CN109543010A - The interactive method and system of fused data library inquiry - Google Patents
The interactive method and system of fused data library inquiry Download PDFInfo
- Publication number
- CN109543010A CN109543010A CN201811231195.5A CN201811231195A CN109543010A CN 109543010 A CN109543010 A CN 109543010A CN 201811231195 A CN201811231195 A CN 201811231195A CN 109543010 A CN109543010 A CN 109543010A
- Authority
- CN
- China
- Prior art keywords
- value
- text
- dialogue
- slot
- current state
- 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
Abstract
The embodiment of the present invention provides the interactive method and system of a kind of fused data library inquiry, which comprises carries out natural language understanding to the dialogue of the user got text above, acquisition understands result information；Result information is understood according to described, updates the current state information of dialogue；If judgement is known in the current state information i.e. comprising intention value, it again include field value, but the slot value of all slots is not included, the maximum slot of entropy is then obtained from presetting database, as target slot, and according to the target slot, generating dialogue, hereafter text, the slot value of the target slot are not included in the current state information；Talk with hereafter text generation natural language according to described, and exports the natural language.The interactive method and system of fused data library inquiry provided in an embodiment of the present invention faster complete conversation tasks by the entropy information of database bracket groove, shorten dialog turns, improve human-computer dialogue efficiency, keep interactive system more intelligent.
Description
Technical field
The present embodiments relate to interactive field more particularly to a kind of human-computer dialogues of fused data library inquiry
Method and system.
Background technique
Task human-computer dialogue is intended to want interactive system acquisition with user man-machine by interacting with user
The relevant slot information of task that conversational system executes, and user is finally made to obtain the result information that the task is performed.It is man-machine
Conversational system executes the round that a task needs to interact with user, reflects the efficiency and intelligence of interactive system, holds
The round that the same task of row needs to interact with user is fewer, and the efficiency of interactive system is higher, more intelligent.
In the prior art, the mode for database being inquired in interactive system includes following three classes: one, by data base querying
The movement produced as system is acted, api interface is called；Two, it in conjunction with the probability of slot value single in the dialog information of user, obtains
The probability that each records into database, is ranked up according to the probability, and the final several records of sequence before examination of choosing return
Back to user；Three, using the method from sequence to sequence, session log is read, the retrieval language of database is generated using decoder
Sentence.
But the interactive system of the mode based on inquiry database in the prior art, the task that executes are needed and are used
The round of family interaction is too many, and the low efficiency of interactive system is not smart enough.
Summary of the invention
A kind of overcome the above problem the purpose of the embodiment of the present invention is that providing or at least be partially solved the above problem
The interactive method and system of fused data library inquiry.
In order to solve the above-mentioned technical problem, on the one hand, the embodiment of the present invention provides a kind of the man-machine of fused data library inquiry
Dialogue method, comprising:
Natural language understanding is carried out to the dialogue of the user got text above, acquisition understands result information, the reason
Solving result information includes one or more of null value, field value, intention value and slot value；
Result information is understood according to described, updates the current state information of dialogue, and the current state information is currently to obtain
The default key message relevant to conversation tasks got, the default key message relevant to conversation tasks include field,
It is intended to and several slots；
If judgement is known i.e. comprising intention value in the current state information, and includes field value, but do not include all
The slot value of slot then obtains the maximum slot of entropy from presetting database, as target slot, and according to the target slot, generation pair
Hereafter text, the slot value of the target slot are not included in the current state information words；
Talk with hereafter text generation natural language according to described, and exports the natural language.
On the other hand, the embodiment of the present invention provides a kind of interactive system of fused data library inquiry, comprising:
Natural language understanding module carries out natural language understanding for the dialogue text above to the user got, obtains
It takes and understands result information, it is described to understand that result information includes one or more of null value, field value, intention value and slot value；
State tracking module updates the current state information of dialogue for understanding result information according to described, described current
Status information is the default key message relevant to conversation tasks currently got, the default pass relevant to conversation tasks
Key information includes field, intention and several slots；
Decision making module, if for judging to know i.e. comprising intention value in the current state information, and include field
Value, but do not include the slot value of all slots, then the maximum slot of entropy is obtained from presetting database, as target slot, and according to
The target slot, generating dialogue, hereafter text, the slot value of the target slot are not included in the current state information；
Spatial term module for talking with hereafter text generation natural language according to described, and exports the nature
Language.
In another aspect, the embodiment of the present invention provides a kind of electronic equipment, comprising:
Memory and processor, the processor and the memory complete mutual communication by bus；It is described to deposit
Reservoir is stored with the program instruction that can be executed by the processor, and it is above-mentioned that the processor calls described program instruction to be able to carry out
Method.
Another aspect, the embodiment of the present invention provide a kind of non-transient computer readable storage medium, are stored thereon with calculating
Machine program realizes above-mentioned method when the computer program is executed by processor.
The interactive method and system of fused data library inquiry provided in an embodiment of the present invention, by database bracket groove
Entropy information faster completes conversation tasks, shortens dialog turns, improves human-computer dialogue efficiency, makes interactive system more
Intelligence.
Detailed description of the invention
Fig. 1 is the interactive method schematic diagram of fused data library inquiry provided in an embodiment of the present invention；
Fig. 2 is the interactive system schematic diagram of fused data library inquiry provided in an embodiment of the present invention；
Fig. 3 is the structural schematic diagram of electronic equipment provided in an embodiment of the present invention.
Specific embodiment
In order to keep the purposes, technical schemes and advantages of the embodiment of the present invention clearer, implement below in conjunction with the present invention
Attached drawing in example, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment
It is a part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiment of the present invention, those of ordinary skill in the art
Every other embodiment obtained without making creative work, shall fall within the protection scope of the present invention.
Fig. 1 is the interactive method schematic diagram of fused data library inquiry provided in an embodiment of the present invention, as shown in Figure 1,
The embodiment of the present invention provides a kind of interactive method of fused data library inquiry, this method comprises:
Step S101, natural language understanding is carried out to the dialogue of the user got text above, acquisition understands that result is believed
Breath, it is described to understand that result information includes one or more of null value, field value, intention value and slot value；
Step S102, result information is understood according to described, update the current state information of dialogue, the current state information
For the default key message relevant to conversation tasks currently got, the default key message packet relevant to conversation tasks
Include field, intention and several slots；
If step S103, judgement is known i.e. comprising intention value in the current state information, and includes field value, but do not wrap
Slot value containing all slots then obtains the maximum slot of entropy, as target slot, and according to the target from presetting database
Slot, generating dialogue, hereafter text, the slot value of the target slot are not included in the current state information；
Step S104, talk with hereafter text generation natural language according to described, and export the natural language.
