CN109616108A - More wheel dialogue interaction processing methods, device, electronic equipment and storage medium - Google Patents
More wheel dialogue interaction processing methods, device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN109616108A CN109616108A CN201811446940.8A CN201811446940A CN109616108A CN 109616108 A CN109616108 A CN 109616108A CN 201811446940 A CN201811446940 A CN 201811446940A CN 109616108 A CN109616108 A CN 109616108A
- Authority
- CN
- China
- Prior art keywords
- interaction
- dialogue
- knowledge
- turn
- dialog
- 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
- 230000003993 interaction Effects 0.000 title claims abstract description 249
- 238000003672 processing method Methods 0.000 title claims abstract description 22
- 230000002452 interceptive effect Effects 0.000 claims abstract description 68
- 238000012545 processing Methods 0.000 claims abstract description 54
- 230000004044 response Effects 0.000 claims abstract description 43
- 238000000034 method Methods 0.000 claims abstract description 31
- 238000004458 analytical method Methods 0.000 claims description 13
- 238000013145 classification model Methods 0.000 claims description 7
- 238000004891 communication Methods 0.000 claims description 3
- 238000013507 mapping Methods 0.000 abstract 3
- 238000007726 management method Methods 0.000 description 13
- 238000010586 diagram Methods 0.000 description 11
- 240000001417 Vigna umbellata Species 0.000 description 10
- 235000011453 Vigna umbellata Nutrition 0.000 description 10
- 238000004590 computer program Methods 0.000 description 9
- 238000002372 labelling Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 230000006870 function Effects 0.000 description 5
- 230000008901 benefit Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000012546 transfer Methods 0.000 description 4
- 244000141359 Malus pumila Species 0.000 description 3
- 235000021016 apples Nutrition 0.000 description 3
- 235000013305 food Nutrition 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 235000013550 pizza Nutrition 0.000 description 3
- 240000009088 Fragaria x ananassa Species 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 235000013399 edible fruits Nutrition 0.000 description 2
- 230000010365 information processing Effects 0.000 description 2
- YNPNZTXNASCQKK-UHFFFAOYSA-N phenanthrene Chemical compound C1=CC=C2C3=CC=CC=C3C=CC2=C1 YNPNZTXNASCQKK-UHFFFAOYSA-N 0.000 description 2
- 235000021012 strawberries Nutrition 0.000 description 2
- 230000026676 system process Effects 0.000 description 2
- 244000046052 Phaseolus vulgaris Species 0.000 description 1
- 235000010627 Phaseolus vulgaris Nutrition 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000012528 membrane Substances 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000013179 statistical model Methods 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/06—Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice
- G10L15/063—Training
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/06—Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice
- G10L15/063—Training
- G10L2015/0631—Creating reference templates; Clustering
- G10L2015/0633—Creating reference templates; Clustering using lexical or orthographic knowledge sources
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/06—Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice
- G10L15/063—Training
- G10L2015/0635—Training updating or merging of old and new templates; Mean values; Weighting
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue
- G10L2015/225—Feedback of the input speech
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Artificial Intelligence (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Machine Translation (AREA)
Abstract
The embodiment of the invention discloses one kind to take turns dialogue interaction processing method, device, electronic equipment and storage medium more.This method comprises: obtaining the target interworking entity determined by current session round when terminating current session round；According to the renewal of knowledge corresponding with target interworking entity interaction knowledge mapping；Wherein, interaction knowledge mapping with by take turns more talk in terminated the corresponding knowledge connection of at least one interworking entity that dialog turns determine；According to the dictionary and/or model that at least one processing module uses in updated interactive knowledge mapping adjustment dialogue interactive system；When receiving dialog information corresponding with Next dialog round, response corresponding with dialog information is made using dialogue interactive system adjusted.It can enable to talk with through the above technical solution between each module of interactive system and the context of dialogue shares identical semantics information and knowledge each other, and then can permit conversation tasks and arbitrarily convert, reach seamless and integrate with.
Description
Technical Field
The present invention relates to information processing technologies, and in particular, to a method and an apparatus for processing multi-turn dialog interactions, an electronic device, and a storage medium.
Background
Man-machine conversation is a sub-direction in the field of artificial intelligence, and popular speaking is to allow people to interact with computers through human language (i.e. natural language).
A dialog interaction system usually needs to go through a series of processing modules from user dialog input to final response output, including speech recognition-intent classification-semantic slot sequence labeling-dialog management-candidate response acquisition-language generation, and the whole link is very long and complex.
In the specific implementation process, the inventor finds that information delivery in a conversation interaction system may have a situation of information inconsistency, and particularly, the situation is more likely to occur in multiple rounds of conversations, so that a poor experience is brought to a user. Exemplified with an ideal dialog interaction:
i want to listen to the song of Zheng Jun- > (system putting song) > who his lovers are- > (system answering singer's relevant knowledge question of Zheng Jun) > having performed TV shows- > (system answering Zheng Jun/Liu Yun's relevant knowledge question) > getting back to the pizza bar- > (system putting Zheng Jun's getting back to the pizza)
The above-described dialogue interaction is ideal, and may not be the case in practice. For example, the second round of the dialog of the user "who his lovers are", although the dialog management module can determine that "he" refers to "zheng jun" by referring to the resolution, if the knowledge map is simply queried according to the entity "zheng jun", it is found that the singer "zheng jun" and the physicist "zheng jun" exist in the knowledge map, and the dialog interaction system cannot necessarily select the correct entity "zheng jun" to answer. As another example, the dialog interaction system may also mistake the "pizza back" intent classification for the navigation task for the user's fourth turn of the dialog.
Disclosure of Invention
In view of this, embodiments of the present invention provide a method, an apparatus, an electronic device, and a storage medium for processing multi-round dialog interactions, so as to solve the problem of missing or inconsistent information transfer in a dialog interaction system.
