CN110222162A - A kind of intelligent answer method based on natural language processing and knowledge mapping - Google Patents
A kind of intelligent answer method based on natural language processing and knowledge mapping Download PDFInfo
- Publication number
- CN110222162A CN110222162A CN201910388931.6A CN201910388931A CN110222162A CN 110222162 A CN110222162 A CN 110222162A CN 201910388931 A CN201910388931 A CN 201910388931A CN 110222162 A CN110222162 A CN 110222162A
- Authority
- CN
- China
- Prior art keywords
- knowledge mapping
- dialogue
- scene
- wheel
- natural language
- 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
- 238000013507 mapping Methods 0.000 title claims abstract description 31
- 238000000034 method Methods 0.000 title claims abstract description 16
- 238000003058 natural language processing Methods 0.000 title claims abstract description 15
- 238000012545 processing Methods 0.000 claims abstract description 7
- 230000004044 response Effects 0.000 description 8
- 230000003993 interaction Effects 0.000 description 7
- 230000009471 action Effects 0.000 description 5
- 208000018530 Disseminated peritoneal leiomyomatosis Diseases 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 4
- 238000013461 design Methods 0.000 description 4
- 238000012790 confirmation Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000032258 transport Effects 0.000 description 3
- 238000002372 labelling Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000003860 storage Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 239000012634 fragment Substances 0.000 description 1
- 230000008570 general process Effects 0.000 description 1
- 230000009191 jumping Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000007474 system interaction Effects 0.000 description 1
- 238000012546 transfer Methods 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
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/332—Query formulation
- G06F16/3329—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
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3344—Query execution using natural language analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Animal Behavior & Ethology (AREA)
- Human Computer Interaction (AREA)
- Machine Translation (AREA)
Abstract
The present invention discloses a kind of intelligent answer method based on natural language processing and knowledge mapping, in the dialog strategy study stage, pass through the study of the indispensable attributes to setting, identify uncertain words wheel, interact the uncertain words wheel with knowledge mapping, processing feedback is carried out by knowledge mapping, exports feedback result;It calls new scene to be intended to if recognizing user under current scene and having, new scene is jumped to from current scene by dialogue nesting and is engaged in the dialogue.The present invention is nested with dialogue by the reasoning for introducing knowledge mapping, solves the disadvantage that existing chat system technology, i.e., for take turns if unpredictable and scene in call other scenes that can not handle the problem of.
Description
Technical field
The present invention relates to field of artificial intelligence, more particularly to a kind of based on natural language processing and knowledge mapping
Intelligent answer method.
Background technique
Intelligent Answer System is generally divided into Task and two kinds of non task type depending on the application, and non task type mainly includes the spare time
Merely, knowledge question, recommendation.
The dialogue management (Dialogue Management, DM) in the dialogue of chat type in non task type is exactly to upper and lower
Text carries out Series Modeling, candidate's reply is scored, sorted and screened, in order to spatial term (Natural
Language Generation, NLG) stage generation preferably reply;Spatial term is the final step of dialogue, is to be
The answer united to user.Attention, intensified learning modeling can also be carried out based on deep learning.DM in the dialogue of knowledge question type
It is exactly the matching for the retrieval and knowledge base that text is carried out on the basis of the type identification of question sentence and classification, in order to NLG rank
The text fragments or knowledge base entity that Duan Shengcheng user wants.DM in recommendation type conversational system is exactly for carrying out user interest
Match and recommendation recalls, sorts, in order to which the NLG stage generates the content preferably recommended to user.
DM in Task dialogue is exactly to engage in the dialogue on the basis of NLU (domain classification and intention assessment, slot filling)
The tracking (Dialogue State Tracking, DST) of state and study (the Dialogue Policy of dialog strategy
Learning, DPL), in order to the study of DPL stage policy and NLG stage clarify demand, guidance user, inquiry, confirmation,
End-of-dialogue language etc..DST is according to current intention (intention), slot value to (slot-value pairs) and before
Be intended to track current state, it is simply that currently processed context.DPL is exactly according to current state
(state) movement of next step is determined.Task dialogue also can be long dialogue for short dialogue (Short Conversation)
(Long Conversation), short to talk with the single-wheel (single turn) for referring to question-response formula, long dialogue refers to you
Carry out my past more wheels (multi-turn), for Task dialogue, long dialogue is more common, and the dialogue of task with traditional type is general
Process is as shown in Figure 1.