Specifically, system obtains the dialogue text above of user first, to the dialogue of the user got text above
Natural language understanding is carried out, acquisition understands result information, understands that result information includes in null value, field value, intention value and slot value
One or more.
By taking one is ordered the Task man-machine interactive system of train ticket as an example, default key message packet relevant to conversation tasks
Including field, intention and several slots, field value is train ticket, it is intended that value may include booking tickets, return ticket and changing label etc., several slots
It may include five slots, this five slots are respectively to originate geosyncline, purpose geosyncline, departure time slot, train type slot and budget gold
Volume slot.Only when the default key message relevant to conversation tasks that system is got can satisfy when needing of execution task,
System could execute corresponding task, and otherwise, system, with user's interaction, will obtain as quickly as possible and appoint with dialogue by constantly
It is engaged in relevant default key message.
When the dialogue of the user got text above is " hello ", system is by the natural language understanding to the text
Later, the information obtained is a greeting term " hello ", since greeting term " hello " is not and conversation tasks phase
The default key message closed, therefore, the understanding result information finally obtained are as follows: null value, null value indicate above from the dialogue of user
Default key message relevant to conversation tasks is not got in text.
When the dialogue of the user got text above is " hello, me can be helped to book tickets ", system is passed through to this article
After this natural language understanding, the information obtained is " ticket booking ", since " ticket booking " is default key relevant to conversation tasks
Information, also, " ticket booking " is able to reflect out the intention of user, therefore, the understanding result information finally obtained are as follows: intention value is to order
Ticket.
When the dialogue of the user got text above is " me please be helped to order a train ticket ", system is passed through to the text
Natural language understanding after, the information that obtains is " ordering " and " train ticket ", since " ordering " and " train ticket " is and conversation tasks
Relevant default key message, also, " ordering " and " train ticket " is able to reflect out the intention of user, and " train ticket " is able to reflect out
Field relevant to conversation tasks, therefore, the understanding result information finally obtained are as follows: field value is train ticket, it is intended that be worth to order
Ticket.
When the dialogue of the user got text above is " me please be helped to order the train ticket of a tomorrow ", system process pair
After the natural language understanding of the text, the information that obtains is " ordering ", " tomorrow " and " train ticket ", due to " ordering ", " tomorrow " and
" train ticket " is default key message relevant to conversation tasks, also, " ordering " and " train ticket " is able to reflect out the meaning of user
Figure, " train ticket " is able to reflect out field relevant to conversation tasks, and " ordering ", " tomorrow " and " train ticket " is able to reflect out user
Departure time, therefore, the understanding result information finally obtained are as follows: field value is train ticket, it is intended that be worth for ticket booking, the departure time
Slot value is tomorrow.
After obtaining the understanding result information for talking with text above, according to result information is understood, the current shape of dialogue is updated
State information, current state information is the default key message relevant to conversation tasks currently got, related to conversation tasks
Default key message include field, be intended to and several slots.
Still by taking said one orders the Task man-machine interactive system of train ticket as an example, default key relevant to conversation tasks
Information include field, be intended to and several slots, field value be train ticket, it is intended that value may include ticket booking, return ticket and change label etc., if
A dry slot may include five slots, this five slots be respectively originate geosyncline, purpose geosyncline, departure time slot, train type slot and
Budget amount slot.Only when the default key message relevant to conversation tasks that system is got can satisfy the need of execution task
When wanting, system could execute corresponding task, otherwise, system will by constantly with user interaction, as quickly as possible obtain with
The relevant default key message of conversation tasks.
Before the dialogue text above for obtaining user, the current state information of dialogue are as follows: having got field value is
Train ticket, it is intended that be worth for ticket booking.When the dialogue of the user got text above is " me please be helped to order the train ticket of a tomorrow "
When, natural language understanding, the understanding result information got are carried out to the text are as follows: departure time slot value is tomorrow, then will be right
The current state information of words updates are as follows: having got field value is train ticket, it is intended that is worth to book tickets, departure time slot value is bright
It.
After having updated the current state information of dialogue, if judgement is known i.e. comprising intention value in preceding status information, again
Comprising field value, but do not include the slot value of all slots, then the maximum slot of entropy is obtained from presetting database, as target
Slot, and according to target slot, dialogue hereafter text is generated, before the slot value of target slot is not included in status information.
Still by taking said one orders the Task man-machine interactive system of train ticket as an example, when the dialogue text above for obtaining user
Before, the current state information of dialogue are as follows: having got field value is train ticket, it is intended that be worth for ticket booking.When the use got
When the dialogue at family text above is " me please be helped to order the train ticket of a tomorrow ", natural language understanding is carried out to the text, is obtained
The understanding result information arrived are as follows: departure time slot value is tomorrow, then updates the current state information of dialogue are as follows: got
Field value is train ticket, it is intended that is worth to book tickets, departure time slot value is tomorrow.
In this case, intention value is contained in status information before system judgement is known, that is, intention value is ticket booking, again
Contain field value train ticket, that is, field value is train ticket, further comprises departure time slot value, that is, departure time slot value is bright
It does not still include the slot value of train type slot and budget amount slot in preceding status information, system is believed according to the preceding state of dialogue
Breath, can not also execute ticket booking task.Since the type of train includes high-speed rail, direct train, express train, rapid and common
Train etc., the type with train include high-speed rail, direct train, express train, rapid and five kinds of ordinary train, user's selection
The probability of the train of any type is 1/5, and the entropy according to the available train type slot of the calculation formula of comentropy is
0.7, budget amount needs are obtained from the dialog text of user, are not classified, available according to the calculation formula of comentropy
The entropy of budget amount slot is 0, then system will select train type slot from presetting database, as target slot, and according to fire
Vehicle class type groove generates dialogue hereafter text, and generating dialogue, hereafter text has got dialog strategy later.
Finally, according to hereafter text generation natural language is talked with, and the natural language is exported, complete the interaction of epicycle.
