CN105930367A - Intelligent chatting robot control method and control device - Google Patents
Intelligent chatting robot control method and control device Download PDFInfo
- Publication number
- CN105930367A CN105930367A CN201610226768.XA CN201610226768A CN105930367A CN 105930367 A CN105930367 A CN 105930367A CN 201610226768 A CN201610226768 A CN 201610226768A CN 105930367 A CN105930367 A CN 105930367A
- Authority
- CN
- China
- Prior art keywords
- statement
- user
- chat
- module
- revert
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/903—Querying
- G06F16/9032—Query formulation
- G06F16/90332—Natural language query formulation or dialogue systems
Abstract
The invention relates to an intelligent chatting robot control method and control device. The control method includes the following steps: acquiring a user input statement of a first user, and matching the user input statement with statements in a chatting database to acquire a first chatting statement having a maximum matching degree greater than a preset value in the chatting database; using the next chatting statement as a robot output statement to be fed back to the first user if the next chatting statement exists in the first chatting statement; using the user input statement of the first user as a new topic robot output statement to be sent to N second users if the next chatting statement does not exist in the first chatting statement, receiving reply statements returned by the N second users, and using the reply statements as robot output statements of the first user to be fed back to the first user. The invention also provides a corresponding intelligent chatting robot control device. The method and the device can allow a chatting robot to possess artificial intelligences, and can improve the Turing test passing rate.
Description
Technical field
The present invention relates to chat robots technical field, particularly relate to a kind of intelligence chat robots controlling party
Method and control device.
Background technology
Chat robots is one of most important content of artificial intelligence.Along with the fast development of Robotics,
User is more and more higher to the functional requirement of chat robots.If user chats with chat robots, point
Unclear chat to as if people or chat robots, then he chat robots can passes through turing test, and this is described
Chat robots has possessed artificial intelligence.
As it is shown in figure 1, conventional chat robot is all based on what chat data storehouse built, in chat data storehouse
Tables of data generally comprise two elementary fields, first character section is characteristic character or words and phrases, second field
It it is the content automatically replied.It is good that content in chat data storehouse needs manually to increase or import manual manufacture in advance
Reply text.The cost in manual manufacture chat data storehouse is the highest, and cannot enumerate all possible from
The dynamic content replied, therefore result in existing chat robots and user conventional to some of user can only input language
Sentence effectively automatically replies, and if other user's read statements of user do not exist in chat data storehouse
When individual features character or words and phrases, chat robots does not then know how to answer, and can only carry out invalid oneself
Dynamic answer, such as answers " " etc., as shown in Figure 2.Further, because chat data storehouse is artificial constructed
, it is not during people chat naturally, to produce next chat statement, so being structure person oneself structure sometimes
Next chat statement out, next the chat statement produced in chatting really with people has a certain distance,
With such manual construction next chat statement out, it is difficult to pass through turing test.
In sum, owing to being limited by the limitation in complete chat data storehouse, there is figure in conventional chat robot
The technical problem that spirit test percent of pass is low.
Summary of the invention
Based on this, it is necessary to for the technical problem that the turing test percent of pass of conventional chat robot existence is low,
A kind of intelligence chat robots control method is provided and controls device.
According to an aspect of the present invention, it is provided that a kind of intelligence chat robots control method, including following step
Rapid:
Obtain user's read statement of first user, by described user's read statement and the language in chat data storehouse
Sentence mates, and obtains the matching degree in described chat data storehouse maximum, and maximum matching degree is more than preset value
First chat statement;
If there is next chat statement in described first chat statement, then using next chat statement described as machine
People's output statement feeds back to described first user;
If a chat statement in the absence of described first chat statement, then the user of described first user is inputted
Statement is sent to N number of second user as the robot output statement of new topic, receives described N number of second and uses
The revert statement that family returns, feeds back to described revert statement as to the robot output statement of first user
First user.
Wherein in an embodiment, described using user's read statement of described first user as new topic
Robot output statement is sent to N number of second user, receives the revert statement that described N number of second user returns,
Described revert statement is fed back to as robot output statement the step of first user, including:
User's read statement of described first user is sent to N number of as the robot output statement of new topic
Second user, receive the revert statement returned at first in described N number of second user, returns described at first
Revert statement feeds back to first user as robot output statement.
Wherein in an embodiment, described using user's read statement of described first user as new topic
Robot output statement be sent to N number of second user, receive and described N number of second user return at first
Revert statement, feeds back to first user using the described revert statement returned at first as robot output statement
After step further comprising the steps of:
User's read statement and the revert statement returned at first of described first user are chatted as continuous print two
It statement stores in chat data storehouse.
Wherein in an embodiment, described method also includes:
If not receiving the revert statement of described second user in Preset Time, the most again choose N number of 3rd use
Family, is sent to described using user's read statement of described first user as the robot output statement of new topic
3rd user, receives the revert statement that described 3rd user returns, and described revert statement is defeated as robot
Go out statement and feed back to first user.
Wherein in an embodiment, if a chat statement in the absence of described first chat statement, then
User's read statement of described first user is sent to N number of second as the robot output statement of new topic
User, receives the revert statement that described N number of second user returns, is used as to first by described revert statement
The robot output statement at family feeds back in the step of first user, and described second user is currently without addition
Chat user, wherein, the value of N be currently without add chat number of users with currently participate in
The ratio of the number of users of chat.