The shortcomings that existing chat system technology is that Task dialogue and the dialogue of non task type lack harmonized programme, for nothing
Method is taken turns if predicting, that is, it is uncertain, unnecessary if take turns and scene in call in other scenes, lack the side of processing
Method.
Summary of the invention
The purpose of the present invention is lack for the dialogue of Task present in existing chat system technology and the dialogue of non task type
Weary harmonized programme, for taking turns if unpredictable, that is, it is uncertain, unnecessary if take turns and scene in call other fields
Jing Shang, lacks the method for processing, and provides a kind of intelligent answer method based on natural language processing and knowledge mapping.
The technical solution adopted to achieve the purpose of the present invention is:
A kind of intelligent answer method based on natural language processing and knowledge mapping is led in the dialog strategy study stage
The study to the indispensable attributes of setting is crossed, identifies uncertain words wheel, interacts the uncertain words wheel with knowledge mapping, by knowledge
Map carries out processing feedback, exports feedback result;
It calls new scene to be intended to if recognizing user under current scene and having, is jumped to by dialogue nesting from current scene
New scene engages in the dialogue.
Preferably, if some words wheel is not present in current scene, inquiry talks about wheel with the presence or absence of some in conversation history,
If it exists, then it jumps in this words wheel and this is talked about to the scene in the history that wheel executes and current scene is set to by nesting, then
Carry out the execution of this words wheel.
Preferably, the inquiry in conversation history is realized with the presence or absence of some words wheel by knowledge mapping.
Compared with prior art, the beneficial effects of the present invention are:
The present invention is nested with dialogue by the reasoning for introducing knowledge mapping in intelligent Answer System, solves existing chat
The shortcomings that its systems technology, i.e., for take turns if unpredictable and scene in call other scenes that can not handle the problem of.
Detailed description of the invention
Fig. 1 show the schematic diagram of the existing intelligent answer method based on natural language processing
Fig. 2 show the schematic diagram of the intelligent answer method based on natural language processing and knowledge mapping;
It is respectively the different schemes figure that intelligent answer rule design is carried out with finite-state automata shown in Fig. 3-5;
Fig. 6 show the schematic diagram of the invention that intelligent answer rule design is carried out with finite-state automata.
Specific embodiment
The present invention is described in further detail below in conjunction with the drawings and specific embodiments.It should be appreciated that described herein
Specific embodiment be only used to explain the present invention, be not intended to limit the present invention.
As shown in Fig. 2, the present invention is based on the intelligent answer methods of natural language processing and knowledge mapping, specifically right
Talk about in the policy learning stage, by the study of the indispensable attributes to setting, identify uncertain words wheel, make the uncertain words wheel with
Knowledge mapping interaction carries out processing feedback by knowledge mapping, exports feedback result;If recognizing user under current scene has calling
New scene is intended to, then jumps to new scene from current scene by dialogue nesting and engage in the dialogue.
In dialogue proposed by the present invention, wheel to unnecessary part (uncertain words are taken turns or taken turns for unforeseen words)
Processing, fully consider necessary part words wheel in, after the interaction of system and user, propose system and knowledge mapping friendship
Mutually, inquiry operation is carried out come the logic of analog subscriber by the logic of knowledge mapping, solved in existing dialogue for can not
The problem of wheel can not be handled if prediction.
Engage in the dialogue in addition, jumping to new scene by the nested current scene realized of dialogue, thus solve it is existing right
The technical issues of calling other scenes cannot achieve in scene in words.
Preferably, if some words wheel is not present in current scene, inquiry talks about wheel with the presence or absence of some in conversation history,
If it exists, then it jumps in this words wheel and this is talked about to the scene in the history that wheel executes and current scene is set to by nesting, then
Carry out the execution of this words wheel.