In order to solve the above problems, embodiments of the present invention mainly provide the following technical solutions:
in a first aspect, an embodiment of the present invention provides a method for processing multiple rounds of dialog interactions, where the method includes:
when the current conversation turn is finished, acquiring a target interaction entity determined by the current conversation turn;
updating an interaction knowledge graph according to knowledge corresponding to the target interaction entity; wherein the interaction knowledge graph is associated with knowledge corresponding to at least one interaction entity determined by an ended conversation turn in a plurality of conversations;
adjusting a dictionary and/or a model used by at least one processing module in the dialogue interaction system according to the updated interaction knowledge graph;
and when receiving the dialogue information corresponding to the next dialogue turn, making a response corresponding to the dialogue information by using the adjusted dialogue interaction system.
In a second aspect, an embodiment of the present invention further provides a multi-turn dialog interaction processing apparatus, where the apparatus includes:
the interactive entity acquisition module is used for acquiring a target interactive entity determined by the current conversation turn when the current conversation turn is finished;
the interactive knowledge map updating module is used for updating the interactive knowledge map according to the knowledge corresponding to the target interactive entity; wherein the interaction knowledge graph is associated with knowledge corresponding to at least one interaction entity determined by an ended conversation turn in a plurality of conversations;
the dialogue interaction system adjusting module is used for adjusting a dictionary and/or a model used by at least one processing module in the dialogue interaction system according to the updated interaction knowledge map;
and the dialogue interaction system processing module is used for making a response corresponding to the dialogue information by using the adjusted dialogue interaction system when the dialogue information corresponding to the next dialogue turn is received.
In a third aspect, an embodiment of the present invention further provides an electronic device, including: at least one processor; and at least one memory, bus connected with the processor; the processor and the memory complete mutual communication through the bus; the processor is used for calling the program instructions in the memory so as to execute the multi-round dialogue interaction processing method in any embodiment of the invention.
In a fourth aspect, an embodiment of the present invention further provides a non-transitory computer-readable storage medium storing computer instructions, where the computer instructions cause the computer to execute the multi-turn dialog interaction processing method according to any embodiment of the present invention.
By the technical scheme, the technical scheme provided by the embodiment of the invention at least has the following advantages:
according to the embodiment of the invention, when the current conversation turn is finished each time, the target interaction entity determined by the current conversation turn is obtained, the interaction knowledge map is updated according to the knowledge corresponding to the target interaction entity, and then the dictionary and/or the model used by at least one processing module in the conversation interaction system is adjusted according to the updated interaction knowledge map, so that when the conversation information corresponding to the next conversation turn is received, the adjusted conversation interaction system can be used for making a corresponding answer. Through the technical scheme, the modules of the conversation interaction system and the conversation context can share the same semantic information and knowledge, so that the conversation tasks can be converted randomly, seamless connection is achieved, and the problem that the semantic information is lost or inconsistent across modules due to errors generated by transmitting the semantic information among the modules of the conversation interaction system is solved.
The foregoing description is only an overview of the technical solutions of the embodiments of the present invention, and the embodiments of the present invention can be implemented according to the content of the description in order to make the technical means of the embodiments of the present invention more clearly understood, and the detailed description of the embodiments of the present invention is provided below in order to make the foregoing and other objects, features, and advantages of the embodiments of the present invention more clearly understandable.
Drawings
Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the embodiments of the invention. Also, like reference numerals are used to refer to like parts throughout the drawings. In the drawings:
fig. 1 is a flowchart illustrating a multi-turn dialog interaction processing method according to an embodiment of the present invention;
fig. 2 is a flowchart illustrating a multi-turn dialog interaction processing method according to a second embodiment of the present invention;
fig. 3 is a flowchart illustrating a multi-turn dialog interaction processing method according to a third embodiment of the present invention;
fig. 4 is a schematic structural diagram illustrating a multi-turn dialogue interaction processing apparatus according to a fourth embodiment of the present invention;
fig. 5 shows a schematic structural diagram of an electronic device according to a fifth embodiment of the present invention.
Detailed Description
Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
Example one
Fig. 1 is a flowchart of a multi-turn dialog interaction processing method according to an embodiment of the present invention, which is applicable to a situation where semantic information transfer is inconsistent when a dialog interaction system processes multiple turns of dialogs, and the method may be executed by a multi-turn dialog interaction processing apparatus according to an embodiment of the present invention, where the apparatus may be implemented in software and/or hardware, and may be generally integrated in a processor. As shown in fig. 1, the method of the embodiment of the present invention specifically includes:
and S110, when the current conversation turn is finished, acquiring the target interaction entity determined by the current conversation turn.
Aiming at the situation that the man-machine interaction dialog is a plurality of turns of dialog, in order to keep semantic information between the plurality of turns of dialog consistent, the content of each turn of dialog of the user is analyzed, and the interactive content of the current turn of dialog is used as the basis for analyzing the dialog information of the next turn of dialog of the user, even the dialog information of a plurality of subsequent turns of dialog.
Therefore, when each round of conversation is finished, namely when the current conversation round is finished, namely the conversation interaction system makes a corresponding response to the current round of conversation of the current user, the target interaction entity in the round of conversation can be determined. The target interactive entity is specifically determined by the conversation content of the current conversation turn, and the entity can be information such as a person name, a place name, an organization name, a number, a date and the like.
For example, if the dialog information of the user in the current dialog turn is "i want to listen to the song of zheng jun", then the target interactive entity in the current dialog turn can be determined as "zheng jun" or "singer zheng jun" according to "zheng jun" and "song".
S120, updating the interaction knowledge map according to knowledge corresponding to the target interaction entity; wherein the interaction knowledge graph is associated with knowledge corresponding to at least one interaction entity determined by an ended conversation turn in the plurality of conversations.
The interaction knowledge map is associated with knowledge corresponding to the interaction entity, and the specific association is the knowledge corresponding to the interaction entity determined by the session turn which has been finished in the multiple sessions, so that when the current session turn is finished, the interaction knowledge map is updated according to the target interaction entity determined by the current session turn, namely the updated interaction knowledge map can be associated with the knowledge corresponding to the target interaction entity.
