspeaker and display screen;Described
It is specifically used for unit 630: in the case where the first interaction of the user is intended to Task, by first text information
It is matched with the keyword of order template, obtains matched first order template.It should be understood that the ASR in foregoing teachings
Module, although first voice messaging is converted to the first text information, ASR module is not aware that first text
First text information can be sent to semantic understanding module (Natural by the meaning of information, therefore, ASR module
Language Understanding, NLU), NLU module is understood that the intention (Intent) of user, in the specific implementation, NLU
It can determine that the first interaction of the user is intended to information type or Task first, then distribute them to and can handle the intention
Program, wherein the program for handling the intention can be " robot (Bot) ".For example user says: " how is weather tomorrow? ",
NLU judges that user's is intended that Consultant, it is intended that classification is weather, therefore weather robot (Bot) is called out to tell that user is bright
It weather condition, if user also wonders posteriori weather, weather Bot will continue as user service, until user anticipates
Figure changes, and when being no longer that weather is intended to, then switching to other Bot is user service.
Briefly, NLU module understands that this process of the intention of user can simply be interpreted as a classifier, the classification
Device belongs to Consultant or Task by the first text information that the first voice messaging is converted into for determining, so as to following
It performs corresponding processing.Wherein, the sentence of Consultant can be " may I ask hotel eatery in a few buildings? " " nearby there is any restaurant's ratio
It is relatively nice " " there is the service for the machine of picking in hotel? " etc., the sentence of Task can be " I needs a pillow " " my needs
Change the sheet " " my TV is broken, and is needed repairing " etc., the application is not especially limited.Wherein, for determining what user was intended to
Classifier can be the disaggregated model obtained by machine learning, and specifically can be is had and task using a large amount of labels
What the text information sample training disaggregated model of type obtained, the application is not especially limited.
It, can be by by described the it should be noted that in the case where the first interaction of the user is intended to Consultant
One text information is matched with consulting template, to obtain the first answer text asked questions, wherein the consulting template
It can be the template in knowledge base, when the first text information and very high knowledge base consulting stencil matching degree, system meeting
Directly reply user's answer (unique response);If a plurality of consulting template in the first text information and knowledge base match and
When very high with degree, system can ask in reply user " you are intended to ask XX problem or XX problem " by loudspeaker or screen, from
And further determine that the intention of user, improve communication efficiency and the usage experience of user;If the first text information and knowledge
Knowledge matching degree in library is very low, system can not answer or refuse answer the problem, such as reply user " this problem I
Do not know ".It should be understood that the example above is merely to illustrate, specific restriction can not be constituted.
The matching unit 630 is obtained for matching first text information with the keyword of order template
Matched first order template.
In the embodiment of the present application, the acquiring unit 610 be also used to it is described obtain user the first voice messaging it
Before, obtain the available information on services of the attendant;The acquiring unit 610 is also used to be obtained according to the information on services
Take the keyword of the order template and the order template.It should be understood that the service that each hotel can provide be it is different,
For example, some hotels can provide laundry service, some hotels do not provide laundry service.Therefore, which is supplying
Before user's use, hotel Fang Huixian configures the keyword of oneself available service in system background.For example,
Certain hotel cannot provide laundry service, may only provide the service of replacement pillow, therefore hotel is in configure system, can
Keyword " pillow " is arranged in system background.When the first voice messaging for receiving user is " I thinks laundry clothes ", by
There is no the keyword of this respect in system background, that is to say, that hotel will return to user " to not without this service, system
Rise, hotel is temporarily without this service " this information, specifically voice can be broadcasted by the loudspeaker of first terminal and be led to
Know, also may be displayed on the screen of first terminal, the application is not especially limited.And when the first voice letter for receiving user
When breath is " I wants to change pillow ", due at this time there is this keyword from the background, that is to say, that there is this service in hotel, so that it may
It is matched to the corresponding order template of " pillow " keyword, to generate corresponding order information.