Below according to the technology of intelligent answer, realization of the invention is described in detail as follows:
In intelligent Answer System, data are structured storages, the data of structured storage according to be intended to or scene into
Row classification, including scene name, intention name, parameter list, response template.Wherein whether parameter list includes parameter name, must
It wants, three component parts of signal language.Response template is for the problem of responding user;Parameter list with from sequence labelling for taking out
The slot position taken is compared, and determines dialogue words wheel.Following example, for " transport cargo " this scene, " your cargo will be from
{ { parameters [" from "] } } transports to { { parameters [" to "] } } ".Here " from ", " to " is exactly parameter, signal language
Respectively " where your origin is? ", " where your destination is? ".It is intended to entitled " transport cargo.By aforementioned four word
In section deposit database, standby later query analysis.
The for another example scene of " reservation turns artificial ", response template are that " telephone number that you reserve is { { parameters [" electricity
Talk about number "], the business of consulting is { { parameters [" business "] } }, will have later contact staff's proactive contact you, please protect
Hold the unimpeded of phone." " telephone number " here, " business " is exactly parameter, is respectively set to necessity, and signal language is respectively " to ask
You leave telephone number? ", " the problem of you will seek advice from please be left? " it is intended to entitled " reservation turns artificial ".Above-mentioned field is stored in number
According in library, it is indispensable after query analysis.
Intelligent response system is generally divided into natural language processing, dialogue state tracking, dialog strategy study and natural language
Four processing stages are generated, it is answer of the system to user that spatial term, which is the final step of dialogue,.As shown in Figure 1.
Wherein, natural language processing be by user input natural language be mapped as user intention and corresponding slot position,
Slot value.Therefore the input of natural language processing is user session sentence, and output is user action obtained after parsing, user action
Including intention and slot position, the technology being related to is intended to identification and slot position identification and filling.It is intended to and slot position collectively forms that " user is dynamic
Make " because machine can not directly understand natural language, the effect of user action is that natural language is mapped as to machine energy
The structuring semantic expressiveness enough understood.
Intention assessment can by but be not limited to sorting algorithm complete.The parameter that slot position is namely intended to, an intention can
And have several or zero parameter, slot position identification can by but be not limited to sequence labelling complete.
Example:
User inputs example: " today, how is Pekinese's weather "
User's intent definition: inquiry weather
Slot position definition:
Slot position one: time
Slot position two: place
In the examples described above, for two necessary slot positions of " inquiry weather " intent definition, they are " time " respectively
" place ".The value of time is " today ", and the value in place is " Beijing ".
The input of dialogue state tracking (DST) is user action and pervious dialogue state, and output is dialogue state tracking
The current dialogue states of judgement.The expression (DST-State Representation) of dialogue state generally includes following three
Point: the filling situation of current slot position, the user action in epicycle dialog procedure, conversation history.
Dialog strategy study (DPL) input be DST output, by preset dialog strategy, select certain movement as
Output.Does for example how user input Beijing today weather? corresponding strategy is exactly all slot positions of one-off recognition, i.e. time
Slot position, slot value is today, the slot position in place, and slot value is Beijing.Secondly the strategy that can also be taken turns using words, such as: it uses
The input at family: inquiry weather.Is system putd question to: may I ask inquiry that day? user answers: today, slot position is exactly time, slot value
It is exactly today;System is inquired again: may I ask inquiry where? user answers: Pekinese, then its slot position is place, and slot value is Beijing.
Usually in intelligent response, intelligent response rule design is carried out using finite-state automata, there are two types of different
Scheme: the first, data are indicated with, and operation is indicated with side;Second, operation is indicated with, data are indicated with side.Scheme
One, as shown in figure 3, point indicates slot position state here, side indicates system acting.The filling of each dialogue state S expression slot position
The state transfer of state, slot position is caused by system acting.When slot position is sky, state is " sky ", is expressed as (0,0);When only
Between slot position when being filled state be (1,0), be expressed as when being all filled (1,1), be expressed as when only place is filled (0,
1).Shared slot position 2 in this example, 22A different state.After system acting issues the movement in " inquiry place ", state will turn
To " S2", if successfully filled, steering state " S2", if it fails, the movement of " inquiring again " will be initiated.Until identification
To state " (1,1) ".
The drawbacks of scheme one: with the increase of slot position quantity, the quantity of dialogue state can also be sharply increased.Specifically,
In the above scheme, the sum of dialogue state is determined by the number of slot position, if slot position has k, the quantity of dialogue state
It is 2kIt is a.