The interactive method of fused data library inquiry provided in an embodiment of the present invention is believed by the entropy of database bracket groove
Breath, faster completes conversation tasks, shortens dialog turns, improve human-computer dialogue efficiency, keep interactive system more intelligent.
On the basis of the above embodiments, further, the dialogue of the described couple of user got text above carries out certainly
Right language understanding, acquisition understand before result information, further includes:
Speech recognition is carried out to the voice data of the user got, obtains the dialogue text above.
Specifically, the object of processing is that the dialogue of the user got is literary above when system carries out natural language understanding
This can directly obtain the dialogue of user if system obtains the dialog information of user by input equipments such as mouse, keyboards
Text above.
If system by voice capture device acquire be user voice dialog information, the use for needing will acquire
The voice data at family carries out speech recognition, obtains the dialogue text above of user.
The interactive method of fused data library inquiry provided in an embodiment of the present invention is believed by the entropy of database bracket groove
Breath, faster completes conversation tasks, shortens dialog turns, improve human-computer dialogue efficiency, keep interactive system more intelligent.
On the basis of the above various embodiments, further, the dialogue of the described couple of user got text above is carried out
Natural language understanding specifically includes:
Word segmentation processing is carried out to the dialogue text above, obtains several words；
Obtain the markup information of each word；
According to the markup information of each word, field value, intention value and the slot value for including in the dialogue text above are identified.
Specifically, to the dialogue of the user got text progress natural language understanding above, specific step is as follows:
The dialogue text above for the user that will acquire by preset algorithm model carries out word segmentation processing, obtains several
Then word obtains the markup information of each word, and according to the markup information of each word, identification, which is talked in text above, includes
Field value, intention value and slot value.
To utilize long short-term memory (Long-Short Term Memory, LSTM) model to the dialogue of user text above
For carrying out natural language understanding, firstly, constructing simulator, LSTM model is trained by simulator, then, will be obtained
To trained LSTM model, LSTM model segments the dialogue of the user got text input above to text above is talked with
Processing, obtains several words, then obtains the markup information of each word, and according to the markup information of each word, identifies in dialogue
Field value, intention value and slot value, the final output for including in text understand result information.
The interactive method of fused data library inquiry provided in an embodiment of the present invention is believed by the entropy of database bracket groove
Breath, faster completes conversation tasks, shortens dialog turns, improve human-computer dialogue efficiency, keep interactive system more intelligent.
It is further, described to talk with hereafter text generation natural language according to described on the basis of the above various embodiments,
And before exporting the natural language, further includes:
If judgement, which is known, does not include intention value in the current state information, field value is not included yet, then from described default
The first fixed text is obtained in database, talks with hereafter text as described；
If judgement is known in the current state information comprising field value, intention value is not included, then from the preset data
The second fixed text is obtained in library, talks with hereafter text as described；
If judgement is known in the current state information comprising intention value, field value is not included, then from the preset data
Third fixed text is obtained in library, talks with hereafter text as described；
If judgement is known i.e. comprising intention value in the current state information, and includes field value, and include all slots
Slot value, then the 4th fixed text is obtained from the presetting database, as the dialogue hereafter text.
Specifically, system obtains the dialogue text above of user first, to the dialogue of the user got text above
Natural language understanding is carried out, acquisition understands result information, understands that result information includes in null value, field value, intention value and slot value
One or more.
After obtaining the understanding result information for talking with text above, according to result information is understood, the current shape of dialogue is updated
State information, current state information is the default key message relevant to conversation tasks currently got, related to conversation tasks
Default key message include field, be intended to and several slots.
After having updated the current state information of dialogue, system, which needs to be produced according to the current state information of dialogue, talks with
Strategy is interacted with user.Different current state informations, system will export different conversation contents.
By taking one is ordered the Task man-machine interactive system of train ticket as an example, default key message packet relevant to conversation tasks
Including field, intention and several slots, field value is train ticket, it is intended that value may include booking tickets, return ticket and changing label etc., several slots
It may include five slots, this five slots are respectively to originate geosyncline, purpose geosyncline, departure time slot, train type slot and budget gold
Volume slot.Only when the default key message relevant to conversation tasks that system is got can satisfy when needing of execution task,
System could execute corresponding task, and otherwise, system, with user's interaction, will obtain as quickly as possible and appoint with dialogue by constantly
It is engaged in relevant default key message.
Also, different field, it will cause user to have unused intention, different intentions will also correspond to different
Slot.
One is ordered in the Task man-machine interactive system of train ticket, and default key message relevant to conversation tasks includes neck
Domain, intention and several slots, field value are train ticket, it is intended that value may include ticket booking, return ticket and change label etc., several slots can be with
Including five slots, this five slots are respectively to originate geosyncline, purpose geosyncline, departure time slot, train type slot and budget amount slot.
In one Task man-machine interactive system made a reservation, default key message relevant to conversation tasks include field,
It is intended to and several slots, corresponding field value is dining room, it is intended that value may include making a reservation, ordering take-away etc., several slots can wrap
Four slots are included, this four slots are respectively geosyncline, consumption time slot and budget amount slot where style of cooking slot, dining room.
Therefore, system preferentially can obtain user by dialogue and want system at no determination field relevant to task
The field of the task of execution, and these key messages are successively obtained according to the sequence of field, intention and slot.Generate dialog strategy
Mode is specific further include:
If not including intention value before judgement is known in status information, does not include field value yet, then obtained from presetting database
The first fixed text is taken, talks with hereafter text as described；
If judgement know before in status information comprising field value, do not include intention value, then obtain the from presetting database
Two fixed texts talk with hereafter text as described；
If judgement know before in status information comprising intention value, do not include field value, then obtain the from presetting database
Three fixed texts talk with hereafter text as described；
If i.e. comprising intention value in status information before judgement is known, and including field value, and include the slot value of all slots,
The 4th fixed text is then obtained from presetting database, talks with hereafter text as described.
Still by taking said one orders the Task man-machine interactive system of train ticket as an example, default key relevant to conversation tasks
Information include field, be intended to and several slots, field value be train ticket, it is intended that value may include ticket booking, return ticket and change label etc., if
A dry slot may include five slots, this five slots be respectively originate geosyncline, purpose geosyncline, departure time slot, train type slot and
Budget amount slot.