In the specific implementation, the order template can by standardizing in advance, logic reasoning or machine learning generate, open
Hair personnel can carry out type division according to the available service in multiple hotels, and each type trains an order template, each
Order template corresponds to one or more service keyword, can according to oneself hotel when hotel side carries out system configuration
The housekeeping service of offer selects keyword, so that system background obtains order template workable for the hotel.For example, Fig. 3 is this
Apply for a kind of configuration interface of the hotel assistant customer service system provided, wherein selectable keyword includes six classifications, also
It is to say, developer has trained the order template of six kinds of housekeeping services, and hotel personnel are when carrying out system configuration, if hotel is only
There is the housekeeping service of replacement pillow, then hotel side can only choose " pillow " keyword, background system will only " pillow " be closed
Key word and corresponding order template can be matched, other templates that do not choose can not be then matched.It should be understood that this
Application determines order template by keyword match, and method is simple and matching speed is fast, improves the response time of system, also,
Also very simple is convenient for the configuration step of hotel user simultaneously, improves the usage experience of user.It should be understood that the example above is only used for
Illustrate, specific restriction can not be constituted.
It should be noted that if the housekeeping service in hotel can segment type, for example the pillow in hotel is divided into " buck wheat pillow
Head " and " fiber pillow ", then hotel side is when carrying out system configuration, it is acceptable further after having selected " pillow " keyword
The attributive classification keyword of pillow is set, and the attributive classification keyword is added into order template, that is to say, that when with
When family needs replacing pillow, system can further inquire user according to attributive classification keyword, specifically need which attribute point
The pillow of class improves user experience, similarly, the attributive classification of sheet connects so that more accurate provides a user housekeeping service
Sending the vehicle classification of machine can also further set, and the application is not especially limited.
The generation unit 640 is used to generate order information according to first text information and the first order template.
In the specific implementation, the order template includes multiple slot positions (Slot), it is described according to first text information with
And the first order template, the generation unit 640 are specifically used for: according to first text information, filling first order
Multiple slot positions of template.Wherein, the slot position includes essential slot position and optionally slot position.It should be understood that the user's
In the case that first interaction is intended to Task, after system must know the Partial key information in first text information,
The task of user's publication can just be completed.For example, user says that " the pillow inner core for being intended to oneself room is replaced with buckwheat
, sleep so pleasant " so system is at least resolved to after Item Information and syntactic information be " pillow increases, 1 ",
It could generate corresponding order information, and " sleep is pleasant " this content, system, which does not parse, can also obtain order information.
Based on this, developer can carry out slot position design in system development, and core information needed for each order is designed as must
Slot position is selected, non-core information design must fill out slot position to be non-.
For example, user has bought the air ticket of 5:00 AM, and under this application scenarios, user, which may may require that, wakes clothes
Business, then must fill out slot position is that can be waken with the time if user does not express clearly for the order template of wake-up service
Time, it is what the specific wake-up time that system, which will question closely user, that is, time slot position must be filled to complete, Er Feibi
It fills out slot position and can be and wake reason, must be filled out for slot position for non-, if there is in this respect in the first text information of user
Hold, so that it may the slot position is filled, if the not content of this respect, system will not question closely user to fill the slot
Position.In the specific implementation, the slot position design of wake-up service can be as shown in table 1 below.It should be understood that the example above and table 1 are only used
In for example, specific restriction can not be constituted.
The design of 1 wake-up service slot position of table
Slot position description | Type | Field name | Citing |
Time | It must fill out | Remind_time | Tomorrow 5:00 AM |
Repetition period | It is non-to fill out | repeat | Daily |
Content | It is non-to fill out | title | It seizes the opportunity |
Alerting pattern | It is non-to fill out | Remind_type | Telephone bell |
In the embodiment of the present application, the determination unit 660 is also used to after the acquisition order information: being ordered described in determining
Whether single information is complete;In the incomplete situation of the order information, phase is obtained according to the missing content of the order information
Art template if answering generates the first inquiry text;The first inquiry text is exported to the user by output equipment;It obtains
Second voice messaging is converted the second text information by the second voice messaging of the user;According to second text
Information determines that the second interaction of the user is intended to Consultant or Task;It is intended to consult in the second interaction of the user
In the case where inquiry type, second text information is matched with consulting template, second is obtained and replies text;It is set by output
It is standby to reply text to user output described second;In the case where the second interaction of the user is intended to Task, root
The missing content of the order information is improved according to second text information;The transmission unit 650 is specifically used for: ordering described
In single complete situation of information, the room number mark and the order information are sent to the terminal of attendant.