Scheme two indicates system acting in Fig. 4 with, indicates slot position state with side such as Fig. 4.In this case, limited
Each dialogue state S in state automata, expression is a kind of system acting, in this example system acting altogether there are three types of, respectively
It is that query time, is answered at inquiry place.The variation of state is as caused by the state change of slot position.
Above two scheme, second of finite-state automata design simpler, it is easier to work using system acting as core
Cheng Shixian.The first finite-state automata enumerates all slot position situations, is more suitable for data-driven using slot position state as core
Machine learning mode.Both schemes respectively have advantage and disadvantage, can be selected according to the actual situation in specific implementation.
In intelligent response, the definition of system acting generally includes three kinds of inquiry, confirmation, answer etc..Inquiry is that system is wanted
Solve the problems, such as that slot position lacks, confirmation is that system will solve the problems, such as fault-tolerance, and answer is the final reply of system.
For taking turns if uncertain, i.e., unnecessary words wheel, the unforeseen words of system are taken turns, and the present invention will be above-mentioned unnecessary
If take turns the interaction of part system and user and be transformed into the interaction of system and knowledge mapping, in general, user will be no longer participate in.
In addition above-mentioned to be interacted with user and all available with the interaction of other general character of the interaction of knowledge mapping in the case of other
Knowledge mapping analysis, inquiry, reasoning etc..Scheme one can be used to complete;It can also be by above two finite-state automata
Combine, operation is indicated with the point in scheme two, is incorporated in scheme one based on the finite-state automata of side expression data
Point indicate data, side indicate operation, as scheme but not limited to this.It is demonstrated used here as association schemes, such as Fig. 5 institute
Show.
In the following, for a example for buying train ticket, as shown in Figure 6, it is assumed that having time, the starting station, these necessity of terminus belong to
Property, similarly, these indispensable attributes can be by the interaction with user, to allow user to answer one by one for other indispensable attributes.But
If being that user says " train ticket that I will be most fast ", this requirement be exactly it is unnecessary if wheel since user is contemplated, I
Just solve this problem with knowledge mapping, particularly as being to inquire whether train ticket has using knowledge mapping come analysis ratiocination
This attribute.That is invention takes turns part in unnecessary words to replace people and system interaction with knowledge mapping.
If user wants to call new scene when talking in this scene, can be completed by nesting.Such as carry out
When attendant consultation service is talked with, user will inquire weather, be completed by " inquiry weather scene ".User needs to send " inquiry day
Gas " message, system can be complete in inquiry weather scene into " inquiry weather scene " by calling the identification of intention assessment model
At attendant consultation service scene is returned after corresponding dialogue, attendant consultation service scene can be further processed.In addition the nesting of multilayer is also
This reason.
In dialogue, some words wheel in scene is not present, and permission system is inquired in pervious conversation history whether there is
(can realize by knowledge mapping) is taken turns if current, and if it exists, then jumps in this words wheel and this is talked about what wheel executed
Scene is set to current scene, after this words wheel has executed, returns and previously takes turns.For multiple nestings and so on.
The above is only a preferred embodiment of the present invention, it is noted that for the common skill of the art
For art personnel, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications
Also it should be regarded as protection scope of the present invention.
Claims (3)
1. a kind of intelligent answer method based on natural language processing and knowledge mapping, which is characterized in that learn in dialog strategy
In stage, by the study of the indispensable attributes to setting, uncertain words wheel is identified, hand over the uncertain words wheel with knowledge mapping
Mutually, processing feedback is carried out by knowledge mapping, exports feedback result;
It calls new scene to be intended to if recognizing user under current scene and having, new field is jumped to from current scene by dialogue nesting
Scape engages in the dialogue.
2. the intelligent answer method based on natural language processing and knowledge mapping as described in claim 1, which is characterized in that if certain
A words wheel is not present in current scene, then inquiry is taken turns with the presence or absence of some word in conversation history, and if it exists, then jumps to this word
The scene talked about in the history that wheel executes in wheel and by this is set to current scene by nesting, then carries out the execution of this words wheel.