According to another aspect of the present invention, also provide for a kind of intelligence chat robots and control device, including:
Chat data storehouse, is used for storing experience chat statement, and described experience chat statement includes the first chat language
Sentence and the next chat statement of the first chat statement;
Retrieval matching module, respectively with chat data storehouse and choose module and be connected, is used for obtaining first user
User's read statement, mates described user's read statement with the statement in chat data storehouse, obtains institute
State the matching degree maximum in chat data storehouse and matching degree the first chat statement more than preset value;If described
There is next chat statement in one chat statement, then using anti-as robot output statement for next chat statement described
Feed described first user;If a chat statement in the absence of described first chat statement, then send and choose finger
Order is to choosing module;
Choose module, be connected with described retrieval matching module, choose instruction for reception, receive and choose finger
N number of second user is chosen after order;
Study module, is connected with described module of choosing, for chatting statement as new topic using described first
Robot output statement is sent to N number of second user;And,
Exercise module, for receiving the revert statement that described N number of second user returns, by described revert statement
First user is fed back to as to the robot output statement of first user.
Wherein in an embodiment, described exercise module described N number of second user of reception returns at first
Revert statement, feeds back to first user using the described revert statement returned at first as robot output statement.
Wherein in an embodiment, described intelligence chat robots controls device and also includes time limit module, institute
State the pre-recorded Preset Time of time limit module, if not receiving the reply language of described second user in Preset Time
Sentence, the most described time limit module send reply time-out instruction to described in chooses module, described in choose module and receive
Again N number of 3rd user is chosen after time-out instruction.
Wherein in an embodiment, described intelligence chat robots controls device and also includes memory module, institute
State memory module for by user's read statement of described first user and the revert statement that returns at first as even
Two continuous chat statements store in chat data storehouse.
Wherein in an embodiment, described in choose module and include:
Chat statistical module, for obtaining currently without the number of users adding chat;
Idle statistical module, for obtaining the current number of users participating in chat;
Ratio module, is used for calculating currently without adding the number of users of chat and the most participating in chat
The ratio of number of users, obtains N value;And,
Randomly select module, in the user never adding chat, randomly select N number of second user.
Above-mentioned intelligent chat robots control method and intelligence chat robots control device and achieve chatting machine
Device people's autonomous learning is chatted, and sends, for first user, user's read statement of coming, when in chat data
Storehouse is retrieved less than meet matching degree require chat statement time using user's read statement of first user as newly
The robot output statement of topic is sent to N number of second user, and N number of second user is exported robot
User's read statement that statement is replied at first is as the robot output statement replied to first user.With existing
Technology is compared, and the present invention greatly improves class human nature, reasonability and the authenticity that chat robots is replied, and makes
Obtain chat robots and possess artificial intelligence, substantially increase the percent of pass of turing test.
Accompanying drawing explanation
Fig. 1 is the chat data storehouse schematic diagram of the existing chat robots built based on chat data storehouse;
Fig. 2 is the chat schematic diagram of the existing chat robots built based on chat data storehouse and user;
Fig. 3 is the process principle figure of intelligence chat robots control method in an embodiment;
Fig. 4 is the chat signal that in an embodiment, application intelligence chat robots control method and user chat
Figure;
Fig. 5 is the structure principle chart that in an embodiment, intelligence chat robots controls device;
Fig. 6 is the structure principle chart choosing module in an embodiment;
Fig. 7 is the structure principle chart of intelligence chat robots system in an embodiment.
Detailed description of the invention
Refer to Fig. 3, a kind of intelligence chat robots control method, comprise the following steps:
Step 102: obtain user's read statement of first user, by user's read statement and chat data storehouse
Statement mate, obtain the matching degree in chat data storehouse maximum, and maximum matching degree be more than preset value
First chat statement.
Concrete, obtain a first user and send user's read statement of coming, retrieve from chat data storehouse
Maximum with this user's read statement matching degree, and maximum matching degree is more than a chat statement in the presence of K%
First chat statement.
Wherein, K is more than 0.If being understandable that K=0, then be exactly optionally from chat data storehouse
Take out a chat statement, be skimble-skamble;The value of K is the biggest, then retrieve successful possibility the least;
When K is too big, can cause retrieve unsuccessfully is too many, so that the first too many chat statement needs execution certainly
The step of primary learning chat, can bring the biggest pressure to system;K is the least, then the first chat statement obtained
The least with the matching degree of user's read statement;But, if K is the least, the robot obtained can be caused again defeated
Go out statement often with user's read statement " irrelevant ";So, K needs to take appropriate value, makes
Obtaining user's output chat statement to replying satisfied, the operating pressure simultaneously making system is not too large.One
In individual preferred embodiment, the value of K is 50~100, it is preferable that the acquiescence value of K is 70.