It should be understood that only according to user the first voice messaging just complete order information probability be it is low-down,
And incomplete order information is sent to hotel staff, attendant can not also understand the specific requirements of user.Therefore,
After using the first text information filling order template, system will judge whether order template is filled completely, if endless
It is whole, it will art template inquires user if being corresponded to according to the content of missing.For example, user says " I wants to change a pillow ",
System analysis to Item Information and syntactic information be " pillow increases, 1 ", and according to the setting on hotel backstage, hotel's pillow
Attribute keyword further comprise " buckwheat pillow/fiber pillow ", art template keyword can be with if being determined according to missing content
For " product name: pillow, Attributions selection: fiber pillow/buckwheat pillow ", thus can be matched to missing content it is corresponding if art
Template: " we provide [product name] type have [Attributions selection], may I ask you need it is any? ", ask to obtain first
Ask text " we provide pillow type have fiber pillow and buckwheat pillow, may I ask you need it is any ".It is obtained in first terminal
After getting the answer of user, order is further improved according to the answer of user, the order for being ultimately sent to attendant can be
Accurately, the wish of user is completely expressed, and then greatly improves the treatment effeciency of attendant, that improves user uses body
It tests.
In the specific implementation, the determination unit 660 is specifically used for: whether determine the first order template must fill out slot position
Filling is complete;The determination unit is specifically used for: the slot position lacked according to the order information obtains art template if corresponding to,
Wherein, the words art template is generated by specification, logic reasoning or machine learning in advance, each slot of the order information
One or more corresponding words art template of position, the words art template are used to generate phase according to the slot position of the missing of the order information
The inquiry text answered;The determination unit 660 is also used to: according to second text information, filling the slot of the order template
Position.That is, whether order information completely determine according to must fill out slot position and fill to finish, when system can not basis
When slot position is filled complete by the first text information, it will questioned closely according to the slot position that must fill out of missing to user, until order mould
The slot position that must fill out of version is stuffed entirely with finishing.Wherein, must each fill out slot position can correspond to one or more inquiry template, system
After what determination lacked must fill out slot position, user is inquired according to corresponding inquiry template, specifically, can be broadcast by loudspeaker
Report the first inquiry text, can also be by showing that screen shows the first inquiry text.The application is not especially limited.In this way,
The method that must fill out slot position is filled by cross-question user, greatly improves the order information for being ultimately sent to attendant
Integrality improves the treatment effeciency of hotel guest room service.For example, it is still exemplified by Table 1, it is assumed that first text information is
" I will catch up with aircraft tomorrow, I needs wake-up service ", slot position at this time can only fill " content " this non-slot position that must be filled out, and must
Filling out slot position " time " can not be filled, and system will question closely user about the content that must fill out slot position " time ", for example, system can be with
By the way that loudspeaker plays that first inquires text " you need me to wake you in the several points? ", thus the second text answered according to user
The order information of information " 5 points of morning " generation " wake-up service ".It should be understood that the example above is merely to illustrate, tool can not be constituted
Body limits.
It should be noted that system receives the voice messaging of user for the first time, that is, after the first voice messaging, it will complete
The interactive process of whole record system and user, that is to say, that record user believes from whole voices after the first voice messaging
Breath, so that dialogue later can all associate with conversation content before, when system passes through the first inquiry of output equipment output
After text, if user does not answer the inquiry, but the problem of asked a Consultant, then system can be first according to knowing
The consulting template known in library finds corresponding answer, if being matched to corresponding consulting template, will answer the Consultant problem,
Then the intersection record before basis comes again the first inquiry text, to meet the various aspects need of user to the greatest extent
It asks.
In the embodiment of the present application, the determination unit 660 is also used to defeated to the user by output equipment described
Out after the second answer text: exporting the first inquiry text to the user by output equipment;Obtain the use
The third voice messaging is converted third text information by the third voice messaging at family;According to the third text information, really
The third interaction of the fixed user is intended to Consultant or Task;It is intended to the feelings of Consultant in the third interaction of the user
Under condition, the third text information is matched with consulting template, third is obtained and replies text;By output equipment to described
After user exports the third answer text, text is inquired to user output described first again by output equipment;?