3. the intelligent answer method based on natural language processing and knowledge mapping as claimed in claim 2, which is characterized in that described
Inquiry is realized with the presence or absence of some words wheel by knowledge mapping in conversation history.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910388931.6A CN110222162A (en) | 2019-05-10 | 2019-05-10 | A kind of intelligent answer method based on natural language processing and knowledge mapping |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910388931.6A CN110222162A (en) | 2019-05-10 | 2019-05-10 | A kind of intelligent answer method based on natural language processing and knowledge mapping |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110222162A true CN110222162A (en) | 2019-09-10 |
Family
ID=67820634
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910388931.6A Pending CN110222162A (en) | 2019-05-10 | 2019-05-10 | A kind of intelligent answer method based on natural language processing and knowledge mapping |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110222162A (en) |
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110727776A (en) * | 2019-10-12 | 2020-01-24 | 一汽轿车股份有限公司 | Automobile question-answer interaction system and method based on artificial intelligence |
CN111104504A (en) * | 2019-12-25 | 2020-05-05 | 天津中科智能识别产业技术研究院有限公司 | Natural language processing and knowledge graph based dialogue method |
CN111400467A (en) * | 2020-03-09 | 2020-07-10 | 上海国民集团健康科技有限公司 | Robot chatting method |
CN111400508A (en) * | 2020-06-08 | 2020-07-10 | 科大讯飞(苏州)科技有限公司 | Necessary attribute determination method, related device and readable storage medium |
CN111611368A (en) * | 2020-05-22 | 2020-09-01 | 北京百度网讯科技有限公司 | Method and device for backtracking common scene conversation in multi-turn conversation |
CN111666388A (en) * | 2020-04-21 | 2020-09-15 | 文思海辉智科科技有限公司 | Dialogue data processing method, device, computer equipment and storage medium |
CN111816173A (en) * | 2020-06-01 | 2020-10-23 | 珠海格力电器股份有限公司 | Dialogue data processing method, device, storage medium and computer equipment |
CN112131360A (en) * | 2020-09-04 | 2020-12-25 | 交通银行股份有限公司太平洋***中心 | Intelligent multi-turn dialogue customization method and system |
JP2021018797A (en) * | 2019-07-23 | 2021-02-15 | バイドゥ オンライン ネットワーク テクノロジー (ベイジン) カンパニー リミテッド | Conversation interaction method, apparatus, computer readable storage medium, and program |
CN112597287A (en) * | 2020-12-15 | 2021-04-02 | 深圳市优必选科技股份有限公司 | Statement processing method, statement processing device and intelligent equipment |
CN112882679A (en) * | 2020-12-21 | 2021-06-01 | 广州橙行智动汽车科技有限公司 | Voice interaction method and device |
CN113050787A (en) * | 2019-12-26 | 2021-06-29 | 上海智臻智能网络科技股份有限公司 | Training method of man-machine conversation model and man-machine conversation method |
CN113569020A (en) * | 2021-06-22 | 2021-10-29 | 复旦大学 | Dialogue management method for fusing knowledge reasoning and knowledge guiding in intelligent conversation system |
CN113850570A (en) * | 2021-09-30 | 2021-12-28 | 中国建筑第七工程局有限公司 | AI-based professional scheme aided decision-making expert system construction method |
US12019990B2 (en) | 2019-12-17 | 2024-06-25 | Beijing Baidu Netcom Science Technology Co., Ltd. | Representation learning method and device based on natural language and knowledge graph |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107870994A (en) * | 2017-10-31 | 2018-04-03 | 北京光年无限科技有限公司 | Man-machine interaction method and system for intelligent robot |
CN107943998A (en) * | 2017-12-05 | 2018-04-20 | 竹间智能科技(上海)有限公司 | A kind of human-machine conversation control system and method for knowledge based collection of illustrative plates |
CN108000526A (en) * | 2017-11-21 | 2018-05-08 | 北京光年无限科技有限公司 | Dialogue exchange method and system for intelligent robot |
CN108256077A (en) * | 2018-01-18 | 2018-07-06 | 重庆邮电大学 | A kind of dynamic expansion knowledge graph inference method towards China Mobile's intelligent customer service |