User's read statement of user can be word, it is also possible to be voice, it might even be possible to be video.One
In individual preferred embodiment, retrieval is completed by the search engine calling chat data storehouse.The institute in chat data storehouse
There is the next chat of a chat statement maximum with user's read statement matching degree of user in chat statement
Statement is clearly the most reasonably replying of this user's read statement of this user, and statement can be expressed as character
String, the calculating of statement matching degree can be converted into the calculating of string matching degree or similarity, can use
Some string matching degree or similarity algorithm, such as Edit Furthest Neighbor (editing distance, it is simply that be used for calculating
From the minimum insertion required for former string (s) converting into target string (t), the number deleted and replace.Obviously when one
Individual statement editing is the minimum insertion needed for another statement, and the number deleted and replace is the least, then mate
Spend the biggest), maximum public substring LCS method (the maximum public substring of obvious two statements is the longest, then this two
Individual statement matching degree is the biggest) etc., do not repeat them here.
Step 104: if the first chat statement exists next chat statement, then using next chat statement as machine
People's output statement feeds back to first user.
Concrete, the first chat statement is to retrieve with user's from all chat statements in chat data storehouse
User's read statement matching degree meets the chat statement that the matching degree in above-mentioned steps 102 requires, obtains first
After chat statement, from chat data storehouse, retrieve next chat statement that the first chat statement is corresponding.If first
There is next chat statement in chat statement, then using next chat statement as the robot output to first user
Statement is sent to first user.Being exported by robot output statement to the mode of first user can be text
Mode, it is also possible to be the mode of voice, it is also possible to be the mode of video, it is also possible to be other modes, such as,
In the way of expression.
Step 106: if a chat statement in the absence of the first chat statement, then the user of first user is inputted
Statement is sent to N number of second user as the robot output statement of new topic, receives N number of second user and returns
The revert statement returned, feeds back to first user using revert statement as to the robot output statement of first user.
Concrete, if a chat statement in the absence of the first chat statement, then carry out autonomous learning chat, will
User's read statement is sent to N number of second user as the robot output statement of new topic to be sought to reply, and
The user's read statement of replying obtained is sent to the first use as to the robot output statement of first user
Family.If the first chat statement exists next chat statement, illustrate chat data storehouse exists to meet matching degree
The revert statement required, therefore, there is no need to carry out autonomous learning chat.In the absence of the first chat statement
One chat statement, illustrates that can not find the user's read statement to first user in chat data storehouse suitably returns
Multiple, so needing to carry out autonomous learning chat.
The intelligent chat robots control method of the present embodiment, retrieves less than meeting coupling in chat data storehouse
Degree require chat statement time carry out autonomous learning chat, by using user's read statement of first user as
The robot output statement of new topic is sent to N number of second user to be sought to reply.The method is greatly enhanced
Class human nature, reasonability and the authenticity that chat robots is replied so that chat robots possesses artificial intelligence,
Substantially increase the percent of pass of turing test.
In one embodiment, step 106 includes: using user's read statement of first user as new topic
Robot output statement be sent to N number of second user, receive the reply returned at first in N number of second user
Statement, feeds back to first user using the revert statement returned at first as robot output statement.
Concrete, the precedence that robot output statement is replied by N number of second user has difference, in order to
Make first user wait the shortest time replied, the user's read statement replied at first is used as to first
The robot output statement that family is replied.If there is two or more user robot output statement to be carried out simultaneously
Reply, then choose one of them chat revert statement as the robot output statement replied to first user.
Specifically can set selection principle according to actual needs, as one of them chat revert statement can be randomly choosed
As the robot output statement replied to first user, it is also possible to select a longest chat to reply
Statement is as the robot output statement etc. replied to first user, and the present embodiment is to this and is not specifically limited.
In one embodiment, further comprising the steps of after step 106: the user of first user is inputted
Statement and the revert statement returned at first store in chat data storehouse as continuous print two chat statement.
Specifically, by user's read statement of first user and the revert statement that returns at first as continuous print two
Individual chat statement stores in chat data storehouse, will user's read statement of first user and replying at first
User's read statement stores to chat data storehouse as experience chat statement, with in abundant chat data storehouse
Hold.
In the present embodiment, in order to directly retrieve the user's input with user after improving from chat data storehouse
The chance of statement matching, the chat statement obtained by autonomous learning stores to database as experience chat statement
Content with abundant chat data storehouse.When user's certain user's read statement in the future is chatted with by autonomous learning
When it increases to the chat statement maximum matching degree in chat data storehouse more than K%, then can be by chat data storehouse
In the next chat statement of this chat statement directly as the returning of this user's read statement of user
Multiple, and without carrying out autonomous learning chat again, contribute to reducing system pressure and improving the speed replying user.
In one embodiment, step 106 is further comprising the steps of: if not receiving second in Preset Time
The revert statement of user, chooses N number of 3rd user the most again, using user's read statement of first user as
The robot output statement of new topic is sent to the 3rd user, receives the revert statement that the 3rd user returns, will
Revert statement feeds back to first user as robot output statement.
Concrete, Preset Time is the t second.When N number of second user does not has to reply within the preset time t second,
The most again choose N number of 3rd user carry out autonomous learning seek reply.When N number of second user does not has in limit
Reply in fixed second time t, then just should again choose some other user and again seek to reply, otherwise
Less than revert statement, first user waits always can think that chat robots has abandoned chat.Further, t is
Natural number more than 0, specifically can arbitrarily set the value of t according to practical situations.An enforcement wherein
In example, the value of t is (1~60) s.If t is the biggest, the time waited is the longest, if t is the least, waits
Time the shortest.If t is too big, first user can be allowed to wait, and the time replied is oversize, if t is the least, and can
Causing another m user to have little time to reply chat robots, therefore, t should take suitable time span.One relatively
In good embodiment, owing in General N the second user, any user strikes the word reply needs time of 3 seconds, because of
This, the value arranging t is 3 seconds.However, it is necessary to explanation, in the present embodiment, the value of t is 3 seconds
Simply an embodiment, is not intended to limit the present invention.