The following dialogue interaction is an example, "I want to listen to Zheng Jun song- > (system plays songs) - > who his lovers are- > (system answers relative knowledge questions of Zheng Jun)," after the first dialogue turn ends, the interactive knowledge map is associated with knowledge corresponding to the Zheng Jun, after the second turn ends, the interactive knowledge map is updated, and the updated interactive knowledge map is associated with knowledge corresponding to the Zheng Jun and the lovers who Zheng Jun.
It is worth noting that the interaction knowledge graph is dynamically updated, and the interaction knowledge graph may not be associated with any knowledge before the start of multiple rounds of conversation, and is dynamically updated as the content of the conversation increases. Specifically, the interaction knowledge graph may be emptied after the end of multiple sessions, or may be emptied after the end of the current session turn and under the condition that session information corresponding to the next session turn is not received after a set time period is exceeded.
And S130, adjusting a dictionary and/or a model used by at least one processing module in the dialogue interaction system according to the updated interaction knowledge graph.
Each module in the dialog interaction system generally needs to define a dictionary and derive a model based on the dictionary to complete a specific task, for example, the speech recognition module performs speech recognition on speech input by a user based on a speech recognition model and a language model, and for example, the intention classification module performs intention classification on text output by the speech recognition module based on a general domain dictionary and a task domain dictionary, and so on.
After the dictionaries and/or models used by the processing modules in the interactive system are adjusted according to the updated interactive knowledge map, the processing modules in the interactive system can share the same semantic information and knowledge.
For example, in the speech recognition module, the related extended vocabulary and the language model statistical information are obtained according to the updated mutual knowledge map and are fused with the basic dictionary and the general language model.
For another example, in modules such as intention classification/semantic groove sequence labeling, relevant extended entities are obtained according to the updated interactive knowledge map, and a domain relevant dictionary is dynamically loaded for feature extraction/feature weight modification of a Chinese word segmentation/statistical model, so that the intention classification and semantic groove sequence recognition rate is improved.
And S140, when receiving the dialogue information corresponding to the next dialogue turn, making a response corresponding to the dialogue information by using the adjusted dialogue interaction system.
When receiving the dialogue information corresponding to the next dialogue turn, the adjusted dialogue interaction system is used for processing, the same semantic information and knowledge can be shared among multiple turns of dialogue, and then response corresponding to the dialogue information and conforming to the user's intention is given.
As an example of a dialog interaction, "① i want to listen to zheng song- > (system dial song) - > ② his lover is- > (system answer to relevant knowledge problem of zheng jun of singer) - > ③ has played tv drama- > (system answer to relevant knowledge problem of zheng jun/liu yu) - > ④ comes back to the rasa bar- > (system dial back to rasa of zheng jun)," at the time of the ④ dialog turn, since the association in the interactive knowledge map is the knowledge corresponding to "zheng jun of singer" with lover of zheng jun of singer "and the like, and further the adjusted dialog system, at the time of the user dialog information processing of the ④ dialog turn, will deal with" going back to rasa "priority" of musing jun of singer "as a place name, and further will not sort" going back to rasa song "as a navigation task by mistake.
The multi-turn interactive dialogue processing method provided by the embodiment of the invention obtains the target interactive entity determined by the current dialogue turn when the current dialogue turn is finished each time, updates the interactive knowledge map according to the knowledge corresponding to the target interactive entity, further adjusts the dictionary and/or the model used by at least one processing module in the interactive dialogue system according to the updated interactive knowledge map, so that the processing modules can share the same semantic information and knowledge, further uses the adjusted interactive dialogue system to make corresponding answers when the dialogue information corresponding to the next dialogue turn is received, so that the dialogue contexts can share the same semantic information and knowledge with each other, allows the dialogue tasks to be converted randomly, achieves seamless connection, and avoids errors generated by the semantic information transmitted among the modules of the interactive dialogue system, thereby causing the problem of missing or inconsistent semantic information across modules.
Example two
Fig. 2 is a flowchart of a multi-turn dialog interaction processing method according to a second embodiment of the present invention. On the basis of the technical scheme, the embodiment of the invention updates the interaction knowledge map according to the knowledge corresponding to the target interaction entity, and specifically comprises the following steps:
inquiring a general knowledge map according to the target interaction entity to acquire knowledge corresponding to the target interaction entity; and updating the interaction knowledge map according to the knowledge corresponding to the target interaction entity.
As shown in fig. 2, the method provided in the embodiment of the present invention specifically includes:
s210, when the current conversation turn is finished, the target interaction entity determined by the current conversation turn is obtained.
S220, inquiring the general knowledge map according to the target interaction entity to acquire knowledge corresponding to the target interaction entity.
After the target interaction entity is determined according to the current conversation turn, the knowledge related to the target interaction entity is inquired in the general knowledge map, and then the knowledge is used as a reference basis for adjusting a dictionary and/or a model used by a processing module in the conversation interaction system.
And S230, updating the interaction knowledge map according to the knowledge corresponding to the target interaction entity, wherein the interaction knowledge map is associated with the knowledge corresponding to at least one interaction entity determined by the finished conversation turn in the multi-turn conversation.
And associating the acquired knowledge related to the target interaction entity to an interaction knowledge map, and further taking the interaction knowledge map as a reference basis for adjusting a dictionary and/or a model used by a processing module in the dialogue interaction system.
Specifically, the interactive knowledge map may associate the knowledge corresponding to the at least one interactive entity determined by the turn of the ended dialog in the multiple turns of the dialog in a manner that the interactive knowledge map includes the knowledge corresponding to the at least one interactive entity determined by the turn of the ended dialog in the multiple turns of the dialog, that is, an interactive knowledge map is constructed, the knowledge corresponding to the at least one interactive entity determined by the turn of the ended dialog in the multiple turns of the dialog is added to the interactive knowledge map, and then a dictionary and/or a model used by at least one processing module in the dialog interaction system may be adjusted according to the knowledge in the interactive knowledge map.