In the case that the third interaction of the user is intended to Task, the order information is improved according to the third text information
Missing content.With reference to foregoing teachings it is found that NLU module is determining that first text information belongs to Consultant or Task
Afterwards, which can be distributed to the Bot that can handle the information, that is to say, that order letter is generated in step S104
The all processes of breath are all to be completed by handling the Bot of the category information, including fill template, determine template integrality, Yi Jisheng
At corresponding answer text etc., until generating last complete order information.Fig. 4 is a kind of acquisition user provided by the present application
The first text information after, the Bot for handling first text information obtains the entire flow figure of order information, in simple terms,
The voice messaging for the user that first terminal obtains every time, system can be all intended to by user of the NLU module to this voice messaging into
Row classification, determines Consultant or Task, performs corresponding processing further according to type, intert Consultant if it is Task,
Consultant is so answered first, repeats Task, until order is completed, the specific processing mode of Task and consulting
The specific processing mode of type specifically refers to foregoing teachings, and which is not described herein again.
For example, still by taking the wake-up service template of table 1 as an example, Fig. 5 is a kind of user provided by the present application and system
Session log schematic diagram, system output first inquiry text " you need me to wake you in the several points? " later, system obtain the
Two text informations are " is the dining room in your hotels in a few buildings? ", system can first determine whether that second text information is to belong at this time
Consultant or Task are found according to the consulting template in knowledge base first after determining that second text information is Consultant
Corresponding answer, and the second answer text is exported, for example, saying that " our hotel eatery is two to user by loudspeaker
Building ", then again by loudspeaker repeat output first inquiry text " you need me to wake you in the several points? ", thus not delaying
While user subscribes housekeeping service, answer user wants the problem of asking as far as possible, promotes the usage experience of user.It should be understood that
The dialogue of system and user can be pure voice dialogue, be also possible to plain text dialogue as shown in Figure 5, can also be voice with
Text combination, conversational mode of the video in conjunction with voice, for example directly broadcasted in the dialogue in Fig. 5 to user speech by phone
Hold, the customer service image in a hotel is directly displayed by television set, bound fraction limb action is broadcasted in Fig. 5 to user speech
Conversation content, the application are not especially limited.
In the specific implementation, complete order information can also be corresponding with words art template, for example, complete order information can be with
Be " Item Title: pillow, type: fiber pillow, quantity: 1 ", can also be " Item Title: compromising, quantity: 2 ", it can be with
It is " service name: wake-up service, time: 5 points of morning ", art template can be that " good, you, which have clicked, [orders at this time
Single information], if confirmation? ", system can be in conjunction with words art template and order information, and generating final user, to reply words art " good
, you have clicked pillow, and type is fiber pillow, quantity 1, slippers, quantity 2, wake-up service, 5 points of morning, if really
Recognize? ", order situation is further finally improved according to the confirmation situation of user, promotes the accuracy rate of order to the greatest extent, is improved
Housekeeping service efficiency and the satisfaction of user.
The transmission unit 650 is used to send the address mark and order information to the terminal of attendant.
In the embodiment of the present application, after attendant receives complete order information and room number mark, Ji Kegen
According to order contents complete user needed for housekeeping service, wherein the order information that attendant receives can be text information,
Voice broadcast order information etc., the application is not especially limited, and optionally, system can be indicated according to room number and order
The order information is sent to the attendant that can handle the order information by information, for example, order information be " Item Title:
Pillow, type: fiber pillow, quantity: 1 ", room identification is rooms 301, and system can send the order information to responsible wine
The attendant in the 3 buildings guest rooms in shop.It should be understood that the example above is merely to illustrate, specific restriction can not be constituted.
It should be noted that above method citing is Chinese order and Chinese human-computer dialogue process, and in practical application, step
After the voice messaging of user is converted text information by rapid S102, if text information is other category of language, such as English text
This information, then matched order information is also corresponding English order template, words art template also will be English words art template, ask
Asking template also will be English inquiry template, be not discussed here.
In above system, by obtaining the first voice messaging of user, the first voice messaging of the user includes user
Room number mark, the first text information is converted by first voice messaging, by first text information and order mould
The keyword of version is matched, and matched first order template is obtained, according to first text information and the first order mould
Version generates order information, to send the room number mark and the order information to the terminal of attendant.By upper
System is stated, when user needs housekeeping service, the hotel Zhi Xuyu assistant's customer service system is interacted, can self-service completion housekeeping service
Order cuts operating costs to reduce the operating pressure of foreground personnel, promotes the working efficiency of hotel's logistics service, and
Since system server can handle multiple user demands simultaneously, to greatly reduce the waiting time of user, further promoted
The usage experience of user.