In one embodiment, if there being multiple second user to return revert statement, then within the preset time t second
Receive the revert statement that all second users return, and respectively by user's read statement and the reception of first user
To each revert statement as continuous print two chat statement join in chat data storehouse.In the present embodiment,
By all second user's revert statements of obtaining in Preset Time in autonomous learning chat process the most empirically
Chat statement stores to chat data storehouse the content in abundant chat data storehouse.As it has been described above, by by autonomous
User's revert statement that study chat obtains stores to chat data storehouse as experience chat statement, it is possible to rich
The content of rich chat data, and then improve the user later directly retrieved from chat data storehouse and to user
The chance of read statement coupling, reduces system pressure, improves the speed replying user.
In one embodiment, the second above-mentioned user is currently without the user adding chat, wherein, N
Value be currently without number of users and the ratio of the current number of users the most participating in chat adding chat
Value.Wherein, choosing of N number of second user specifically includes following steps:
First, obtain currently without number of users N1 adding chat.
Secondly, current number of users N2 participating in chat is obtained.
Afterwards, calculate currently without adding number of users N1 of chat and the current number of users participating in chat
The ratio of mesh N2, and reduced value result rounds up, and obtains N,
Finally, never add and the user of chat randomly selects N number of second user.
Further, the 3rd above-mentioned user is currently without the user adding chat, and is not yet selected
As the user of the second user, during the choosing of N number of 3rd user, finally chat from currently without adding
My god, and be not yet selected in the user as the second user and randomly select N number of 3rd user, other step
All with above-mentioned N number of second user to choose process identical, do not repeat them here.
In the present embodiment, for alleviating system pressure and improving computational efficiency, N value is set for currently without adding
Enter number of users N1 of chat and the ratio of current number of users N2 participating in chat.However, it is desirable to
Illustrating, the present invention is not limited to this, and the quantity of such as N can also randomly select, therefore, above
Using N value is currently without adding number of users N1 of chat and the current number of users participating in chat
The ratio of mesh N2 is an embodiment, is not intended to limit the present invention.
Below in conjunction with specific embodiment correspondence above-mentioned intelligent chat robots control method and user's chat
Process is described in detail.For convenience of description, taking 70 with K in the present embodiment, preset time t is to enter for 3 seconds
Row explanation.
As shown in Figure 4, application intelligence chat robots control method and user's chat during it is an embodiment
Chat schematic diagram.
Saying " my belly is how it hurts " with intelligence chat robots when first user, intelligence chat robots is in chat
The chat statement retrieving maximum coupling in database has " my stomach-ache ", and matching degree is more than 70%, thus
By " my stomach-ache " as the first chat statement.Then, retrieving chat statement from chat data storehouse is
" I has a stomachache what if?" " what's the matter, why has a stomachache for next corresponding chat statement?Catch a cold
??" reply to user U.
Saying " Xiao Wang is looking for you " with chat robots when first, intelligence chat robots is in chat data storehouse
Do not retrieve the chat statement being more than 70% with " Xiao Wang is looking for you " matching degree.Now, intelligence chatting machine
Device people carries out autonomous learning chat, and intelligence chat robots obtains currently without the number of users adding chat
N1=3000, obtains current number of users N2=1000 participating in chat, calculates
Never randomly select 3 the second users, by first user with the user of intelligence chat robots chat
User's read statement is sent to the second user as the robot output statement of new topic.One of them user is
First reply " which Xiao Wang?", turnaround time is 2 seconds, within 3 seconds limited, so intelligence chat
Robot is by " which Xiao Wang?" reply to user U, and by " Xiao Wang is looking for you ", " which Xiao Wang?”
Join in chat data storehouse as continuous print chat statement, when other two second users also reply in 3s
Time, also statement of chatting accordingly is joined in chat data storehouse.
Above-mentioned whole process is entirely what intelligence chat robots was automatically performed.Intelligence chat robots is to many
User's autonomous learning, the statement replying to user obtained is actually other users and replies to the language of robot
Sentence, so while be that chat robots is chatted with user, be substantially in chat between different user, but use
Family feels it is that chat robots is being chatted with oneself, so the reply of chat robots is very reasonable, very
Truly such that it is able to pass through turing test;Along with chat robots learns to multi-user self, chat data
Storehouse can be more and more abundanter, and the first chat statement is inevitable with the matching degree of user's read statement of user more and more higher,
So that the reply of chat robots is more reasonable, the most truly, figure spirit can increasingly be improved
The percent of pass of test.
Concrete, the intelligent chat robots of the present embodiment is the robot of general justice, including energy and user session
All simulating human behaviors or thought and simulation other biological machinery, also include some computer program,
Such as qq chat robots program etc..