The dialogue for the user that system is got text above, and the natural language understanding to the text, acquisition understand result
Information, according to the current state information for understanding result information update dialogue, if after updating the current state information talked with, dialogue
Current state information remains as null value, and the current state information of dialogue remains as null value, indicates that system does not obtain so far
Get any default key message relevant to conversation tasks.At this point, system obtains the first fixed text from presetting database,
First fixed text can be with are as follows: and " owner, may I ask have you any instructions to give ", using first fixed text as dialogue hereafter text,
Then, corresponding natural language is generated, and is exported, after user hears the conversation content of system output, next conversation content
In will embody field relevant to conversation tasks, for example, the next conversation content of user can be with are as follows: " me please be help to order one
Open train ticket ".
If after the current state information for updating dialogue, the current state information of dialogue are as follows: got field value as fire
Ticket, at this point, system obtains the second fixed text from presetting database, the second fixed text can be with are as follows: " may I ask you to be intended to order
Train ticket when ", using second fixed text as dialogue, hereafter then text generates corresponding natural language, and defeated
Out, it after user hears the conversation content of system output, will be embodied in next conversation content relevant to conversation tasks
It is intended to, for example, if user wants to return ticket, the next conversation content of user can be with are as follows: " I wants to return ticket ", if user does want to order
Ticket, the next conversation content of user can be with are as follows: " I wants to stipulate day ".
If after the current state information for updating dialogue, the current state information of dialogue are as follows: having got intention value is to order
Ticket, at this point, system obtains third fixed text from presetting database, third fixed text can be with are as follows: " may I ask you are intended to which is ordered
It train ticket ", using the third fixed text as dialogue, hereafter then text generates corresponding natural language, and defeated
Out, it after user hears the conversation content of system output, will be embodied in next conversation content relevant to conversation tasks
Field, for example, if user wants to order ferry ticket, the next conversation content of user can be with are as follows: and " I wants to order ferry ticket ", if user is true
Want to order train ticket in fact, the next conversation content of user can be with are as follows: " I wants to stipulate the train ticket in day ".
If after the current state information for updating dialogue, the current state information of dialogue are as follows: got field value as fire
Ticket, it is intended that be worth to book tickets, origin slot value is Beijing, destination slot value is Shanghai, departure time slot value is tomorrow, train class
Type groove value be high-speed rail, 800 yuan of budget amount groove location or less.At this point, system can be with according to the current state information of dialogue
Execution task then executes corresponding task, and according to task execution as a result, obtaining the 4th fixed text from presetting database
This, if booking tickets successfully, the 4th fixed text can be with are as follows: " ticket has made a reservation for complete ".
The interactive method of fused data library inquiry provided in an embodiment of the present invention is believed by the entropy of database bracket groove
Breath, faster completes conversation tasks, shortens dialog turns, improve human-computer dialogue efficiency, keep interactive system more intelligent.
Fig. 2 is the interactive system schematic diagram of fused data library inquiry provided in an embodiment of the present invention, as shown in Fig. 2,
The embodiment of the present invention provides a kind of interactive system of fused data library inquiry, for executing described in any of the above-described embodiment
Method, specifically include natural language understanding module 201, state tracking module 202, decision making module 203 and natural language
Generation module 204, in which:
Natural language understanding module 201 is used to carry out natural language understanding to the dialogue of the user got text above,
Acquisition understands result information, described to understand that result information includes one of null value, field value, intention value and slot value or more
Kind；State tracking module 202 updates the current state information of dialogue, the current shape for understanding result information according to described
State information is the default key message relevant to conversation tasks currently got, the default key relevant to conversation tasks
Information includes field, intention and several slots；If decision making module 203 is known in the current state information i.e. for judging
Comprising intention value, and include field value, but do not include the slot value of all slots, then it is maximum to obtain entropy from presetting database
Slot generates dialogue hereafter text, the slot value of the target slot is not included in described as target slot, and according to the target slot
In current state information；Spatial term module 204 is used to talk with hereafter text generation natural language according to described, and exports
The natural language.
Specifically, system obtains the dialogue text above of user first, is obtained by 201 Duis of natural language understanding module
The dialogue of the user arrived text above carries out natural language understanding, and acquisition understands result information, understand result information include null value,
One or more of field value, intention value and slot value.
After obtaining the understanding result information for talking with text above, by state tracking module 202 according to understanding that result believes
Breath, updates the current state information of dialogue, current state information is the default key relevant to conversation tasks currently got
Information, default key message relevant to conversation tasks include field, intention and several slots.
After having updated the current state information of dialogue, if state letter before being known by the judgement of decision making module 203
Include intention value in breath, and include field value, but do not include the slot value of all slots, then obtains entropy from presetting database
Maximum slot, as target slot, and according to target slot, generating dialogue, hereafter text, the slot value of target slot are not included in preceding state
In information.
Finally, by spatial term module 204 according to dialogue hereafter text generation natural language, and export it is described from
Right language completes the interaction of epicycle.
The interactive system of fused data library inquiry provided in an embodiment of the present invention is believed by the entropy of database bracket groove
Breath, faster completes conversation tasks, shortens dialog turns, improve human-computer dialogue efficiency, keep interactive system more intelligent.
On the basis of the above embodiments, further, the interactive system further include:
Speech recognition module carries out speech recognition for the voice data to the user got, and it is described right to obtain
Talk about text above.
Specifically, the object of processing is that the dialogue of the user got is literary above when system carries out natural language understanding
This can directly obtain the dialogue of user if system obtains the dialog information of user by input equipments such as mouse, keyboards
Text above.
If system by voice capture device acquire be user voice dialog information, also need to know by voice
The voice data for the user that other module will acquire carries out speech recognition, obtains the dialogue text above of user.
The interactive system of fused data library inquiry provided in an embodiment of the present invention is believed by the entropy of database bracket groove
Breath, faster completes conversation tasks, shortens dialog turns, improve human-computer dialogue efficiency, keep interactive system more intelligent.
On the basis of the above various embodiments, further, the dialogue of the described couple of user got text above is carried out
Natural language understanding specifically includes:
Word segmentation processing is carried out to the dialogue text above, obtains several words；
Obtain the markup information of each word；
According to the markup information of each word, field value, intention value and the slot value for including in the dialogue text above are identified.