It may also be that the interaction knowledge map includes a link to knowledge corresponding to at least one interaction entity determined by an ended turn of a dialog in the multiple turns of dialogs, that is, the interaction knowledge map does not directly include knowledge, but the knowledge corresponding to the at least one interaction entity determined by the ended turn of the dialog in the multiple turns of dialogs may be obtained through the link in the interaction knowledge map, and then the corresponding knowledge may be obtained according to the link in the interaction knowledge map, thereby adjusting a dictionary and/or a model used by at least one processing module in the dialog interaction system.
And S240, adjusting the dictionary and/or the model used by at least one processing module in the dialogue interaction system according to the updated interaction knowledge graph.
For example, the determined target interaction entity is "singer zheng jun", and all knowledge corresponding to the entity "singer zheng jun" in the general knowledge graph is further associated to the interaction knowledge graph, such as which songs sung by the singer zheng jun are, who lovers are, what activities have recently been attended, and so on.
Furthermore, a dictionary and/or model used by at least one processing module in the dialog interaction system may be adjusted based on the interaction knowledge graph, wherein the processing module in the dialog interaction system comprises: the system comprises a voice recognition module, a semantic understanding module, a dialogue management module and a language generation module, wherein the semantic understanding module can be further specifically divided into an intention classification submodule and a semantic slot sequence labeling submodule.
The processing modules in the dialogue interaction system respectively correspond to the dictionaries and/or models to complete specific tasks, for example, the voice recognition module completes voice recognition tasks based on the voice recognition dictionaries and the voice models, the intention classification submodule completes intention classification tasks based on the general field dictionaries and the task field dictionaries, the semantic groove sequence labeling submodule completes semantic groove labeling tasks based on the task semantic groove dictionaries and the named entity dictionaries, the dialogue management module determines response operation of the computer for user dialogue information based on the semantic groove transfer inheritance dictionaries, the task semantic groove dictionaries and the named entity dictionaries, and the language generation module generates question responses corresponding to user dialogue information based on the language generation dictionaries and the task field dictionaries.
Taking a speech recognition module as an example, pronunciations of all words are stored in a speech recognition dictionary, the speech recognition dictionary is used for connecting an acoustic model and a language model, the words can be found out through the speech recognition dictionary when phonemes in user dialogue information are recognized, and the core of the language model is to predict the probability of occurrence of a sentence or a group of word sequences so as to generate text information corresponding to the user dialogue information.
Further, adjusting a lexicon and/or a model used by at least one processing module in the dialog interaction system according to the updated interaction knowledge graph, wherein the lexicon and/or the model comprises at least one of the following items:
adjusting a relevant dictionary and a language model weight in the voice recognition module according to the updated interactive knowledge map;
adjusting a relevant dictionary and an intention classification model weight in a semantic understanding module according to the updated interactive knowledge map; and
and adjusting the relevant dictionaries in the dialogue management module and the language generation module according to the updated interactive knowledge map.
For example, according to a target interaction entity 'singer Zheng Jun', a knowledge triplet 'Zheng Jun-lover-Liu Run' is obtained from the general knowledge map, then the updated interaction knowledge map can be associated with the knowledge triplet, and the voice recognition dictionary and the language model weight in the voice recognition module are adjusted according to the interaction knowledge map. If the speech recognition dictionary does not include "liuyun-liuyun", then "liuyun-liuyun" is added to the speech recognition dictionary, and the language model is adjusted to increase the prediction probability of the word "liuyun".
For another example, the interactive knowledge map updated according to the target interactive entity "singer royal membrane" may associate the knowledge triplet "royal membrane-song-red bean", and adjust the general field dictionary and task field dictionary in the intention classification submodule according to the interactive knowledge map, so that the intention classification submodule recognizes the "red bean" as the weight of the song field word is increased, and recognizes the "red bean" as the weight of the food field word is decreased.
And S250, when receiving the dialogue information corresponding to the next dialogue turn, making a response corresponding to the dialogue information by using the adjusted dialogue interaction system.
When receiving the dialogue information corresponding to the next dialogue turn, the adjusted dialogue interaction system is used for processing, the same semantic information and knowledge can be shared among multiple turns of dialogue, and then response corresponding to the dialogue information and conforming to the user's intention is given.
For example, when the dialogue interaction is "① i want to listen to song of royal pof- > (system dial song) - > ② red bean- > (system response)," the adjusted dialogue interaction system is used to respond to the user dialogue information "red bean" of ② interaction turn, the song red bean is played instead of providing some knowledge about food red bean.
It is worth pointing out that since the interaction knowledge map associates the knowledge corresponding to the interaction entity "singer joss" at the end of the ① th interaction turn, the adjusted dialog interaction system can still give correct responses when the user makes a dialog corresponding to the interaction entity "singer joss" at the n + m th interaction turn, after that, even if the user switches topics, the dialog is not related to the interaction entity "singer joss".
In the technical scheme, the conversation contexts can share the same semantic information and knowledge with each other, the conversation tasks are allowed to be converted randomly, seamless connection is achieved, and the problem that cross-module semantic information is lost or inconsistent due to errors caused by the fact that the semantic information is transmitted among modules of a conversation interaction system is solved.
EXAMPLE III
Fig. 3 is a flowchart of a multi-turn dialog interaction processing method according to a third embodiment of the present invention. On the basis of the above technical solution, the embodiment of the present invention uses the adjusted dialog interaction system to make a response corresponding to the dialog information, specifically:
performing voice recognition on the dialogue information based on the adjusted related dictionary and the adjusted voice model by using a voice recognition module in the adjusted dialogue interaction system;
performing semantic analysis on the recognition result output by the voice recognition by using a semantic understanding module in the adjusted dialogue interaction system based on the adjusted related dictionary and intention classification model;
performing reference resolution, semantic slot inheritance and task parameter analysis operations on analysis results output by a semantic understanding module by using a dialog management module in the adjusted dialog interaction system based on the adjusted related dictionary and the adjusted interaction knowledge map, generating candidate responses corresponding to dialog information, and making final responses corresponding to the dialog information;
and generating natural language text corresponding to the final response based on the adjusted related dictionary by using a language generation module in the adjusted dialogue interaction system.