CN108549662A (en) * | 2018-03-16 | 2018-09-18 | 北京云知声信息技术有限公司 | The supplement digestion procedure and device of semantic analysis result in more wheel sessions |
CN109271504A (en) * | 2018-11-07 | 2019-01-25 | 爱因互动科技发展(北京)有限公司 | The method of the reasoning dialogue of knowledge based map |
CN109616108A (en) * | 2018-11-29 | 2019-04-12 | 北京羽扇智信息科技有限公司 | More wheel dialogue interaction processing methods, device, electronic equipment and storage medium |
-
2019
- 2019-05-10 CN CN201910388931.6A patent/CN110222162A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107870994A (en) * | 2017-10-31 | 2018-04-03 | 北京光年无限科技有限公司 | Man-machine interaction method and system for intelligent robot |
CN108000526A (en) * | 2017-11-21 | 2018-05-08 | 北京光年无限科技有限公司 | Dialogue exchange method and system for intelligent robot |
CN107943998A (en) * | 2017-12-05 | 2018-04-20 | 竹间智能科技(上海)有限公司 | A kind of human-machine conversation control system and method for knowledge based collection of illustrative plates |
CN108256077A (en) * | 2018-01-18 | 2018-07-06 | 重庆邮电大学 | A kind of dynamic expansion knowledge graph inference method towards China Mobile's intelligent customer service |
CN108549662A (en) * | 2018-03-16 | 2018-09-18 | 北京云知声信息技术有限公司 | The supplement digestion procedure and device of semantic analysis result in more wheel sessions |
CN109271504A (en) * | 2018-11-07 | 2019-01-25 | 爱因互动科技发展(北京)有限公司 | The method of the reasoning dialogue of knowledge based map |
CN109616108A (en) * | 2018-11-29 | 2019-04-12 | 北京羽扇智信息科技有限公司 | More wheel dialogue interaction processing methods, device, electronic equipment and storage medium |
Cited By (25)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2021018797A (en) * | 2019-07-23 | 2021-02-15 | バイドゥ オンライン ネットワーク テクノロジー (ベイジン) カンパニー リミテッド | Conversation interaction method, apparatus, computer readable storage medium, and program |
JP7150770B2 (en) | 2019-07-23 | 2022-10-11 | バイドゥ オンライン ネットワーク テクノロジー(ペキン) カンパニー リミテッド | Interactive method, device, computer-readable storage medium, and program |
US11322153B2 (en) | 2019-07-23 | 2022-05-03 | Baidu Online Network Technology (Beijing) Co., Ltd. | Conversation interaction method, apparatus and computer readable storage medium |
CN110727776B (en) * | 2019-10-12 | 2023-05-12 | 一汽奔腾轿车有限公司 | Automobile question-answering interaction system and interaction method based on artificial intelligence |
CN110727776A (en) * | 2019-10-12 | 2020-01-24 | 一汽轿车股份有限公司 | Automobile question-answer interaction system and method based on artificial intelligence |
US12019990B2 (en) | 2019-12-17 | 2024-06-25 | Beijing Baidu Netcom Science Technology Co., Ltd. | Representation learning method and device based on natural language and knowledge graph |
CN111104504A (en) * | 2019-12-25 | 2020-05-05 | 天津中科智能识别产业技术研究院有限公司 | Natural language processing and knowledge graph based dialogue method |
CN113050787A (en) * | 2019-12-26 | 2021-06-29 | 上海智臻智能网络科技股份有限公司 | Training method of man-machine conversation model and man-machine conversation method |
CN111400467A (en) * | 2020-03-09 | 2020-07-10 | 上海国民集团健康科技有限公司 | Robot chatting method |
CN111400467B (en) * | 2020-03-09 | 2023-05-16 | 上海国民集团健康科技有限公司 | Robot chatting method |
CN111666388B (en) * | 2020-04-21 | 2023-11-10 | 文思海辉智科科技有限公司 | Dialogue data processing method, device, computer equipment and storage medium |
CN111666388A (en) * | 2020-04-21 | 2020-09-15 | 文思海辉智科科技有限公司 | Dialogue data processing method, device, computer equipment and storage medium |
CN111611368A (en) * | 2020-05-22 | 2020-09-01 | 北京百度网讯科技有限公司 | Method and device for backtracking common scene conversation in multi-turn conversation |
CN111816173A (en) * | 2020-06-01 | 2020-10-23 | 珠海格力电器股份有限公司 | Dialogue data processing method, device, storage medium and computer equipment |
CN111816173B (en) * | 2020-06-01 | 2024-06-07 | 珠海格力电器股份有限公司 | Dialogue data processing method and device, storage medium and computer equipment |