Specifically, to the dialogue of the user got text progress natural language understanding above, specific step is as follows:
The dialogue text above for the user that will acquire by preset algorithm model carries out word segmentation processing, obtains several
Then word obtains the markup information of each word, and according to the markup information of each word, identification, which is talked in text above, includes
Field value, intention value and slot value.
To utilize long short-term memory (Long-Short Term Memory, LSTM) model to the dialogue of user text above
For carrying out natural language understanding, firstly, constructing simulator, LSTM model is trained by simulator, then, will be obtained
To trained LSTM model, LSTM model segments the dialogue of the user got text input above to text above is talked with
Processing, obtains several words, then obtains the markup information of each word, and according to the markup information of each word, identifies in dialogue
Field value, intention value and slot value, the final output for including in text understand result information.
The interactive system of fused data library inquiry provided in an embodiment of the present invention is believed by the entropy of database bracket groove
Breath, faster completes conversation tasks, shortens dialog turns, improve human-computer dialogue efficiency, keep interactive system more intelligent.
On the basis of the above various embodiments, further, the decision making module is also used to:
If judgement, which is known, does not include intention value in the current state information, field value is not included yet, then from described default
The first fixed text is obtained in database, talks with hereafter text as described；
If judgement is known in the current state information comprising field value, intention value is not included, then from the preset data
The second fixed text is obtained in library, talks with hereafter text as described；
If judgement is known in the current state information comprising intention value, field value is not included, then from the preset data
Third fixed text is obtained in library, talks with hereafter text as described；
If judgement is known i.e. comprising intention value in the current state information, and includes field value, and include all slots
Slot value, then the 4th fixed text is obtained from the presetting database, as the dialogue hereafter text.
Specifically, system obtains the dialogue text above of user first, to the dialogue of the user got text above
Natural language understanding is carried out, acquisition understands result information, understands that result information includes in null value, field value, intention value and slot value
One or more.
After obtaining the understanding result information for talking with text above, according to result information is understood, the current shape of dialogue is updated
State information, current state information is the default key message relevant to conversation tasks currently got, related to conversation tasks
Default key message include field, be intended to and several slots.
After having updated the current state information of dialogue, system, which needs to be produced according to the current state information of dialogue, talks with
Strategy is interacted with user.Different current state informations, system will export different conversation contents.
By taking one is ordered the Task man-machine interactive system of train ticket as an example, default key message packet relevant to conversation tasks
Including field, intention and several slots, field value is train ticket, it is intended that value may include booking tickets, return ticket and changing label etc., several slots
It may include five slots, this five slots are respectively to originate geosyncline, purpose geosyncline, departure time slot, train type slot and budget gold
Volume slot.Only when the default key message relevant to conversation tasks that system is got can satisfy when needing of execution task,
System could execute corresponding task, and otherwise, system, with user's interaction, will obtain as quickly as possible and appoint with dialogue by constantly
It is engaged in relevant default key message.
Also, different field, it will cause user to have unused intention, different intentions will also correspond to different
Slot.
One is ordered in the Task man-machine interactive system of train ticket, and default key message relevant to conversation tasks includes neck
Domain, intention and several slots, field value are train ticket, it is intended that value may include ticket booking, return ticket and change label etc., several slots can be with
Including five slots, this five slots are respectively to originate geosyncline, purpose geosyncline, departure time slot, train type slot and budget amount slot.
In one Task man-machine interactive system made a reservation, default key message relevant to conversation tasks include field,
It is intended to and several slots, corresponding field value is dining room, it is intended that value may include making a reservation, ordering take-away etc., several slots can wrap
Four slots are included, this four slots are respectively geosyncline, consumption time slot and budget amount slot where style of cooking slot, dining room.
Therefore, system preferentially can obtain user by dialogue and want system at no determination field relevant to task
The field of the task of execution, and these key messages are successively obtained according to the sequence of field, intention and slot.Generate dialog strategy
Mode is specific further include:
If not including intention value before judgement is known in status information, does not include field value yet, then obtained from presetting database
The first fixed text is taken, talks with hereafter text as described；
If judgement know before in status information comprising field value, do not include intention value, then obtain the from presetting database
Two fixed texts talk with hereafter text as described；
If judgement know before in status information comprising intention value, do not include field value, then obtain the from presetting database
Three fixed texts talk with hereafter text as described；
If i.e. comprising intention value in status information before judgement is known, and including field value, and include the slot value of all slots,
The 4th fixed text is then obtained from presetting database, talks with hereafter text as described.
Still by taking said one orders the Task man-machine interactive system of train ticket as an example, default key relevant to conversation tasks
Information include field, be intended to and several slots, field value be train ticket, it is intended that value may include ticket booking, return ticket and change label etc., if
A dry slot may include five slots, this five slots be respectively originate geosyncline, purpose geosyncline, departure time slot, train type slot and
Budget amount slot.
The dialogue for the user that system is got text above, and the natural language understanding to the text, acquisition understand result
Information, according to the current state information for understanding result information update dialogue, if after updating the current state information talked with, dialogue
Current state information remains as null value, and the current state information of dialogue remains as null value, indicates that system does not obtain so far
Get any default key message relevant to conversation tasks.At this point, system obtains the first fixed text from presetting database,
First fixed text can be with are as follows: and " owner, may I ask have you any instructions to give ", using first fixed text as dialogue hereafter text,
Then, corresponding natural language is generated, and is exported, after user hears the conversation content of system output, next conversation content
In will embody field relevant to conversation tasks, for example, the next conversation content of user can be with are as follows: " me please be help to order one
Open train ticket ".
If after the current state information for updating dialogue, the current state information of dialogue are as follows: got field value as fire
Ticket, at this point, system obtains the second fixed text from presetting database, the second fixed text can be with are as follows: " may I ask you to be intended to order
Train ticket when ", using second fixed text as dialogue, hereafter then text generates corresponding natural language, and defeated
Out, it after user hears the conversation content of system output, will be embodied in next conversation content relevant to conversation tasks
It is intended to, for example, if user wants to return ticket, the next conversation content of user can be with are as follows: " I wants to return ticket ", if user does want to order
Ticket, the next conversation content of user can be with are as follows: " I wants to stipulate day ".