Referring to Fig. 5, based on above-mentioned intelligent chat robots control method, the present invention also provides for a kind of intelligence
Chat robots control device, including chat data storehouse 502, retrieval matching module 504, choose module 506,
Study module 508, exercise module 510, time limit module 512 and memory module 514.Wherein:
Chat data storehouse 502 is used for storing experience chat statement, and experience chat statement includes the first chat statement
Next chat statement with the first chat statement.
Concrete, chat statement in chat data storehouse 502 is ordered into, under major part chat statement has
One chat statement, only last the chat statement at the end of chat do not have next chat statement.Chat
Day data storehouse 502 can carry out branch by the form of text to chat data and store, it would however also be possible to employ tables of data
Form carry out branch store, each chat statement is as a line in text or tables of data, according to people
The PS of class chat, sequentially stores chat statement.The previous chat language of one chat statement
Sentence is then stored in the previous row that this chat statement is expert at, and the later chat statement of a chat statement is then
It is stored in rear a line that this chat statement is expert at.Data acquisition range in chat data storehouse 502 includes
Chat data, including chat data or human chat's data of chat robots, these data can be text,
The message form such as voice, video, these data can import chat data storehouse 502 in batches, it is also possible to increases in real time
Add into chat data storehouse 502 to amount formula.
Retrieval matching module 504 respectively with chat data storehouse 502 and choose module 506 and be connected, be used for obtaining
User's read statement of first user, mates user's read statement with the statement in chat data storehouse,
Obtain the matching degree maximum in chat data storehouse and matching degree the first chat statement more than preset value;If first
There is next chat statement in chat statement, then as robot output statement, next chat statement is fed back to the
One user;If a chat statement, then send and choose instruction to choosing module in the absence of the first chat statement.
Concrete, retrieval matching module 504 is retrieved defeated with the user of first user from chat data storehouse 502
Enter the maximum and maximum matching degree of statement matching degree the first chat statement more than K%.In one embodiment, K's
Value is 50~100, and in a preferred embodiment, the acquiescence value of K is 70.
In the present embodiment, retrieval matching module 504 is the search engine being integrated in chat data storehouse 502,
This search engine can carry out the fuzzy matching of statement, supports text, voice and the retrieval of video and fuzzy matching,
Can retrieve defeated with the user of user from all chat statements in chat data storehouse 502 by search engine
Enter the chat statement that statement matching degree is maximum.In one embodiment, it is also possible to by cloud storage and cloud
Computing technique improves the speed of chat data storehouse 502 and search engine thereof.
Choose module 506 to be connected with retrieval matching module 504, choose instruction for reception, receive and choose
N number of second user is chosen after instruction.
Study module 508 is connected with choosing module 506, for the statement machine as new topic of chatting first
Device people's output statement is sent to N number of second user.
Exercise module 510 is for receiving the revert statement that N number of second user returns, using revert statement as giving
The robot output statement of first user feeds back to first user.
Concrete, exercise module 510 receives the revert statement returned at first in N number of second user, will at first
The revert statement returned feeds back to first user as robot output statement.
Time limit module 512 respectively with choose module 506 and exercise module 510 is connected, time limit module 512 is pre-
First record Preset Time, if exercise module 510 does not receives the revert statement of the second user in Preset Time,
Then time limit module 512 sends reply time-out instruction to choosing module 506, chooses module 506 and receives time-out and refer to
Again N number of 3rd user is chosen after order.
Memory module 514 is connected with chat data 502 and exercise module 510 respectively, and memory module 514 is used
In user's read statement of first user and the revert statement that returns at first are chatted statement as continuous print two
Store in chat data storehouse 502.
Concrete, user's read statement and the user replied at first of first user are inputted by memory module 514
Statement stores to chat data storehouse 502 as experience chat statement, with in abundant chat data storehouse 502
Hold.Further, all second user's revert statements obtained in Preset Time are all made by memory module 514
The content in abundant chat data storehouse is stored to chat data storehouse for experience chat statement.If in preset time t
Have multiple second user to return revert statement in Miao, then Connection Module 510 receives all second user's returns
Revert statement, memory module 514 is respectively by user's read statement of first user and each reply of receiving
Statement joins in chat data storehouse as continuous print two chat statement.
As shown in Figure 6, in one embodiment, above-mentioned module 506 of choosing includes:
Chat statistical module 506a, for obtaining currently without the number of users adding chat.
Idle statistical module 506b, for obtaining the current number of users participating in chat.
Ratio module 506c, for calculating currently without adding the number of users of chat and currently merely participating in
The ratio of it number of users, obtains N value.
Concrete, in order to avoid ratio result to occur, decimal impact is chosen, and in one embodiment, reduced value is tied
Fruit rounds up, and obtains N value.
Randomly select module 506d, in the user never adding chat, randomly select N number of second user.
Further, randomly select module 506 to be additionally operable to choose N number of 3rd user, as N number of second user
When all not having to reply within the preset time t second, randomly select module 506 and again choose N number of 3rd user and enter
Row autonomous learning is sought to reply.3rd user is currently without the user adding chat, and is not yet selected
User as the second user.When choosing N number of three user, randomly select module 506d from currently without
Add chat, and be not yet selected in the user as the second user and randomly select N number of 3rd user.