CN111400508B (en) * | 2020-06-08 | 2020-09-18 | 科大讯飞(苏州)科技有限公司 | Necessary attribute determination method, related device and readable storage medium |
CN111400508A (en) * | 2020-06-08 | 2020-07-10 | 科大讯飞(苏州)科技有限公司 | Necessary attribute determination method, related device and readable storage medium |
CN112131360A (en) * | 2020-09-04 | 2020-12-25 | 交通银行股份有限公司太平洋***中心 | Intelligent multi-turn dialogue customization method and system |
CN112597287A (en) * | 2020-12-15 | 2021-04-02 | 深圳市优必选科技股份有限公司 | Statement processing method, statement processing device and intelligent equipment |
CN112597287B (en) * | 2020-12-15 | 2024-03-15 | 深圳市优必选科技股份有限公司 | Statement processing method, statement processing device and intelligent equipment |
CN112882679B (en) * | 2020-12-21 | 2022-07-01 | 广州橙行智动汽车科技有限公司 | Voice interaction method and device |
CN112882679A (en) * | 2020-12-21 | 2021-06-01 | 广州橙行智动汽车科技有限公司 | Voice interaction method and device |
CN113569020B (en) * | 2021-06-22 | 2023-10-10 | 复旦大学 | Dialogue management method for integrating knowledge reasoning and knowledge guiding in intelligent dialogue system |
CN113569020A (en) * | 2021-06-22 | 2021-10-29 | 复旦大学 | Dialogue management method for fusing knowledge reasoning and knowledge guiding in intelligent conversation system |
CN113850570A (en) * | 2021-09-30 | 2021-12-28 | 中国建筑第七工程局有限公司 | AI-based professional scheme aided decision-making expert system construction method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110222162A (en) | A kind of intelligent answer method based on natural language processing and knowledge mapping | |
CN110704641B (en) | Ten-thousand-level intention classification method and device, storage medium and electronic equipment | |
CN109493166B (en) | Construction method for task type dialogue system aiming at e-commerce shopping guide scene | |
CN110175227B (en) | Dialogue auxiliary system based on team learning and hierarchical reasoning | |
CN107329967B (en) | Question answering system and method based on deep learning | |
CN102194005B (en) | Chat robot system and automatic chat method | |
CN111046132A (en) | Customer service question and answer processing method and system for retrieving multiple rounds of conversations | |
CN113515613A (en) | Intelligent robot integrating chatting, knowledge and task question answering | |
CN109688281A (en) | A kind of intelligent sound exchange method and system | |
CN104008160A (en) | Method and system of indistinct logic chatting robot for realizing parallel topic control | |
CN113268610B (en) | Intent jump method, device, equipment and storage medium based on knowledge graph | |
CN109829036A (en) | A kind of dialogue management method and relevant apparatus | |
CN109933654A (en) | A kind of dialogue management method based on State Tree | |
CN110321564A (en) | A kind of more wheel dialogue intension recognizing methods | |
CN116959433B (en) | Text processing method, device, electronic equipment and storage medium | |
CN109344242A (en) | A kind of dialogue answering method, device, equipment and storage medium | |
CN106127526A (en) | Intelligent robot system and method for work thereof | |
CN112364147A (en) | Cross-domain multi-turn dialogue method based on knowledge graph and implementation system | |
CN111400458A (en) | Automatic generalization method and device | |
Li et al. | The sppd system for schema guided dialogue state tracking challenge | |
CN112214589A (en) | Method for multi-round session framework based on cold start | |
CN111104504A (en) | Natural language processing and knowledge graph based dialogue method | |
CN110175733A (en) | A kind of public opinion information processing method and server | |
CN111625629A (en) | Task-based conversational robot response method, device, robot and storage medium | |
CN113139044A (en) | Question-answering multi-turn dialogue method supporting multi-intention switching for 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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190910 |
|
RJ01 | Rejection of invention patent application after publication |