It is a kind of structural schematic diagram of server provided herein referring to Fig. 7, Fig. 7.The server can be in Fig. 1
Director server, wherein data processing method provided by the present application can realize in cloud service cluster as shown in Figure 7,
It can be realized in single computer node and memory node, the application is not especially limited, and the cloud service cluster includes extremely
A few calculate node 710 and at least one memory node 720.
Calculate node 710 includes one or more processors 711, communication interface 712 and memory 713.Wherein, processor
711, it can be connected by bus 714 between communication interface 712 and memory 713.
Processor 711 includes one or more general processor, wherein general processor, which can be, is capable of handling electronics
Any kind of equipment of instruction, including it is central processing unit (Central Processing Unit, CPU), microprocessor, micro-
Controller, primary processor, controller and ASIC (Application Specific Integrated Circuit, dedicated collection
At circuit) etc..It can be only for the application specific processor of calculate node 710 or can be total to other calculate nodes 710
It enjoys.Processor 711 executes various types of stored digital instructions, such as the software or firmware journey that are stored in memory 713
Sequence, it can make calculate node 710 provide wider a variety of services.For example, processor 711 is able to carry out speech recognition, semantic reason
The code of the modules such as solution and order generation, to execute at least part for the method being discussed herein.
Communication interface 712 can be wireline interface (such as Ethernet interface), for other calculate nodes or user into
Row communication.When communication interface 712 is wireline interface, communication interface 712 can use the protocol suite on TCP/IP, for example,
RAAS agreement, remote function calls (Remote Function Call, RFC) agreement, Simple Object Access Protocol (Simple
Object Access Protocol, SOAP) agreement, Simple Network Management Protocol (Simple Network Management
Protocol, SNMP) agreement, Common Object Request Broker Architecture (Common Object Request Broker
Architecture, CORBA) agreement and distributed protocol etc..
Memory 713 may include volatile memory (Volatile Memory), such as random access memory
(Random Access Memory, RAM);Memory also may include nonvolatile memory (Non-Volatile
), such as read-only memory (Read-Only Memory, ROM), flash memory (Flash Memory), hard disk Memory
(Hard Disk Drive, HDD) or solid state hard disk (Solid-State Drive, SSD) memory can also include above-mentioned kind
The combination of the memory of class.
Memory node 720 includes one or more storage controls 721 and storage array 722.Wherein, storage control
It can be connected by bus 724 between 721 and storage array 722.
Storage control 721 includes one or more general processor, wherein general processor, which can be, to be capable of handling
Any kind of equipment, including CPU, microprocessor, microcontroller, primary processor, controller and ASIC of e-command etc.
Deng.It can be only for single memory node 720 application specific processor or can be with calculate node 710 or other storages
Node 720 is shared.It is appreciated that in the present embodiment, each memory node includes a storage control, implement in others
In example, a storage control can also be shared with multiple memory nodes, be not especially limited herein.
Memory array 722 may include multiple memories 723.Memory 723 can be nonvolatile memory, such as
ROM, flash memory, HDD or SSD memory can also include the combination of the memory of mentioned kind.For example, storage array can
To be made of multiple HDD or multiple SDD, it is made of alternatively, storage array can be HDD and SDD.Wherein, Duo Gecun
Reservoir storage control 721 by assistance under be combined in different ways and form memory group, to provide than single
The higher storage performance of memory and offer technology of data copy.Optionally, memory array 723 may include one or more
A data center.Multiple data centers, which can be set, to be in the same localities, alternatively, do not make to have respectively in different places herein
Body limits.Memory array 723 can store program code and program data.Wherein, program code includes speech recognition
Block code, semantic understanding block code and order production block code etc..Program data include: order template data,
Template data, words art template data etc. are seeked advice from, can also include order template, consulting template, words art template and other languages
The training sample set data of sound model or template, for using the training sample set data generator number for processor 711
Order template, consulting template and words art template in etc..