Referring to Fig. 7, control device based on above-mentioned intelligent chat robots, the present invention provides again a kind of intelligence
Chat system, including multiple users and some above-mentioned based on autonomous learning intelligent chat robots, intelligence
Communicate to connect between chat robots and each user.
The intelligent chatting system based on autonomous learning of the present embodiment can include an intelligent chat robots also
Can include multiple intelligence chat robots, the structure and working principle of concrete intelligence chat robots is the most detailed
Describe to the greatest extent, do not repeat them here.Each intelligence chat robots have multiple user chat, chat robots and
Being connected by internet, mobile Internet, LAN, cloud, Internet of Things or social network communication between user
Connect.
In the development of artificial intelligence, independent learning ability is the important directions that robot turns to intelligence,
When machine has really possessed the ability of autonomous learning, machine will be no longer machine.Chat robots wants to pass through
Turing test, is necessary for the chat statement so that self and human chat's statement extraordinary image, the most identical, so
And, the chat statement of user is not used for autonomous learning by conventional chat robot.Above-mentioned intelligence chat
Robot control method, intelligence chat robots control device and intelligent chatting system achieve chat robots
Autonomous learning is chatted, and sends, for first user, user's read statement of coming, when in chat data storehouse
Retrieve less than meet matching degree require chat statement time using user's read statement of first user as new topic
Robot output statement be sent to N number of second user, and by N number of second user to robot output statement
The user's read statement replied at first is as the robot output statement replied to first user.With prior art
Comparing, the present invention greatly improves class human nature, reasonability and the authenticity that chat robots is replied so that chat
It robot artificial intelligence, substantially increases the percent of pass of turing test.
Each technical characteristic of embodiment described above can combine arbitrarily, for making description succinct, the most right
The all possible combination of each technical characteristic in above-described embodiment is all described, but, if these skills
There is not contradiction in the combination of art feature, is all considered to be the scope that this specification is recorded.
Embodiment described above only have expressed the several embodiments of the present invention, and it describes more concrete and detailed,
But can not therefore be construed as limiting the scope of the patent.It should be pointed out that, for this area
For those of ordinary skill, without departing from the inventive concept of the premise, it is also possible to make some deformation and change
Entering, these broadly fall into protection scope of the present invention.Therefore, the protection domain of patent of the present invention should be with appended power
Profit requires to be as the criterion.
Claims (10)
1. an intelligent chat robots control method, it is characterised in that comprise the following steps:
Obtain user's read statement of first user, by described user's read statement and the language in chat data storehouse
Sentence mates, and obtains the matching degree in described chat data storehouse maximum, and maximum matching degree is more than preset value
First chat statement;
If there is next chat statement in described first chat statement, then using next chat statement described as machine
People's output statement feeds back to described first user;
If a chat statement in the absence of described first chat statement, then the user of described first user is inputted
Statement is sent to N number of second user as the robot output statement of new topic, receives described N number of second and uses
The revert statement that family returns, feeds back to described revert statement as to the robot output statement of first user
First user.
Intelligence chat robots control method the most according to claim 1, it is characterised in that described general
User's read statement of described first user is sent to N number of second use as the robot output statement of new topic
Family, receives the revert statement that described N number of second user returns, is exported as robot by described revert statement
Statement feeds back to the step of first user, including:
User's read statement of described first user is sent to N number of as the robot output statement of new topic
Second user, receive the revert statement returned at first in described N number of second user, returns described at first
Revert statement feeds back to first user as robot output statement.
Intelligence chat robots control method the most according to claim 2, it is characterised in that described
User's read statement of described first user is sent to N number of second as the robot output statement of new topic
User, receives the revert statement returned at first in described N number of second user, by the described reply returned at first
Statement is further comprising the steps of after feeding back to the step of first user as robot output statement:
User's read statement and the revert statement returned at first of described first user are chatted as continuous print two
It statement stores in chat data storehouse.
Intelligence chat robots control method the most according to claim 1, it is characterised in that described side
Method also includes:
If not receiving the revert statement of described second user in Preset Time, the most again choose N number of 3rd use
Family, is sent to described using user's read statement of described first user as the robot output statement of new topic
3rd user, receives the revert statement that described 3rd user returns, and described revert statement is defeated as robot
Go out statement and feed back to first user.
Intelligence chat robots control method the most according to claim 1, it is characterised in that if described
A chat statement in the absence of described first chat statement, then make user's read statement of described first user
Robot output statement for new topic is sent to N number of second user, receives described N number of second user and returns
Revert statement, described revert statement is fed back to the first use as to the robot output statement of first user
In the step at family, described second user is currently without the user adding chat, and wherein, the value of N is for working as
Before do not add the ratio of the number of users of chat and the current number of users the most participating in chat.
6. an intelligent chat robots controls device, it is characterised in that including:
Chat data storehouse, is used for storing experience chat statement, and described experience chat statement includes the first chat language
Sentence and the next chat statement of the first chat statement;
Retrieval matching module, respectively with chat data storehouse and choose module and be connected, is used for obtaining first user
User's read statement, mates described user's read statement with the statement in chat data storehouse, obtains institute
State the matching degree maximum in chat data storehouse and matching degree the first chat statement more than preset value;If described
There is next chat statement in one chat statement, then using anti-as robot output statement for next chat statement described
Feed described first user;If a chat statement in the absence of described first chat statement, then send and choose finger
Order is to choosing module;
Choose module, be connected with described retrieval matching module, choose instruction for reception, receive and choose finger
N number of second user is chosen after order;
Study module, is connected with described module of choosing, for chatting statement as new topic using described first
Robot output statement is sent to N number of second user;And,
Exercise module, for receiving the revert statement that described N number of second user returns, by described revert statement
First user is fed back to as to the robot output statement of first user.