As shown in fig. 3, the method provided in the embodiment of the present invention specifically includes:
and S310, when the current conversation turn is finished, acquiring the target interaction entity determined by the current conversation turn.
And S320, inquiring the general knowledge map according to the target interaction entity to acquire knowledge corresponding to the target interaction entity.
And S330, updating the interaction knowledge map according to the knowledge corresponding to the target interaction entity, wherein the interaction knowledge map is associated with the knowledge corresponding to at least one interaction entity determined by the finished conversation turn in the multi-turn conversation.
And S340, adjusting a dictionary and/or a model used by at least one processing module in the dialogue interaction system according to the updated interaction knowledge graph.
For the explanation of S310-S340, please refer to the foregoing embodiments, which are not described herein.
And S350, when receiving the dialogue information corresponding to the next dialogue turn, performing voice recognition on the dialogue information based on the adjusted related dictionary and the adjusted voice model by using the voice recognition module in the adjusted dialogue interaction system.
The voice recognition module recognizes the voice of the user into a text through the acoustic model and the language model, and after the dictionary and the language model corresponding to the voice recognition module are adjusted according to the updated interaction knowledge map, the voice recognition module can better accord with the intention of the user when recognizing the dialogue information of the next dialogue turn, thereby greatly improving the accuracy of the voice recognition.
And S360, performing semantic analysis on the recognition result output by the voice recognition based on the adjusted related dictionary and intention classification model by using a semantic understanding module in the adjusted dialogue interaction system.
And the intention classification submodule completes the intention classification task based on the adjusted general field dictionary and the task field dictionary, and then the semantic slot sequence labeling submodule completes the semantic slot sequence labeling task based on the task semantic slot dictionary and the named entity dictionary according to the intention classification result. Among them, entities such as a person name, a place name, a facility name, a number, a date, and the like are stored in the named entity dictionary.
Furthermore, a semantic analysis result of the semantic understanding module can be obtained, for example, from the sentence "how is the weather in the Mingtian day in Beijing," the semantic understanding module can obtain the following semantic results:
domain (Domain): weather (weather)
Intent (Intent): weather inquiry
Word Slot (Slot): city (city) Beijing
Day (date) is tomorrow
When most of knowledge associated with the interactive knowledge map is music knowledge, namely, when most of topics involved in multiple rounds of conversations of the user are music, if the user says a sentence of 'Chengdu', the semantic understanding module classifies the 'Chengdu' into songs instead of classifying the 'Chengdu' into places, and further intends to determine that the songs are played.
And S370, carrying out reference resolution, semantic slot inheritance and cross-task parameter analysis operations on the analysis result output by the semantic understanding module by using the adjusted conversation management module in the conversation interaction system based on the adjusted related dictionary and the adjusted interaction knowledge map, generating candidate responses corresponding to the conversation information, and further making a final response corresponding to the conversation information.
The session types of the candidate answers may include a task type, a chatting type and a question and answer type.
As a conversation interaction example, the "① i want to listen to the song of zheng- > (the system dials a song) - > ② his lovers are- > (the system answers the relevant knowledge problem of zheng jun of the singer) - > ③ has played a tv drama- > (the system answers the relevant knowledge problem of zheng jun/liu yung) - > ④ to get back to the rasa bar- > (the system dials back to rasa of zheng jun)," for the conversation of ② conversation turns "who his lovers are", the conversation management module in the adjusted conversation interaction system can determine not only that "he" refers to "zheng jun" by referring to the resolution, but also that "zheng jun" refers to "singer zheng jun" by interacting the knowledge map instead of "physicist", thereby being able to query the answer based on "zheng jun" in the general knowledge map to generate a correct response.
The reference resolution, the semantic slot inheritance and the task parameter resolution are all highly related to a specific task scene, for example, a purposeful (to) related entity is needed in a navigation task, a hotel search needs a location (location) entity, although the semantic slot names (to/location) of the two tasks are different, the dialogue management module can know that the two semantic slot indexes are entities of geographic information through knowledge related to an interaction knowledge map, and therefore, the ambiguity problem is easier to process.
For example, when "① i want to listen to song- > (system dial song) - > ② red bean- > (system response)" of royal phenanthrene, there is ambiguity in the user dialogue information "red bean" of the ② th interactive turn, and when the dialogue management module makes a response to the "red bean", the candidate response generated according to the "song red bean" is ranked before the candidate response generated according to the "food red bean" according to the associated knowledge in the interactive knowledge map, and the candidate response generated according to the "song red bean" is output as a final response.
And then, a chat task is performed, if the dialog system is provided with personal equipment, favorite fruits are apples, and a user asks a question that the user likes to eat the strawberries, the dialog system can know that the strawberries refer to the fruits and can simultaneously give a response that the user matched that the personal equipment in the personal equipment likes to apples, so that the user can correctly and accurately answer the response that the user likes to eat the apples and the like.
And S380, generating a natural language text corresponding to the final response based on the adjusted related dictionary by using the language generation module in the adjusted interactive dialogue system.
If the conversation management module determines that the next operation of the computer is to output voice information as a response corresponding to the user conversation, the language generation module is used for generating a natural language text based on the adjusted related dictionary, and then the natural language text is converted into the voice of a flow field by using a voice synthesis technology and is played to the user.
According to the technical scheme provided by the embodiment of the invention, the interactive knowledge map is used as the reference information to adjust the dictionaries and/or models used by each processing module in the dialogue interaction system so as to keep the semantic information transmitted among the modules of the dialogue interaction system consistent, and further, when the adjusted dialogue interaction system makes a response corresponding to the dialogue information by referring to the interactive knowledge map, the ambiguity problem is solved more easily.
Example four
Fig. 4 is a schematic structural diagram of a multi-turn dialog interaction processing apparatus according to a fourth embodiment of the present invention, which is applicable to a case where semantic information transfer is inconsistent when a dialog interaction system processes multiple turns of dialogs, and the apparatus may be implemented in software and/or hardware, and may be generally integrated in a processor. As shown in fig. 4, the apparatus specifically includes: an interactive entity acquisition module 410, an interactive knowledge graph update module 420, a dialog interaction system adaptation module 430, and a dialog interaction system processing module 440, wherein,
an interactive entity obtaining module 410, configured to obtain a target interactive entity determined by a current conversation turn when the current conversation turn is ended;
an interaction knowledge graph updating module 420, configured to update an interaction knowledge graph according to knowledge corresponding to the target interaction entity; wherein the interaction knowledge graph is associated with knowledge corresponding to at least one interaction entity determined by an ended conversation turn in a plurality of conversations;
a dialogue interaction system adjusting module 430, configured to adjust a dictionary and/or a model used by at least one processing module in the dialogue interaction system according to the updated interaction knowledge graph;
and the dialogue interaction system processing module 440 is configured to, when receiving the dialogue information corresponding to the next dialogue turn, make a response corresponding to the dialogue information using the adjusted dialogue interaction system.
The multi-turn dialogue interaction processing device provided by the embodiment of the invention obtains the target interaction entity determined by the current dialogue turn when the current dialogue turn is finished each time, updating the interactive knowledge map according to the knowledge corresponding to the target interactive entity, and further adjusting dictionaries and/or models used by at least one processing module in the interactive system according to the updated interactive knowledge map so that the processing modules can share the same semantic information and knowledge, and then when receiving the dialogue information corresponding to the next dialogue turn, using the adjusted dialogue interaction system to make a corresponding answer, therefore, the conversation contexts can share the same semantic information and knowledge with each other, the conversation tasks are allowed to be converted randomly, seamless connection is achieved, and the problem that the semantic information is lost or inconsistent across modules due to errors generated by the transmission of the semantic information among modules of a conversation interaction system is solved.
Further, the interaction knowledge map updating module 420 is specifically configured to query a general knowledge map according to the target interaction entity, and acquire knowledge corresponding to the target interaction entity; and updating the interaction knowledge graph according to the knowledge corresponding to the target interaction entity.
Wherein the interaction knowledge-graph comprises knowledge corresponding to at least one interaction entity determined by an ended conversation turn in a plurality of conversations, or,
the interaction knowledge graph includes links to knowledge corresponding to at least one interactive entity determined by an ended turn of dialog in a plurality of turns of dialog.
Specifically, the processing module in the dialog interaction system includes: the system comprises a voice recognition module, a semantic understanding module, a dialogue management module and a language generation module.
Further, the dialog interaction system adjustment module 430 specifically includes at least one unit,
the first adjusting unit is used for adjusting a relevant dictionary and language model weight in the voice recognition module according to the updated interaction knowledge graph;
the second adjusting unit is used for adjusting the relevant dictionary and the intention classification model weight in the semantic understanding module according to the updated interactive knowledge map; and
and the third adjusting unit is used for adjusting the relevant dictionaries in the dialogue management module and the language generation module according to the updated interaction knowledge graph.
Specifically, the dialogue interaction system processing module 440 is configured to, when receiving dialogue information corresponding to a next dialogue turn, perform voice recognition on the dialogue information based on the adjusted related dictionary and voice model by using the adjusted voice recognition module in the dialogue interaction system;
performing semantic analysis on the recognition result output by the voice recognition based on the adjusted related dictionary and intention classification model by using an adjusted semantic understanding module in the dialogue interaction system;
performing reference resolution, semantic slot inheritance and task parameter analysis operations on an analysis result output by the semantic understanding module by using an adjusted related dictionary and the adjusted interaction knowledge map by using an adjusted conversation management module in the conversation interaction system, generating a candidate response corresponding to the conversation information, and further making a final response corresponding to the conversation information;
and generating a natural language text corresponding to the final response based on the adjusted related dictionary by using the adjusted language generation module in the dialogue interaction system.
Wherein the session type of the candidate answer includes at least one of: task type, chatting type, and question and answer type.
Since the multi-turn dialog interaction processing apparatus described in the embodiment of the present invention is an apparatus capable of executing the multi-turn dialog interaction processing method described in the embodiment of the present invention, based on the multi-turn dialog interaction processing method described in the embodiment of the present invention, a person skilled in the art can understand a specific implementation manner of the multi-turn dialog interaction processing apparatus described in the embodiment of the present invention and various variations thereof, so how the multi-turn dialog interaction processing apparatus implements the multi-turn dialog interaction processing method described in the embodiment of the present invention is not described in detail herein. The scope of the present application is intended to be covered by the following claims as long as those skilled in the art can implement the apparatus for performing the multi-turn interactive processing method in the embodiments of the present invention.
EXAMPLE five
An embodiment of the present invention provides an electronic device, as shown in fig. 5, including: at least one processor (processor) 51; and at least one memory (memory)52, a bus 53 connected to the processor 51; wherein,
the processor 51 and the memory 52 complete mutual communication through the bus 53;
the memory 52 is a non-transitory computer-readable storage medium, and can be used for storing software programs, computer-executable programs, and modules, such as program instructions/modules corresponding to a multi-turn dialog interaction processing method according to an embodiment of the present invention (for example, as shown in fig. 4, the interactive entity acquiring module 410, the interactive knowledge map updating module 420, the dialog interaction system adjusting module 430, and the dialog interaction system processing module 440). The processor 51 is configured to call program instructions/modules in the memory 52 to execute the steps in one of the multi-round dialog interaction processing methods in the above-described method embodiments.
The memory 52 may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store data created according to use of the electronic device, and the like. Further, the memory 52 may include high speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, the memory 52 may optionally include memory located remotely from the processor 51, which may be connected to the terminal device via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
EXAMPLE six
An embodiment of the present invention provides a non-transitory computer-readable storage medium, where the non-transitory computer-readable storage medium stores computer instructions, where the computer instructions cause the computer to execute a multi-turn dialog interaction processing method provided in the foregoing method embodiments, where the method includes: .
When the current conversation turn is finished, acquiring a target interaction entity determined by the current conversation turn;
updating an interaction knowledge graph according to knowledge corresponding to the target interaction entity; wherein the interaction knowledge graph is associated with knowledge corresponding to at least one interaction entity determined by an ended conversation turn in a plurality of conversations;
adjusting a dictionary and/or a model used by at least one processing module in the dialogue interaction system according to the updated interaction knowledge graph;
and when receiving the dialogue information corresponding to the next dialogue turn, making a response corresponding to the dialogue information by using the adjusted dialogue interaction system.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include forms of volatile memory in a computer readable medium, Random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). The memory is an example of a computer-readable medium.
Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in the process, method, article, or apparatus that comprises the element.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The above embodiments are merely examples of the present application, and are not intended to limit the present application, and the technical features of the embodiments may be combined and arranged within the scope of the present invention. Various modifications and changes may occur to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.
Claims (10)
1. A multi-turn dialogue interaction processing method is characterized by comprising the following steps:
when the current conversation turn is finished, acquiring a target interaction entity determined by the current conversation turn;
updating an interaction knowledge graph according to knowledge corresponding to the target interaction entity; wherein the interaction knowledge graph is associated with knowledge corresponding to at least one interaction entity determined by an ended conversation turn in a plurality of conversations;
adjusting a dictionary and/or a model used by at least one processing module in the dialogue interaction system according to the updated interaction knowledge graph;
and when receiving the dialogue information corresponding to the next dialogue turn, making a response corresponding to the dialogue information by using the adjusted dialogue interaction system.
2. The method of claim 1, wherein updating the interaction knowledge graph based on knowledge corresponding to the target interaction entity comprises:
inquiring a general knowledge graph according to the target interaction entity to acquire knowledge corresponding to the target interaction entity;
and updating the interaction knowledge graph according to the knowledge corresponding to the target interaction entity.
3. The method of claim 1 or 2, wherein the interaction knowledge graph comprises knowledge corresponding to at least one interaction entity determined by an ended dialog turn in a plurality of dialogs, or,
the interaction knowledge graph includes links to knowledge corresponding to at least one interactive entity determined by an ended turn of dialog in a plurality of turns of dialog.
4. The method of claim 1, wherein the processing module in the dialog interaction system comprises: the system comprises a voice recognition module, a semantic understanding module, a dialogue management module and a language generation module.
5. The method of claim 4, wherein the adjusting the lexicon and/or model used by at least one processing module in the dialog interaction system according to the updated interaction knowledge graph comprises at least one of the following:
adjusting a relevant dictionary and a language model weight in the voice recognition module according to the updated interaction knowledge graph;
adjusting a relevant dictionary and an intention classification model weight in the semantic understanding module according to the updated interactive knowledge map; and
and adjusting the relevant dictionaries in the dialogue management module and the language generation module according to the updated interaction knowledge graph.
6. The method of claim 5, wherein making a response corresponding to the dialog information using the adjusted dialog interaction system comprises:
performing voice recognition on the dialogue information based on the adjusted relevant dictionary and the adjusted voice model by using the adjusted voice recognition module in the dialogue interaction system;
performing semantic analysis on the recognition result output by the voice recognition based on the adjusted related dictionary and intention classification model by using an adjusted semantic understanding module in the dialogue interaction system;
performing reference resolution, semantic slot inheritance and task parameter analysis operations on an analysis result output by the semantic understanding module by using an adjusted related dictionary and the adjusted interaction knowledge map by using an adjusted conversation management module in the conversation interaction system, generating a candidate response corresponding to the conversation information, and further making a final response corresponding to the conversation information;
and generating a natural language text corresponding to the final response based on the adjusted related dictionary by using the adjusted language generation module in the dialogue interaction system.
7. The method of claim 6, wherein the session type of the candidate answer comprises at least one of: task type, chatting type, and question and answer type.
8. A multi-turn dialog interaction processing apparatus, comprising:
the interactive entity acquisition module is used for acquiring a target interactive entity determined by the current conversation turn when the current conversation turn is finished;
the interactive knowledge map updating module is used for updating the interactive knowledge map according to the knowledge corresponding to the target interactive entity; wherein the interaction knowledge graph is associated with knowledge corresponding to at least one interaction entity determined by an ended conversation turn in a plurality of conversations;
the dialogue interaction system adjusting module is used for adjusting a dictionary and/or a model used by at least one processing module in the dialogue interaction system according to the updated interaction knowledge map;
and the dialogue interaction system processing module is used for making a response corresponding to the dialogue information by using the adjusted dialogue interaction system when the dialogue information corresponding to the next dialogue turn is received.
9. An electronic device, comprising:
at least one processor;
and at least one memory, bus connected with the processor; wherein,
the processor and the memory complete mutual communication through the bus;
the processor is used for calling the program instructions in the memory to execute the multi-turn dialogue interaction processing method of any one of claim 1 to claim 7.