If after the current state information for updating dialogue, the current state information of dialogue are as follows: having got intention value is to order
Ticket, at this point, system obtains third fixed text from presetting database, third fixed text can be with are as follows: " may I ask you are intended to which is ordered
It train ticket ", using the third fixed text as dialogue, hereafter then text generates corresponding natural language, and defeated
Out, it after user hears the conversation content of system output, will be embodied in next conversation content relevant to conversation tasks
Field, for example, if user wants to order ferry ticket, the next conversation content of user can be with are as follows: and " I wants to order ferry ticket ", if user is true
Want to order train ticket in fact, the next conversation content of user can be with are as follows: " I wants to stipulate the train ticket in day ".
If after the current state information for updating dialogue, the current state information of dialogue are as follows: got field value as fire
Ticket, it is intended that be worth to book tickets, origin slot value is Beijing, destination slot value is Shanghai, departure time slot value is tomorrow, train class
Type groove value be high-speed rail, 800 yuan of budget amount groove location or less.At this point, system can be with according to the current state information of dialogue
Execution task then executes corresponding task, and according to task execution as a result, obtaining the 4th fixed text from presetting database
This, if booking tickets successfully, the 4th fixed text can be with are as follows: " ticket has made a reservation for complete ".
The embodiment of the present invention provides a kind of interactive system of fused data library inquiry, for executing any of the above-described implementation
Method described in example, the system provided through this embodiment execute the specific steps of above-mentioned a certain method as described in the examples
Identical as above-mentioned corresponding embodiment, details are not described herein again.
The interactive system of fused data library inquiry provided in an embodiment of the present invention is believed by the entropy of database bracket groove
Breath, faster completes conversation tasks, shortens dialog turns, improve human-computer dialogue efficiency, keep interactive system more intelligent.
Fig. 3 is the structural schematic diagram of electronic equipment provided in an embodiment of the present invention, as shown in figure 3, the equipment includes: place
Manage device 301, memory 302 and bus 303；
Wherein, processor 301 and memory 302 complete mutual communication by the bus 303；
Processor 301 is used to call the program instruction in memory 302, to execute provided by above-mentioned each method embodiment
Method, for example, natural language understanding is carried out to the dialogue of the user got text above, acquisition understands result information,
It is described to understand that result information includes one or more of null value, field value, intention value and slot value；
Result information is understood according to described, updates the current state information of dialogue, and the current state information is currently to obtain
The default key message relevant to conversation tasks got, the default key message relevant to conversation tasks include field,
It is intended to and several slots；
If judgement is known i.e. comprising intention value in the current state information, and includes field value, but do not include all
The slot value of slot then obtains the maximum slot of entropy from presetting database, as target slot, and according to the target slot, generation pair
Hereafter text, the slot value of the target slot are not included in the current state information words；
Talk with hereafter text generation natural language according to described, and exports the natural language.
The embodiment of the present invention provides a kind of computer program product, and the computer program product is non-transient including being stored in
Computer program on computer readable storage medium, the computer program include program instruction, when described program instructs quilt
When computer executes, computer is able to carry out method provided by above-mentioned each method embodiment, for example,
Natural language understanding is carried out to the dialogue of the user got text above, acquisition understands result information, the reason
Solving result information includes one or more of null value, field value, intention value and slot value；
Result information is understood according to described, updates the current state information of dialogue, and the current state information is currently to obtain
The default key message relevant to conversation tasks got, the default key message relevant to conversation tasks include field,
It is intended to and several slots；
If judgement is known i.e. comprising intention value in the current state information, and includes field value, but do not include all
The slot value of slot then obtains the maximum slot of entropy from presetting database, as target slot, and according to the target slot, generation pair
Hereafter text, the slot value of the target slot are not included in the current state information words；
Talk with hereafter text generation natural language according to described, and exports the natural language.
The embodiment of the present invention provides a kind of non-transient computer readable storage medium, the non-transient computer readable storage
Medium storing computer instruction, the computer instruction make the computer execute side provided by above-mentioned each method embodiment
Method, for example,
Natural language understanding is carried out to the dialogue of the user got text above, acquisition understands result information, the reason
Solving result information includes one or more of null value, field value, intention value and slot value；
Result information is understood according to described, updates the current state information of dialogue, and the current state information is currently to obtain
The default key message relevant to conversation tasks got, the default key message relevant to conversation tasks include field,
It is intended to and several slots；
If judgement is known i.e. comprising intention value in the current state information, and includes field value, but do not include all
The slot value of slot then obtains the maximum slot of entropy from presetting database, as target slot, and according to the target slot, generation pair
Hereafter text, the slot value of the target slot are not included in the current state information words；
Talk with hereafter text generation natural language according to described, and exports the natural language.
Those of ordinary skill in the art will appreciate that: realize that all or part of the steps of above method embodiment can pass through
The relevant hardware of program instruction is completed, and program above-mentioned can be stored in a computer readable storage medium, the program
When being executed, step including the steps of the foregoing method embodiments is executed；And storage medium above-mentioned includes: ROM, RAM, magnetic disk or light
The various media that can store program code such as disk.
The embodiments such as device and equipment described above are only schematical, wherein described be used as separate part description
Unit may or may not be physically separated, component shown as a unit may or may not be
Physical unit, it can it is in one place, or may be distributed over multiple network units.It can be according to the actual needs
Some or all of the modules therein is selected to achieve the purpose of the solution of this embodiment.Those of ordinary skill in the art are not paying
In the case where creative labor, it can understand and implement.
Through the above description of the embodiments, those skilled in the art can be understood that each embodiment can
It realizes by means of software and necessary general hardware platform, naturally it is also possible to pass through hardware.Based on this understanding, on
Stating technical solution, substantially the part that contributes to existing technology can be embodied in the form of software products in other words, should
Computer software product may be stored in a computer readable storage medium, such as ROM/RAM, magnetic disk, CD, including several fingers
It enables and using so that a computer equipment (can be personal computer, server or the network equipment etc.) executes each implementation
Method described in certain parts of example or embodiment.
Finally, it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations；Although
Present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that: it still may be used
To modify the technical solutions described in the foregoing embodiments or equivalent replacement of some of the technical features；
And these are modified or replaceed, technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution spirit and
Range.