Intelligence chat robots the most according to claim 6 controls device, it is characterised in that described white silk
Practise module and receive the revert statement returned at first in described N number of second user, by the described reply returned at first
Statement feeds back to first user as robot output statement.
Intelligence chat robots the most according to claim 6 controls device, it is characterised in that also include
Time limit module, the described pre-recorded Preset Time of time limit module, if do not receive described second in Preset Time
The revert statement of user, the most described time limit module sends and chooses module, described choosing described in reply time-out instruction extremely
Delivery block chooses N number of 3rd user after receiving time-out instruction again.
Intelligence chat robots the most according to claim 6 controls device, it is characterised in that also include
Memory module, described memory module is used for user's read statement of described first user and returning of returning at first
Multiple statement stores in chat data storehouse as continuous print two chat statement.
Intelligence chat robots the most according to claim 6 controls device, it is characterised in that described
Choose module to include:
Chat statistical module, for obtaining currently without the number of users adding chat;
Idle statistical module, for obtaining the current number of users participating in chat;
Ratio module, chats with participation with current currently without the number of users of addition chat for calculating
The ratio of number of users, obtain N value;And,
Randomly select module, in the user never adding chat, randomly select N number of second user.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610226768.XA CN105930367B (en) | 2016-04-12 | 2016-04-12 | Intelligent chat robot control method and control device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610226768.XA CN105930367B (en) | 2016-04-12 | 2016-04-12 | Intelligent chat robot control method and control device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105930367A true CN105930367A (en) | 2016-09-07 |
CN105930367B CN105930367B (en) | 2020-06-09 |
Family
ID=56838915
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610226768.XA Active CN105930367B (en) | 2016-04-12 | 2016-04-12 | Intelligent chat robot control method and control device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105930367B (en) |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107016070A (en) * | 2017-03-22 | 2017-08-04 | 北京光年无限科技有限公司 | A kind of interactive method and device for intelligent robot |
CN109465823A (en) * | 2018-11-06 | 2019-03-15 | 泰康保险集团股份有限公司 | Study of Intelligent Robot Control method and device, electronic equipment, storage medium |
CN109760041A (en) * | 2017-11-10 | 2019-05-17 | 贝斯平环球公司 | Cloud management system and its operating method based on chat robots |
US10318927B2 (en) | 2017-07-17 | 2019-06-11 | ExpertHiring, LLC | Method and system for managing, matching, and sourcing employment candidates in a recruitment campaign |
CN110019700A (en) * | 2017-09-13 | 2019-07-16 | 阿里巴巴集团控股有限公司 | Data processing method and equipment |
CN110059000A (en) * | 2019-03-15 | 2019-07-26 | 深圳壹账通智能科技有限公司 | Chat drama test control method, device, computer equipment and storage medium |
CN111027976A (en) * | 2019-11-13 | 2020-04-17 | 支付宝(杭州)信息技术有限公司 | Method for obtaining transaction identity information of fraudulent party |
CN114650263A (en) * | 2017-06-29 | 2022-06-21 | 谷歌有限责任公司 | Proactive provision of new content to group chat participants |
US11887594B2 (en) | 2017-03-22 | 2024-01-30 | Google Llc | Proactive incorporation of unsolicited content into human-to-computer dialogs |
US11929069B2 (en) | 2017-05-03 | 2024-03-12 | Google Llc | Proactive incorporation of unsolicited content into human-to-computer dialogs |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101076060A (en) * | 2007-03-30 | 2007-11-21 | 腾讯科技(深圳)有限公司 | Chatting robot system and automatic chatting method |
US20080189367A1 (en) * | 2007-02-01 | 2008-08-07 | Oki Electric Industry Co., Ltd. | User-to-user communication method, program, and apparatus |
CN101431573A (en) * | 2007-11-08 | 2009-05-13 | 上海赢思软件技术有限公司 | Method and equipment for implementing automatic customer service through human-machine interaction technology |
CN103390047A (en) * | 2013-07-18 | 2013-11-13 | 天格科技(杭州)有限公司 | Chatting robot knowledge base and construction method thereof |
US20140122407A1 (en) * | 2012-10-26 | 2014-05-01 | Xiaojiang Duan | Chatbot system and method having auto-select input message with quality response |
CN105068661A (en) * | 2015-09-07 | 2015-11-18 | 百度在线网络技术(北京)有限公司 | Man-machine interaction method and system based on artificial intelligence |
-
2016
- 2016-04-12 CN CN201610226768.XA patent/CN105930367B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20080189367A1 (en) * | 2007-02-01 | 2008-08-07 | Oki Electric Industry Co., Ltd. | User-to-user communication method, program, and apparatus |
CN101076060A (en) * | 2007-03-30 | 2007-11-21 | 腾讯科技(深圳)有限公司 | Chatting robot system and automatic chatting method |