In the above-described embodiments, can come wholly or partly by software, hardware, firmware or any combination thereof real
It is existing.When implemented in software, it can entirely or partly realize in the form of a computer program product.The computer program
Product includes one or more computer instructions.When loading on computers and executing the computer program instructions, all or
It partly generates according to process or function described in the embodiment of the present application.The computer can be general purpose computer, dedicated meter
Calculation machine, computer network or other programmable devices.The computer instruction can store in computer readable storage medium
In, or from a computer readable storage medium to the transmission of another computer readable storage medium, for example, the computer
Instruction can pass through wired (such as coaxial cable, optical fiber, number from a web-site, computer, server or data center
User's line) or wireless (such as infrared, wireless, microwave etc.) mode into another web-site, computer, server or data
The heart is transmitted.The computer readable storage medium can be any usable medium or include that computer can access
The data storage devices such as one or more usable mediums integrated server, data center.The usable medium can be magnetism
Medium, (for example, floppy disk, storage dish, tape), optical medium (for example, DVD) or semiconductor medium (such as solid state memory disc
Solid State Disk (SSD)) etc..
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the program can be stored in a computer-readable storage medium
In, the program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein, the storage medium can be magnetic
Dish, CD, read-only memory (Read-Only Memory, ROM) or random access memory (Random Access
Memory, RAM) etc..
The foregoing is merely several embodiments of the application, those skilled in the art is according to can be with disclosed in application documents
Various changes or modifications are carried out without departing from spirit and scope to the application.
Claims (10)
1. a kind of method of data processing, which is characterized in that the described method includes:
Obtain the first voice messaging of user, wherein the first voice messaging of the user includes that the room number of user identifies;
The first text information is converted by first voice messaging;
First text information is matched with the keyword of order template, obtains matched first order template, wherein
The order template is used to generate order information according to text information;
According to first text information and the first order template, order information is generated;
The room number mark and the order information are sent to the terminal of attendant.
2. the method according to claim 1, wherein described convert the first text for first voice messaging
After information, the method also includes:
According to first text information, determine that the first interaction of the user is intended to Consultant or Task;
In the case where the first interaction of the user is intended to Consultant, first text information and consulting template are carried out
Matching obtains first and replies text;
Described first is exported to the user by output equipment and replies text, wherein the output equipment includes at least loudspeaking
One or more of device and display screen;
It is described to match first text information with the keyword of order template, obtain matched first order template packet
It includes:
In the case where the first interaction of the user is intended to Task, by the pass of first text information and order template
Key word is matched, and matched first order template is obtained.
3. according to the method described in claim 2, it is characterized in that, after the acquisition order information, the method also includes:
Determine whether the order information is complete;
In the incomplete situation of the order information, accordingly art mould is obtained according to the missing content of the order information
Version generates the first inquiry text;
The first inquiry text is exported to the user by output equipment;
Second voice messaging is converted the second text information by the second voice messaging for obtaining the user;
According to second text information, determine that the second interaction of the user is intended to Consultant or Task;
In the case where the second interaction of the user is intended to Consultant, second text information and consulting template are carried out
Matching obtains second and replies text;
Described second, which is exported, to the user by output equipment replies text;
In the case where the second interaction of the user is intended to Task, the order is improved according to second text information
The missing content of information;
It is described to include: to the terminal of the attendant transmission room number mark and the order information
In the complete situation of the order information, the room number mark and the order are sent to the terminal of attendant
Information.
4. according to the method described in claim 3, it is characterized in that, described export described the to the user by output equipment
After two reply text, the method also includes:
The first inquiry text is exported to the user by output equipment;
The third voice messaging is converted third text information by the third voice messaging for obtaining the user;
According to the third text information, determine that the third interaction of the user is intended to Consultant or Task;
In the case where the interaction of the third of the user is intended to Consultant, the third text information and consulting template are carried out
Matching obtains third and replies text;
It is defeated to the user again by output equipment after exporting the third answer text to the user by output equipment
The first inquiry text out;
In the case where the interaction of the third of the user is intended to Task, the order is improved according to the third text information
The missing content of information.
5. the method according to claim 1, wherein it is described obtain user the first voice messaging before, it is described
Method further include:
Obtain the available information on services of the attendant;
According to the information on services, the keyword of the order template and the order template is obtained.
6. 5 any method according to claim 1, which is characterized in that the order template passes through specification, logic in advance
Inference method or machine learning generate, and the order template includes multiple slot positions,
It is described according to first text information and the first order template, generating order information includes:
According to first text information, multiple slot positions of the first order template are filled;
The slot position includes that must fill out slot position and slot position is filled out in choosing, the determination order information whether completely include:
Determine the first order template must fill out whether slot position fills completely;
It is described that according to the acquisition of the missing content of the order information, accordingly art template includes:
According to the slot position that the order information lacks, art template if corresponding to is obtained, wherein the words art template by advising in advance
Model, logic reasoning or machine learning generate, and each slot position of the order information corresponds to what one or more said art template,
The words art template is used to generate corresponding inquiry text according to the slot position of the missing of the order information;
The missing content for improving the order information according to second text information includes:
According to second text information, the slot position of the order template is filled.