Claims (10)
1. a kind of interactive method of fused data library inquiry characterized by comprising
Natural language understanding is carried out to the dialogue of the user got text above, acquisition understands result information, the understanding knot
Fruit information includes one or more of null value, field value, intention value and slot value；
Result information is understood according to described, updates the current state information of dialogue, and the current state information is currently to get
Default key message relevant to conversation tasks, the default key message relevant to conversation tasks include field, be intended to
With several slots；
If judgement is known i.e. comprising intention value in the current state information, and includes field value, but do not include all slots
Slot value then obtains the maximum slot of entropy from presetting database, as target slot, and according to the target slot, generates under dialogue
The slot value of text, the target slot is not included in the current state information；
Talk with hereafter text generation natural language according to described, and exports the natural language.
2. interactive method according to claim 1, which is characterized in that the dialogue of the described couple of user got is above
Text carries out natural language understanding, and acquisition understands before result information, further includes:
Speech recognition is carried out to the voice data of the user got, obtains the dialogue text above.
3. interactive method according to claim 1, which is characterized in that the dialogue of the described couple of user got is above
Text carries out natural language understanding, specifically includes:
Word segmentation processing is carried out to the dialogue text above, obtains several words；
Obtain the markup information of each word；
According to the markup information of each word, field value, intention value and the slot value for including in the dialogue text above are identified.
4. interactive method according to claim 1, which is characterized in that described according to the dialogue hereafter text generation
Natural language, and before exporting the natural language, further includes:
If judgement, which is known, does not include intention value in the current state information, field value is not included yet, then from the preset data
The first fixed text is obtained in library, talks with hereafter text as described；
If judgement is known in the current state information comprising field value, intention value is not included, then from the presetting database
The second fixed text is obtained, talks with hereafter text as described；
If judgement is known in the current state information comprising intention value, field value is not included, then from the presetting database
Third fixed text is obtained, talks with hereafter text as described；
If judgement is known i.e. comprising intention value in the current state information, and includes field value, and include the slot of all slots
Value then obtains the 4th fixed text from the presetting database, talks with hereafter text as described.
5. a kind of interactive system of fused data library inquiry characterized by comprising
Natural language understanding module carries out natural language understanding for the dialogue text above to the user got, obtains reason
Result information is solved, it is described to understand that result information includes one or more of null value, field value, intention value and slot value；
State tracking module updates the current state information of dialogue, the current state for understanding result information according to described
Information is the default key message relevant to conversation tasks currently got, the default crucial letter relevant to conversation tasks
Breath includes field, intention and several slots；
Decision making module, if for judging to know i.e. comprising intention value in the current state information, and include field value, but
Slot value not comprising all slots then obtains the maximum slot of entropy, as target slot, and according to the mesh from presetting database
Slot is marked, generating dialogue, hereafter text, the slot value of the target slot are not included in the current state information；
Spatial term module for talking with hereafter text generation natural language according to described, and exports the natural language.
6. interactive system according to claim 5, which is characterized in that further include:
Speech recognition module carries out speech recognition for the voice data to the user got, obtains in the dialogue
Text.
7. interactive system according to claim 5, which is characterized in that the dialogue of the described couple of user got is above
Text carries out natural language understanding, specifically includes:
Word segmentation processing is carried out to the dialogue text above, obtains several words；
Obtain the markup information of each word；
According to the markup information of each word, field value, intention value and the slot value for including in the dialogue text above are identified.
8. interactive system according to claim 5, which is characterized in that the decision making module is also used to:
If judgement, which is known, does not include intention value in the current state information, field value is not included yet, then from the preset data
The first fixed text is obtained in library, talks with hereafter text as described；
If judgement is known in the current state information comprising field value, intention value is not included, then from the presetting database
The second fixed text is obtained, talks with hereafter text as described；
If judgement is known in the current state information comprising intention value, field value is not included, then from the presetting database
Third fixed text is obtained, talks with hereafter text as described；
If judgement is known i.e. comprising intention value in the current state information, and includes field value, and include the slot of all slots
Value then obtains the 4th fixed text from the presetting database, talks with hereafter text as described.
9. a kind of electronic equipment characterized by comprising
Memory and processor, the processor and the memory complete mutual communication by bus；The memory
It is stored with the program instruction that can be executed by the processor, the processor calls described program instruction to be able to carry out right such as and wants
Seek 1 to 4 any method.
10. a kind of non-transient computer readable storage medium, is stored thereon with computer program, which is characterized in that when the meter
When calculation machine program is executed by processor, the method as described in Claims 1-4 is any is realized.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811231195.5A CN109543010A (en) | 2018-10-22 | 2018-10-22 | The interactive method and system of fused data library inquiry |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811231195.5A CN109543010A (en) | 2018-10-22 | 2018-10-22 | The interactive method and system of fused data library inquiry |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109543010A true CN109543010A (en) | 2019-03-29 |
Family
ID=65844156