CN101431573A (en) * | 2007-11-08 | 2009-05-13 | 上海赢思软件技术有限公司 | Method and equipment for implementing automatic customer service through human-machine interaction technology |
US20140122407A1 (en) * | 2012-10-26 | 2014-05-01 | Xiaojiang Duan | Chatbot system and method having auto-select input message with quality response |
CN103390047A (en) * | 2013-07-18 | 2013-11-13 | 天格科技(杭州)有限公司 | Chatting robot knowledge base and construction method thereof |
CN105068661A (en) * | 2015-09-07 | 2015-11-18 | 百度在线网络技术(北京)有限公司 | Man-machine interaction method and system based on artificial intelligence |
Non-Patent Citations (1)
Title |
---|
王智慧: "基于自然语言理解的自动应答技术及应用研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107016070A (en) * | 2017-03-22 | 2017-08-04 | 北京光年无限科技有限公司 | A kind of interactive method and device for intelligent robot |
US11887594B2 (en) | 2017-03-22 | 2024-01-30 | Google Llc | Proactive incorporation of unsolicited content into human-to-computer dialogs |
CN107016070B (en) * | 2017-03-22 | 2020-06-02 | 北京光年无限科技有限公司 | Man-machine conversation method and device for intelligent robot |
US11929069B2 (en) | 2017-05-03 | 2024-03-12 | Google Llc | Proactive incorporation of unsolicited content into human-to-computer dialogs |
CN114650263A (en) * | 2017-06-29 | 2022-06-21 | 谷歌有限责任公司 | Proactive provision of new content to group chat participants |
US10318927B2 (en) | 2017-07-17 | 2019-06-11 | ExpertHiring, LLC | Method and system for managing, matching, and sourcing employment candidates in a recruitment campaign |
US11087283B2 (en) | 2017-07-17 | 2021-08-10 | ExpertHiring, LLC | Method and system for managing, matching, and sourcing employment candidates in a recruitment campaign |
CN110019700A (en) * | 2017-09-13 | 2019-07-16 | 阿里巴巴集团控股有限公司 | Data processing method and equipment |
CN110019700B (en) * | 2017-09-13 | 2023-01-17 | 阿里巴巴集团控股有限公司 | Data processing method and device |
CN109760041B (en) * | 2017-11-10 | 2022-05-06 | 贝斯平环球公司 | Chat robot-based cloud management system and operation method thereof |
CN109760041A (en) * | 2017-11-10 | 2019-05-17 | 贝斯平环球公司 | Cloud management system and its operating method based on chat robots |
CN109465823A (en) * | 2018-11-06 | 2019-03-15 | 泰康保险集团股份有限公司 | Study of Intelligent Robot Control method and device, electronic equipment, storage medium |
CN110059000A (en) * | 2019-03-15 | 2019-07-26 | 深圳壹账通智能科技有限公司 | Chat drama test control method, device, computer equipment and storage medium |
CN111027976B (en) * | 2019-11-13 | 2022-06-14 | 支付宝(杭州)信息技术有限公司 | Method for obtaining transaction identity information of fraudulent party |
CN111027976A (en) * | 2019-11-13 | 2020-04-17 | 支付宝(杭州)信息技术有限公司 | Method for obtaining transaction identity information of fraudulent party |
Also Published As
Publication number | Publication date |
---|---|
CN105930367B (en) | 2020-06-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105930367A (en) | Intelligent chatting robot control method and control device | |
CN105051728B (en) | Data base querying translation system | |
US7814048B2 (en) | Knowledge extraction from online discussion forums | |
CN110008322A (en) | Art recommended method and device under more wheel session operational scenarios | |
CN105975530B (en) | Robot conversation control method and system based on chat big data knowledge base | |
CN106202301A (en) | A kind of intelligent response system based on degree of depth study | |
CN106020488A (en) | Man-machine interaction method and device for conversation system | |
CN109791549A (en) | Machine customer interaction towards dialogue | |
JP6853752B2 (en) | Dialogue scenario corpus creation support system | |
CN116452169B (en) | Online recruitment generation type recommendation system and method | |
CN106407405A (en) | A social contact system based on love and marriage matching degree search | |
Kephart et al. | A symbiotic cognitive computing perspective on autonomic computing | |
CN112035714A (en) | Man-machine conversation method based on character companions | |
CN111767386B (en) | Dialogue processing method, device, electronic equipment and computer readable storage medium | |
CN109919793A (en) | Activity participates in analysis and recommended method | |
CN110019695A (en) | A kind of automatic chatting response method | |
CN110020007A (en) | Robot interactive control method and device, storage medium, server | |
CN112199154A (en) | Distributed collaborative sampling central optimization-based reinforcement learning training system and method | |
CN112581203A (en) | Providing explanatory product recommendations in a session | |
CN108509610A (en) | A kind of search method and system of activity and companion | |
Debowski et al. | Conversational agents in creative work–a systematic literature review and research agenda for remote design thinking | |
CN117421416B (en) | Interactive search method and device and electronic equipment | |
CN110472113A (en) | A kind of intelligent interaction engine optimization method, device, equipment | |
CN105991399A (en) | Method and system for realizing questioning over network | |
KR101393258B1 (en) | A system of proposing the creative keyword and a method of proposing the related keyword using thereof |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |