CN106777081A - Method and device for determining conversational system acknowledgment strategy - Google Patents
Method and device for determining conversational system acknowledgment strategy Download PDFInfo
- Publication number
- CN106777081A CN106777081A CN201611146416.XA CN201611146416A CN106777081A CN 106777081 A CN106777081 A CN 106777081A CN 201611146416 A CN201611146416 A CN 201611146416A CN 106777081 A CN106777081 A CN 106777081A
- Authority
- CN
- China
- Prior art keywords
- user
- active user
- intended
- information
- advance
- 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
- 238000000034 method Methods 0.000 title claims abstract description 18
- 230000008447 perception Effects 0.000 claims abstract description 37
- 238000000605 extraction Methods 0.000 claims description 6
- 239000000284 extract Substances 0.000 claims description 4
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- User Interface Of Digital Computer (AREA)
Abstract
The present invention provides a kind of method and device for determining conversational system acknowledgment strategy, wherein, method includes, obtains the current dialog information of user input；According to current dialog information, and according to the advance perception intention assessment device for building, determine that active user perceives and active user is intended to；Perceived according to active user and active user is intended to, and according to the advance decision-making device for building, determine acknowledgment strategy.Method and device provided by the present invention for determining conversational system acknowledgment strategy, current dialog information according to user input is determined the current perception of user and is intended to, acknowledgment strategy is adjusted so as to combine the current perception of user with being intended to, response flexibility ratio when conversational system is exchanged with user can be improved.
Description
Technical field
The present invention relates to chat robots field and artificial intelligence conversational system field, more particularly to one kind is for determining
The method and device of conversational system acknowledgment strategy.
Background technology
With continuing to develop for artificial intelligence technology, it is also required to carry out adaptability for the man-machine conversational system for being exchanged
Ground development, to lift interactive experience when user is exchanged with conversational system, so as to improve trust of the user to conversational system
Degree.Existing conversational system, is typically with simple sentence matching technique to realize that the dialog information to user input carries out response,
That is, when being exchanged with user, for same a word that user is input under different scenes, existing conversational system is made
Answer be often identical.That is, existing conversational system can not be adjusted according to the actual perceived of user and intention
Whole acknowledgment strategy, flexibility is not high, not enough intelligently.
The content of the invention
The technical problem to be solved in the present invention is to provide a kind of method and device for determining conversational system acknowledgment strategy,
Current dialog information according to user input determine the current perception of user be intended to so that with reference to the current perception of user with
It is intended to adjust acknowledgment strategy, response flexibility ratio when conversational system is exchanged with user can be improved.
In order to solve the above technical problems, the technical scheme that the present invention is provided is：
On the one hand, the present invention provides a kind of method for determining conversational system acknowledgment strategy, including,
Obtain the current dialog information of user input；
According to current dialog information, and according to the advance perception intention assessment device for building, determine active user perceive and
Active user is intended to；
Perceived according to active user and active user is intended to, and according to the advance decision-making device for building, determine acknowledgment strategy.
Further, current dialog information includes, text information, user voice information, user image information.
Further, the current dialog information of user input is obtained, is specifically included, using the advance Word Input device for building
The text information in current session is extracted, the user voice extracted using the advance sound extractor for building in current session is believed
Breath, and, the user image information in current session is extracted using the advance Extraction of Image device for building；
According to current dialog information, and according to the advance perception intention assessment device for building, determine active user perceive and
Active user is intended to, and specifically includes, and by the text information in current session, user voice information, user image information input is pre-
In the perception intention assessment device for first building, to obtain, active user perceives and active user is intended to；
Perceived according to active user and active user be intended to, and according to the advance decision-making device for building, determine acknowledgment strategy,
Specifically include, by active user perceive and active user be intended to input in advance build decision-making device in, with obtain acknowledgment strategy and
Number is persistently taken turns in the corresponding response of acknowledgment strategy.
Further, also include, perceived according to active user and active user is intended to, update and perceive intention assessment device.
On the other hand, the present invention provides a kind of device for determining conversational system acknowledgment strategy, including,
Dialog information acquisition module, the current dialog information for obtaining user input；
Perceive and be intended to identification module, for according to current dialog information, and according to the advance perception intention assessment for building
Device, determines that active user perceives and active user is intended to；
Response decision-making module, for being intended to according to active user's perception and active user, and determining according to advance structure
Plan device, determines acknowledgment strategy.
Further, current dialog information includes, text information, user voice information, user image information.
Further, dialog information acquisition module, specifically for extracting current right using the advance Word Input device for building
Text information in words, the user voice information in current session is extracted using the advance sound extractor for building, and, use
The Extraction of Image device for building in advance extracts the user image information in current session；
Perceive and be intended to identification module, specifically for by the text information in current session, user voice information, Yong Huying
As in the perception intention assessment device that information input builds in advance, to obtain, active user perceives and active user is intended to；
Response decision-making module, specifically for being perceived active user and active user being intended to the decision-making that input builds in advance
In device, number is persistently taken turns to obtain acknowledgment strategy and the corresponding response of acknowledgment strategy.
Further, also include, identifier update module, be intended to active user for being perceived according to active user, more
It is new to perceive intention assessment device.
Method and device provided by the present invention for determining conversational system acknowledgment strategy, according to the current right of user input
Words information is determined the current perception of user and is intended to, so that acknowledgment strategy is adjusted with being intended to reference to the current perception of user,
Response flexibility ratio when conversational system is exchanged with user can be improved.
Brief description of the drawings
Fig. 1 is the flow chart for determining the method for conversational system acknowledgment strategy provided in an embodiment of the present invention；
Fig. 2 is the block diagram for determining the device of conversational system acknowledgment strategy provided in an embodiment of the present invention.
Specific embodiment
The present invention is further illustrated below by specific embodiment, it should be understood, however, that, these embodiments are only
It is used for specifically describing in more detail, and is not to be construed as limiting the present invention in any form.
Embodiment one
With reference to Fig. 1, the method for determining conversational system acknowledgment strategy that the present embodiment is provided, including,
Step S1：Obtain the current dialog information of user input；
Step S2：According to current dialog information, and according to the advance perception intention assessment device for building, determine active user
Perceive and active user is intended to；
Step S3：Perceived according to active user and active user is intended to, and according to the advance decision-making device for building, it is determined that should
Answer strategy.
Method for determining conversational system acknowledgment strategy provided in an embodiment of the present invention, according to the current right of user input
Words information is determined the current perception of user and is intended to, so that acknowledgment strategy is adjusted with being intended to reference to the current perception of user,
Response flexibility ratio when conversational system is exchanged with user can be improved.
Preferably, current dialog information includes, text information, user voice information, user image information.
It is further preferred that obtaining the current dialog information of user input, specifically include, carried using the advance word for building
The text information during device extracts current session is taken, the user voice in current session is extracted using the advance sound extractor for building
Information, and, the user image information in current session is extracted using the advance Extraction of Image device for building；
According to current dialog information, and according to the advance perception intention assessment device for building, determine active user perceive and
Active user is intended to, and specifically includes, and by the text information in current session, user voice information, user image information input is pre-
In the perception intention assessment device for first building, to obtain, active user perceives and active user is intended to；
Perceived according to active user and active user be intended to, and according to the advance decision-making device for building, determine acknowledgment strategy,
Specifically include, by active user perceive and active user be intended to input in advance build decision-making device in, with obtain acknowledgment strategy and
Number is persistently taken turns in the corresponding response of acknowledgment strategy.
In the present embodiment, by extracting the text information in current session, user voice information, and, user image letter
Breath, obtains user and perceives and be intended to relevant information from multiple angles, so that susceptibility of the conversational system to user profile can be improved,
And then preferably catch user's perception and be intended to, so as to realize the flexible response of conversational system and user.
Specifically, the present embodiment perceives meaning when the current dialog information to user input perceive with intention assessment
Figure identifier is according to current dialog information and preceding N wheel dialog history information come the Perception Area corresponding to judging current dialog information
Between be intended to it is interval, so that it is determined that active user perceives and active user is intended to.
More specifically, after knowing that active user perceives and active user is intended to, the decision-making device in the present embodiment can
Perceived according to active user and be intended to provide acknowledgment strategy with active user, and number is persistently taken turns in the corresponding response of acknowledgment strategy.Need
It is noted that in the present embodiment, the dialogue wheel number that number refers to being carried out using the acknowledgment strategy response is persistently taken turns in response.Example
Such as, when knowing that it is hungry that user is currently perceived as, and being currently intended to purchase, then ensuing 5 wheel acknowledgment strategy is all and pushes away
User's purchase food is recommended, and now, it is 5 that number is persistently taken turns in response.It should be noted that the response corresponding to acknowledgment strategy is persistently taken turns
Number can be set and be adjusted with reference to actual demand, and the present embodiment is not especially limited.
Furthermore, it is necessary to explanation, active user can only be got in current dialog information and is perceived or active user's meaning
(for example, user only has the intention of purchase, but there is not clearly perception) during a kind of in figure, the decision-making device in the present embodiment remains unchanged
Corresponding acknowledgment strategy can be given.That is, the practicality of the method for the present embodiment is high.
It is further preferred that also including, perceived according to active user and active user is intended to, updated and perceive intention assessment
Device.
In the present embodiment, perceiving intention assessment device can be carried out during user is constantly interacted with conversational system
Self-renewing, so as to improve the accuracy and practicality for perceiving intention assessment device.
Embodiment two
With reference to the another aspect that Fig. 2, the present embodiment are provided, the present invention provides a kind of for determining conversational system acknowledgment strategy
Device, including,
Dialog information acquisition module 1, the current dialog information for obtaining user input；
Perceive and be intended to identification module 2, for according to current dialog information, and according to the advance perception intention assessment for building
Device, determines that active user perceives and active user is intended to；
Response decision-making module 3, for being intended to according to active user's perception and active user, and determining according to advance structure
Plan device, determines acknowledgment strategy.
Device for determining conversational system acknowledgment strategy provided in an embodiment of the present invention, according to the current right of user input
Words information is determined the current perception of user and is intended to, so that acknowledgment strategy is adjusted with being intended to reference to the current perception of user,
Response flexibility ratio when conversational system is exchanged with user can be improved.
Preferably, current dialog information includes, text information, user voice information, user image information.
It is further preferred that dialog information acquisition module 1, specifically for being extracted using the advance Word Input device for building
Text information in current session, the user voice information in current session is extracted using the advance sound extractor for building, with
And, the user image information in current session is extracted using the advance Extraction of Image device for building；
Perceive and be intended to identification module 2, specifically for by the text information in current session, user voice information, Yong Huying
As in the perception intention assessment device that information input builds in advance, to obtain, active user perceives and active user is intended to；
Response decision-making module 3, specifically for being perceived active user and active user being intended to the decision-making that input builds in advance
In device, number is persistently taken turns to obtain acknowledgment strategy and the corresponding response of acknowledgment strategy.
In the present embodiment, by extracting the text information in current session, user voice information, and, user image letter
Breath, obtains user and perceives and be intended to relevant information from multiple angles, so that susceptibility of the conversational system to user profile can be improved,
And then preferably catch user's perception and be intended to, so as to realize the flexible response of conversational system and user.
Specifically, the present embodiment perceives meaning when the current dialog information to user input perceive with intention assessment
Figure identifier is according to current dialog information and preceding N wheel dialog history information come the Perception Area corresponding to judging current dialog information
Between be intended to it is interval, so that it is determined that active user perceives and active user is intended to.
More specifically, after knowing that active user perceives and active user is intended to, the decision-making device in the present embodiment can
Perceived according to active user and be intended to provide acknowledgment strategy with active user, and number is persistently taken turns in the corresponding response of acknowledgment strategy.Need
It is noted that in the present embodiment, the dialogue wheel number that number refers to being carried out using the acknowledgment strategy response is persistently taken turns in response.Example
Such as, when knowing that it is hungry that user is currently perceived as, and being currently intended to purchase, then ensuing 5 wheel acknowledgment strategy is all and pushes away
User's purchase food is recommended, and now, it is 5 that number is persistently taken turns in response.It should be noted that the response corresponding to acknowledgment strategy is persistently taken turns
Number can be set and be adjusted with reference to actual demand, and the present embodiment is not especially limited.
Furthermore, it is necessary to explanation, active user can only be got in current dialog information and is perceived or active user's meaning
(for example, user only has the intention of purchase, but there is not clearly perception) during a kind of in figure, the decision-making device in the present embodiment remains unchanged
Corresponding acknowledgment strategy can be given.That is, the practicality of the device of the present embodiment is high.
It is further preferred that also include, identifier update module, for being perceived and active user's meaning according to active user
Figure, updates and perceives intention assessment device.
In the present embodiment, perceiving intention assessment device can be carried out during user is constantly interacted with conversational system
Self-renewing, so as to improve the accuracy and practicality for perceiving intention assessment device.
Although present invention has been a certain degree of description, it will be apparent that, do not departing from the spirit and scope of the present invention
Under the conditions of, the appropriate change of each condition can be carried out.It is appreciated that the invention is not restricted to the embodiment, and it is attributed to right
It is required that scope, its equivalent for including each factor.
Claims (8)
1. a kind of method for determining conversational system acknowledgment strategy, it is characterised in that including：
Obtain the current dialog information of user input；
According to the current dialog information, and according to the advance perception intention assessment device for building, determine active user perceive and
Active user is intended to；
Perceived according to the active user and the active user is intended to, and according to the advance decision-making device for building, determine response
Strategy.
2. the method for determining conversational system acknowledgment strategy according to claim 1, it is characterised in that described current right
Words information includes, text information, user voice information, user image information.
3. the method for determining conversational system acknowledgment strategy according to claim 1, it is characterised in that
The current dialog information for obtaining user input, specifically includes, and extracts current using the advance Word Input device for building
Text information in dialogue, the user voice information in current session is extracted using the advance sound extractor for building, and, adopt
The user image information in current session is extracted with the advance Extraction of Image device for building；
It is described according to the current dialog information, and according to the advance perception intention assessment device for building, determine that active user feels
Know and be intended to active user, specifically include, by the text information in current session, the user voice information, the use
In the perception intention assessment device that family image information input builds in advance, to obtain, active user perceives and active user is intended to；
It is described to be perceived according to the active user and active user intention, and according to the advance decision-making device for building, it is determined that
Acknowledgment strategy, specifically includes, and the active user is perceived and the active user is intended in the decision-making device that input builds in advance,
Number is persistently taken turns to obtain acknowledgment strategy and the corresponding response of the acknowledgment strategy.
4. the method for determining conversational system acknowledgment strategy according to claim 1, it is characterised in that also include, root
Perceived according to the active user and the active user is intended to, update the perception intention assessment device.
5. a kind of device for determining conversational system acknowledgment strategy, it is characterised in that including：
Dialog information acquisition module, the current dialog information for obtaining user input；
Perceive and be intended to identification module, for according to the current dialog information, and according to the advance perception intention assessment for building
Device, determines that active user perceives and active user is intended to；
Response decision-making module, is intended to for being perceived according to the active user with the active user, and according to structure in advance
Decision-making device, determine acknowledgment strategy.
6. the device for determining conversational system acknowledgment strategy according to claim 5, it is characterised in that described current right
Words information includes, text information, user voice information, user image information.
7. the device for determining conversational system acknowledgment strategy according to claim 5, it is characterised in that
The dialog information acquisition module, specifically for extracting the text in current session using the advance Word Input device for building
Word information, the user voice information in current session is extracted using the advance sound extractor for building, and, using structure in advance
Extraction of Image device extract current session in user image information；
Described perception is intended to identification module, specifically for by the text information in current session, the user voice letter
Breath, in the perception intention assessment device that the user image information input builds in advance, is perceived and current use with obtaining active user
Family is intended to；
The response decision-making module, specifically for being perceived the active user and the active user being intended to the advance structure of input
In the decision-making device built, number is persistently taken turns to obtain acknowledgment strategy and the corresponding response of the acknowledgment strategy.