7. a kind of hotel assistant customer service system, which is characterized in that including acquiring unit, conversion unit, matching unit, generation unit
And transmission unit, wherein
The acquiring unit is used to obtain the first voice messaging of user, and the first voice messaging of the user includes the room of user
Between number mark;
The conversion unit is used to convert the first text information for first voice messaging;
The matching unit is for matching first text information with the keyword of order template, acquisition matched the
One order template;
The generation unit is used to generate order information according to first text information and the first order template;
The transmission unit is used to send the address mark and order information to the terminal of attendant.
8. a kind of server, which is characterized in that including processor and memory;The memory is for storing instruction, described
Processor executes when executing described instruction as claimed in any one of claims 1 to 6 for executing described instruction, the processor
Method.
9. a kind of computer non-transitory storage media, which is characterized in that the computer non-transitory storage media is stored with calculating
Machine program, which is characterized in that realize when the computer program is executed by calculating equipment such as any one of claims 1 to 6 institute
The method stated.
10. a kind of computer program product, which is characterized in that when the computer program product is read and executed by a computer
When, realize such as method described in any one of claims 1 to 6.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910410960.8A CN110211001A (en) | 2019-05-17 | 2019-05-17 | A kind of hotel assistant customer service system, data processing method and relevant device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910410960.8A CN110211001A (en) | 2019-05-17 | 2019-05-17 | A kind of hotel assistant customer service system, data processing method and relevant device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110211001A true CN110211001A (en) | 2019-09-06 |
Family
ID=67787625
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910410960.8A Pending CN110211001A (en) | 2019-05-17 | 2019-05-17 | A kind of hotel assistant customer service system, data processing method and relevant device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110211001A (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111143565A (en) * | 2019-12-29 | 2020-05-12 | 杭州睿沃科技有限公司 | K-means self-learning system |
CN111160318A (en) * | 2020-01-07 | 2020-05-15 | 百度在线网络技术(北京)有限公司 | Electronic equipment control method and device |
CN111510561A (en) * | 2020-06-18 | 2020-08-07 | 北京云迹科技有限公司 | Method, device, storage medium and equipment for distributing task requests |
CN112199498A (en) * | 2020-09-27 | 2021-01-08 | 中国建设银行股份有限公司 | Man-machine conversation method, device, medium and electronic equipment for endowment service |
CN112286366A (en) * | 2020-12-30 | 2021-01-29 | 北京百度网讯科技有限公司 | Method, apparatus, device and medium for human-computer interaction |
CN113705943A (en) * | 2020-05-22 | 2021-11-26 | 纬创资通股份有限公司 | Task management method and system based on voice talkback function and mobile device |
CN113743631A (en) * | 2021-09-13 | 2021-12-03 | 广州任行信息科技有限公司 | Overseas guest room order data storage method and system |
CN115412864A (en) * | 2022-09-01 | 2022-11-29 | 上海美来客数字科技有限公司 | System for pushing message to appointed personnel for talkback broadcasting and responding based on 4G network |
TWI823651B (en) * | 2022-10-27 | 2023-11-21 | 犀動智能科技股份有限公司 | Intelligent message push system and smart message push method |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107423275A (en) * | 2017-06-27 | 2017-12-01 | 北京小度信息科技有限公司 | Sequence information generation method and device |
CN107886948A (en) * | 2017-11-16 | 2018-04-06 | 百度在线网络技术(北京)有限公司 | Voice interactive method and device, terminal, server and readable storage medium storing program for executing |
CN108446322A (en) * | 2018-02-10 | 2018-08-24 | 灯塔财经信息有限公司 | A kind of implementation method and device of intelligent Answer System |
CN109165274A (en) * | 2018-07-16 | 2019-01-08 | 厦门快商通信息技术有限公司 | A kind of on-line automatic client service method, system and electronic equipment |
-
2019
- 2019-05-17 CN CN201910410960.8A patent/CN110211001A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107423275A (en) * | 2017-06-27 | 2017-12-01 | 北京小度信息科技有限公司 | Sequence information generation method and device |
CN107886948A (en) * | 2017-11-16 | 2018-04-06 | 百度在线网络技术(北京)有限公司 | Voice interactive method and device, terminal, server and readable storage medium storing program for executing |
CN108446322A (en) * | 2018-02-10 | 2018-08-24 | 灯塔财经信息有限公司 | A kind of implementation method and device of intelligent Answer System |
CN109165274A (en) * | 2018-07-16 | 2019-01-08 | 厦门快商通信息技术有限公司 | A kind of on-line automatic client service method, system and electronic equipment |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111143565A (en) * | 2019-12-29 | 2020-05-12 | 杭州睿沃科技有限公司 | K-means self-learning system |
CN111160318A (en) * | 2020-01-07 | 2020-05-15 | 百度在线网络技术(北京)有限公司 | Electronic equipment control method and device |
CN113705943A (en) * | 2020-05-22 | 2021-11-26 | 纬创资通股份有限公司 | Task management method and system based on voice talkback function and mobile device |
CN113705943B (en) * | 2020-05-22 | 2024-04-26 | 纬创资通股份有限公司 | Task management method and system based on voice intercom function and mobile device |
CN111510561A (en) * | 2020-06-18 | 2020-08-07 | 北京云迹科技有限公司 | Method, device, storage medium and equipment for distributing task requests |
CN112199498A (en) * | 2020-09-27 | 2021-01-08 | 中国建设银行股份有限公司 | Man-machine conversation method, device, medium and electronic equipment for endowment service |
CN112286366A (en) * | 2020-12-30 | 2021-01-29 | 北京百度网讯科技有限公司 | Method, apparatus, device and medium for human-computer interaction |
CN113743631A (en) * | 2021-09-13 | 2021-12-03 | 广州任行信息科技有限公司 | Overseas guest room order data storage method and system |
CN115412864A (en) * | 2022-09-01 | 2022-11-29 | 上海美来客数字科技有限公司 | System for pushing message to appointed personnel for talkback broadcasting and responding based on 4G network |
TWI823651B (en) * | 2022-10-27 | 2023-11-21 | 犀動智能科技股份有限公司 | Intelligent message push system and smart message push method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110211001A (en) | A kind of hotel assistant customer service system, data processing method and relevant device | |
CN106647586B (en) | A kind of virtual computer room visual control management system and method based on B/S framework | |
CN110235154B (en) | Associating meetings with items using feature keywords | |
CN106777135B (en) | Service scheduling method, device and robot service system | |
WO2018036555A1 (en) | Session processing method and apparatus | |
CN107977236B (en) | Question-answering system generation method, terminal device, storage medium and question-answering system | |
EP3360313B1 (en) | Feedback controller for data transmissions | |
CN108513705A (en) | Selective sensor poll | |
WO2018067368A1 (en) | Hierarchical annotation of dialog acts | |
KR102137818B1 (en) | Smart ordering system and method using artificial intelligence | |
CN104813311A (en) | System and methods for virtual agent recommendation for multiple persons | |
CN110083706A (en) | A kind of internet of things oriented micro services frame and its service combining method | |
CN110084616A (en) | Intelligence pays a return visit method, apparatus, computer installation and storage medium | |
CN109410948A (en) | Communication means, device, system, computer equipment and readable storage medium storing program for executing | |
CN110297896A (en) | A kind of application method and intelligent customer service system of intelligent customer service platform | |
CN114721833A (en) | Intelligent cloud coordination method and device based on platform service type | |
WO2022048648A1 (en) | Method and apparatus for achieving automatic model construction, electronic device, and storage medium | |
CN104869048A (en) | Grouping processing method, device and system of MicroBlog data | |
CN108306851A (en) | Information acquisition method, providing method, apparatus and system | |
Schneider et al. | Feedback in context: Supporting the evolution of IT-ecosystems | |
CN115982331A (en) | Information interaction method, device and equipment in session scene | |
CN110096516B (en) | User-defined database interaction dialog generation method and system | |
CN111026773A (en) | Information processing method, information processing device, electronic equipment and storage medium | |
CN109710747A (en) | Information processing method, device and electronic equipment | |
CN106325515B (en) | A kind of service-oriented man-machine interactive system and implementation method |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190906 |
|
RJ01 | Rejection of invention patent application after publication |