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811231195.5A Pending CN109543010A (en) | 2018-10-22 | 2018-10-22 | The interactive method and system of fused data library inquiry |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109543010A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110096579A (en) * | 2019-04-23 | 2019-08-06 | 南京硅基智能科技有限公司 | A kind of more wheel dialogue methods |
| CN110110057A (en) * | 2019-04-22 | 2019-08-09 | 南京硅基智能科技有限公司 | A kind of dynamic words art group system and its application in Intelligent dialogue robot |
| CN110442697A (en) * | 2019-08-06 | 2019-11-12 | 上海灵羚科技有限公司 | A kind of man-machine interaction method, system, computer equipment and storage medium |
| CN111209297A (en) * | 2019-12-31 | 2020-05-29 | 深圳云天励飞技术有限公司 | Data query method and device, electronic equipment and storage medium |
| CN111353035A (en) * | 2020-03-11 | 2020-06-30 | 镁佳(北京)科技有限公司 | Man-machine conversation method and device, readable storage medium and electronic equipment |
| CN111538818A (en) * | 2020-03-26 | 2020-08-14 | 深圳云天励飞技术有限公司 | Data query method and device, electronic equipment and storage medium |
| CN111639167A (en) * | 2020-05-21 | 2020-09-08 | 北京声智科技有限公司 | Task conversation method and device |
| CN111737146A (en) * | 2020-07-21 | 2020-10-02 | 中国人民解放军国防科技大学 | Statement generation method for dialog system evaluation |
| CN112581938A (en) * | 2019-09-30 | 2021-03-30 | 华为技术有限公司 | Voice breakpoint detection method, device and equipment based on artificial intelligence |
| WO2021121436A1 (en) * | 2019-12-21 | 2021-06-24 | 华南理工大学 | Dialogue strategy method for task-oriented dialogue system |
| WO2021129262A1 (en) * | 2019-12-26 | 2021-07-01 | 思必驰科技股份有限公司 | Server-side processing method and server for actively initiating conversation, and voice interaction system capable of actively initiating conversation |
| WO2024066391A1 (en) * | 2022-09-30 | 2024-04-04 | 中兴通讯股份有限公司 | Intent-based driving method and apparatus for telecommunication network management product |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104360897A (en) * | 2014-10-29 | 2015-02-18 | 百度在线网络技术（北京）有限公司 | Conversation processing method and conversation management system |
| US20170011117A1 (en) * | 2014-03-26 | 2017-01-12 | Huawei Technologies Co., Ltd. | Help Processing Method and Device Based on Semantic Recognition |
-
2018
- 2018-10-22 CN CN201811231195.5A patent/CN109543010A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170011117A1 (en) * | 2014-03-26 | 2017-01-12 | Huawei Technologies Co., Ltd. | Help Processing Method and Device Based on Semantic Recognition |
| CN104360897A (en) * | 2014-10-29 | 2015-02-18 | 百度在线网络技术（北京）有限公司 | Conversation processing method and conversation management system |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110110057A (en) * | 2019-04-22 | 2019-08-09 | 南京硅基智能科技有限公司 | A kind of dynamic words art group system and its application in Intelligent dialogue robot |
| CN110096579A (en) * | 2019-04-23 | 2019-08-06 | 南京硅基智能科技有限公司 | A kind of more wheel dialogue methods |
| CN110442697A (en) * | 2019-08-06 | 2019-11-12 | 上海灵羚科技有限公司 | A kind of man-machine interaction method, system, computer equipment and storage medium |
| CN110442697B (en) * | 2019-08-06 | 2023-09-12 | 中电金信软件(上海)有限公司 | Man-machine interaction method, system, computer equipment and storage medium |
| CN112581938B (en) * | 2019-09-30 | 2024-04-09 | 华为技术有限公司 | Speech breakpoint detection method, device and equipment based on artificial intelligence |
| CN112581938A (en) * | 2019-09-30 | 2021-03-30 | 华为技术有限公司 | Voice breakpoint detection method, device and equipment based on artificial intelligence |
| WO2021121436A1 (en) * | 2019-12-21 | 2021-06-24 | 华南理工大学 | Dialogue strategy method for task-oriented dialogue system |
| WO2021129262A1 (en) * | 2019-12-26 | 2021-07-01 | 思必驰科技股份有限公司 | Server-side processing method and server for actively initiating conversation, and voice interaction system capable of actively initiating conversation |
| CN111209297A (en) * | 2019-12-31 | 2020-05-29 | 深圳云天励飞技术有限公司 | Data query method and device, electronic equipment and storage medium |
| CN111209297B (en) * | 2019-12-31 | 2024-05-03 | 深圳云天励飞技术有限公司 | Data query method, device, electronic equipment and storage medium |
| CN111353035A (en) * | 2020-03-11 | 2020-06-30 | 镁佳(北京)科技有限公司 | Man-machine conversation method and device, readable storage medium and electronic equipment |
| CN111538818A (en) * | 2020-03-26 | 2020-08-14 | 深圳云天励飞技术有限公司 | Data query method and device, electronic equipment and storage medium |
| CN111639167A (en) * | 2020-05-21 | 2020-09-08 | 北京声智科技有限公司 | Task conversation method and device |
| CN111639167B (en) * | 2020-05-21 | 2024-04-16 | 北京声智科技有限公司 | Task dialogue method and device |
| CN111737146B (en) * | 2020-07-21 | 2020-11-13 | 中国人民解放军国防科技大学 | Statement generation method for dialog system evaluation |
| CN111737146A (en) * | 2020-07-21 | 2020-10-02 | 中国人民解放军国防科技大学 | Statement generation method for dialog system evaluation |
| WO2024066391A1 (en) * | 2022-09-30 | 2024-04-04 | 中兴通讯股份有限公司 | Intent-based driving method and apparatus for telecommunication network management product |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109543010A (en) | The interactive method and system of fused data library inquiry | |
| US11386893B2 (en) | Human-computer interaction processing system, method, storage medium, and electronic device | |
| US10311454B2 (en) | Customer interaction and experience system using emotional-semantic computing | |
| US10679134B2 (en) | Automated ontology development | |
| CN106407178B (en) | A kind of session abstraction generating method, device, server apparatus and terminal device | |
| US9575936B2 (en) | Word cloud display | |
| US9477752B1 (en) | Ontology administration and application to enhance communication data analytics | |
| US20180082679A1 (en) | Optimal human-machine conversations using emotion-enhanced natural speech using hierarchical neural networks and reinforcement learning | |
| CN105701088B (en) | The method and apparatus manually talked with are switched to from machine dialogue | |
| US20180373744A1 (en) | System and methods for using conversational similarity for dimension reduction in deep analytics | |
| US20160260434A1 (en) | Selective speech recognition for chat and digital personal assistant systems | |
| EP2650776A1 (en) | System and methods for delivering advanced natural language interaction applications | |
| US20210150398A1 (en) | Conversational interchange optimization | |
| CN107992543A (en) | Question and answer exchange method and device, computer equipment and computer-readable recording medium | |
| CN108345692A (en) | A kind of automatic question-answering method and system | |
| CN108763495B (en) | Interactive method, system, electronic equipment and storage medium | |
| CN105117387B (en) | A kind of intelligent robot interactive system | |
| CN109145104A (en) | For talking with interactive method and apparatus | |
| CN108877782A (en) | Audio recognition method and device | |
| US10255346B2 (en) | Tagging relations with N-best | |
| US20150193532A1 (en) | Labeling/naming of themes | |
| CN109739965A (en) | Moving method and device, equipment, the readable storage medium storing program for executing of cross-cutting dialog strategy | |
| CN108829668A (en) | Generation method, device, computer equipment and the storage medium of text information | |
| CN110427470A (en) | Question and answer processing method, device and electronic equipment | |
| CN103942056B (en) | A kind of man-machine interaction method and device based on scene |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190329 |