8. the device for determining conversational system acknowledgment strategy according to claim 5, it is characterised in that also include, knows
Other device update module, is intended to for being perceived according to the active user with the active user, updates the perception intention assessment
Device.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611146416.XA CN106777081A (en) | 2016-12-13 | 2016-12-13 | Method and device for determining conversational system acknowledgment strategy |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611146416.XA CN106777081A (en) | 2016-12-13 | 2016-12-13 | Method and device for determining conversational system acknowledgment strategy |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106777081A true CN106777081A (en) | 2017-05-31 |
Family
ID=58880833
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611146416.XA Pending CN106777081A (en) | 2016-12-13 | 2016-12-13 | Method and device for determining conversational system acknowledgment strategy |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106777081A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110046221A (en) * | 2019-03-01 | 2019-07-23 | 平安科技（深圳）有限公司 | A kind of machine dialogue method, device, computer equipment and storage medium |
| CN111538820A (en) * | 2020-04-10 | 2020-08-14 | 出门问问信息科技有限公司 | Exception reply processing device and computer readable storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110144999A1 (en) * | 2009-12-11 | 2011-06-16 | Samsung Electronics Co., Ltd. | Dialogue system and dialogue method thereof |
| CN104462024A (en) * | 2014-10-29 | 2015-03-25 | 百度在线网络技术（北京）有限公司 | Method and device for generating dialogue action strategy model |
| 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 |
| CN105512228A (en) * | 2015-11-30 | 2016-04-20 | 北京光年无限科技有限公司 | Bidirectional question-answer data processing method and system based on intelligent robot |
| CN106095834A (en) * | 2016-06-01 | 2016-11-09 | 竹间智能科技（上海）有限公司 | Intelligent dialogue method and system based on topic |
-
2016
- 2016-12-13 CN CN201611146416.XA patent/CN106777081A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110144999A1 (en) * | 2009-12-11 | 2011-06-16 | Samsung Electronics Co., Ltd. | Dialogue system and dialogue method thereof |
| CN104462024A (en) * | 2014-10-29 | 2015-03-25 | 百度在线网络技术（北京）有限公司 | Method and device for generating dialogue action strategy model |
| 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 |
| CN105512228A (en) * | 2015-11-30 | 2016-04-20 | 北京光年无限科技有限公司 | Bidirectional question-answer data processing method and system based on intelligent robot |
| CN106095834A (en) * | 2016-06-01 | 2016-11-09 | 竹间智能科技（上海）有限公司 | Intelligent dialogue method and system based on topic |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110046221A (en) * | 2019-03-01 | 2019-07-23 | 平安科技（深圳）有限公司 | A kind of machine dialogue method, device, computer equipment and storage medium |
| WO2020177282A1 (en) * | 2019-03-01 | 2020-09-10 | 平安科技（深圳）有限公司 | Machine dialogue method and apparatus, computer device, and storage medium |
| CN110046221B (en) * | 2019-03-01 | 2023-12-22 | 平安科技（深圳）有限公司 | Machine dialogue method, device, computer equipment and storage medium |
| CN111538820A (en) * | 2020-04-10 | 2020-08-14 | 出门问问信息科技有限公司 | Exception reply processing device and computer readable storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3451328B1 (en) | Method and apparatus for verifying information | |
| Eslinger | The web of preaching: New options in homiletic method | |
| EP4239607A4 (en) | Intelligent weighing apparatus based on computer vision technology, and intelligent weighing method | |
| CN107146624B (en) | A kind of method for identifying speaker and device | |
| EP3706069A3 (en) | Image processing method, image processing apparatus, learnt model manufacturing method, and image processing system | |
| CN107918640A (en) | Sample determines method and device | |
| MX2017005187A (en) | Social scores for network elements. | |
| WO2014160651A3 (en) | Presenting true product dimensions within augmented reality | |
| CN109712144A (en) | Processing method, training method, equipment and the storage medium of face-image | |
| EP3777653A4 (en) | Artificial neutral network deep learning-based method, apparatus, learning strategy, and system for analyte analysis | |
| WO2005026891A3 (en) | Systems and methods for gating medical procedures | |
| EP4280526A3 (en) | Code rate for control information | |
| CN109558892A (en) | A kind of target identification method neural network based and system | |
| CN104867076A (en) | Teaching examination question publishing method and system | |
| CN107195300A (en) | voice control method and system | |
| EP4033412A3 (en) | Method and apparatus with neural network training | |
| EP3886511A3 (en) | Message identification method and apparatus | |
| CN105681318A (en) | Information push method and device based on voice recognition of artificial intelligence | |
| EP4008412A4 (en) | Exercise management system, server system, terminal device, and exercise management method | |
| CN106777081A (en) | Method and device for determining conversational system acknowledgment strategy | |
| US20200306589A1 (en) | Systems and methods for real-time feedback and athletic training on a smart user device | |
| EP3918987A4 (en) | Learned model, learning method, and program, and medical information acquisition device, method, and program | |
| Lee et al. | Smartphone-assisted pronunciation learning technique for ambient intelligence | |
| EP4191606A4 (en) | Medical assistance system, medical assistance method, and computer program | |
| CN106909930A (en) | A kind of model and method of the man-machine automatic switchover based on government affairs machine question answering system |
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 |