10. A non-transitory computer-readable storage medium storing computer instructions for causing a computer to perform the multi-turn dialog interaction processing method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811446940.8A CN109616108B (en) | 2018-11-29 | 2018-11-29 | Multi-turn dialogue interaction processing method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811446940.8A CN109616108B (en) | 2018-11-29 | 2018-11-29 | Multi-turn dialogue interaction processing method and device, electronic equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109616108A true CN109616108A (en) | 2019-04-12 |
| CN109616108B CN109616108B (en) | 2022-05-31 |
Family
ID=66005502
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811446940.8A Active CN109616108B (en) | 2018-11-29 | 2018-11-29 | Multi-turn dialogue interaction processing method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109616108B (en) |
Cited By (38)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110032633A (en) * | 2019-04-17 | 2019-07-19 | 腾讯科技（深圳）有限公司 | More wheel dialog process method, apparatus and equipment |
| CN110111795A (en) * | 2019-04-23 | 2019-08-09 | 维沃移动通信有限公司 | A kind of method of speech processing and terminal device |
| CN110175226A (en) * | 2019-05-09 | 2019-08-27 | 厦门邑通软件科技有限公司 | A kind of dialogue decision-making technique based on various dimensions scene analysis |
| CN110222162A (en) * | 2019-05-10 | 2019-09-10 | 天津中科智能识别产业技术研究院有限公司 | A kind of intelligent answer method based on natural language processing and knowledge mapping |
| CN110347816A (en) * | 2019-07-15 | 2019-10-18 | 腾讯科技（上海）有限公司 | A kind of information recommendation method and device |
| CN110543556A (en) * | 2019-09-05 | 2019-12-06 | 出门问问信息科技有限公司 | Dialogue configuration method, storage medium and electronic equipment |
| CN110706702A (en) * | 2019-10-15 | 2020-01-17 | 芜湖盟博科技有限公司 | Infinite stage multi-turn dialogue method for speech recognition |
| CN110737765A (en) * | 2019-10-25 | 2020-01-31 | 上海喜马拉雅科技有限公司 | Dialogue data processing method for multi-turn dialogue and related device |
| CN110827831A (en) * | 2019-11-15 | 2020-02-21 | 广州洪荒智能科技有限公司 | Voice information processing method, device, equipment and medium based on man-machine interaction |
| CN110909159A (en) * | 2019-09-29 | 2020-03-24 | 珠海格力电器股份有限公司 | User intention identification method, device, terminal and storage medium |
| CN111026886A (en) * | 2019-12-26 | 2020-04-17 | 成都航天科工大数据研究院有限公司 | Multi-round dialogue processing method for professional scene |
| CN111178940A (en) * | 2019-12-11 | 2020-05-19 | 北京明略软件系统有限公司 | Method and system for automatically generating sales call technology map |
| CN111177358A (en) * | 2019-12-31 | 2020-05-19 | 华为技术有限公司 | Intent recognition method, server and storage medium |
| CN111222323A (en) * | 2019-12-30 | 2020-06-02 | 深圳市优必选科技股份有限公司 | Word slot extraction method, word slot extraction device and electronic equipment |
| CN111341308A (en) * | 2020-02-12 | 2020-06-26 | 百度在线网络技术（北京）有限公司 | Method and apparatus for outputting information |
| CN111339283A (en) * | 2020-05-15 | 2020-06-26 | 支付宝(杭州)信息技术有限公司 | Method and device for providing customer service answers aiming at user questions |
| CN111651615A (en) * | 2020-07-24 | 2020-09-11 | 上海茂声智能科技有限公司 | A method and system for human-computer interaction based on knowledge graph |
| WO2020216064A1 (en) * | 2019-04-24 | 2020-10-29 | 京东方科技集团股份有限公司 | Speech emotion recognition method, semantic recognition method, question-answering method, computer device and computer-readable storage medium |
| CN111966803A (en) * | 2020-08-03 | 2020-11-20 | 深圳市欢太科技有限公司 | Dialogue simulation method, dialogue simulation device, storage medium and electronic equipment |
| WO2020238341A1 (en) * | 2019-05-31 | 2020-12-03 | 华为技术有限公司 | Speech recognition method, apparatus and device, and computer-readable storage medium |
| CN112164401A (en) * | 2020-09-18 | 2021-01-01 | 广州小鹏汽车科技有限公司 | Voice interaction method, server and computer-readable storage medium |
| CN112215010A (en) * | 2019-07-10 | 2021-01-12 | 北京猎户星空科技有限公司 | Semantic recognition method and equipment |
| CN112364147A (en) * | 2020-12-01 | 2021-02-12 | 四川长虹电器股份有限公司 | Cross-domain multi-turn dialogue method based on knowledge graph and implementation system |
| CN112445902A (en) * | 2019-09-04 | 2021-03-05 | 深圳Tcl数字技术有限公司 | Method for identifying user intention in multi-turn conversation and related equipment |
| CN112463939A (en) * | 2020-11-12 | 2021-03-09 | 深圳市欢太科技有限公司 | Man-machine conversation method, system, service device and computer storage medium |
| CN112487137A (en) * | 2019-09-12 | 2021-03-12 | 甲骨文国际公司 | Streamlining conversation processing using integrated shared resources |
| CN112581955A (en) * | 2020-11-30 | 2021-03-30 | 广州橙行智动汽车科技有限公司 | Voice control method, server, voice control system and readable storage medium |
| CN112911073A (en) * | 2019-04-30 | 2021-06-04 | 五竹科技(北京)有限公司 | Intelligent knowledge graph construction method and device for outbound process conversation content |
| CN113470625A (en) * | 2021-06-30 | 2021-10-01 | 西安力传智能技术有限公司 | Voice conversation processing method, device, equipment and storage medium |
| CN113868428A (en) * | 2020-06-30 | 2021-12-31 | 中国移动通信有限公司研究院 | Service recommendation method and device, electronic equipment and computer readable storage medium |
| CN114144773A (en) * | 2019-08-01 | 2022-03-04 | 国际商业机器公司 | Adjusting conversational flow based on behavior in human-machine cognitive interactions |
| CN115081454A (en) * | 2021-03-11 | 2022-09-20 | 阿里巴巴新加坡控股有限公司 | Dialogue understanding method and device, electronic equipment and storage medium |
| CN115186147A (en) * | 2022-05-31 | 2022-10-14 | 华院计算技术(上海)股份有限公司 | Method and device for generating conversation content, storage medium and terminal |
| CN115204188A (en) * | 2022-05-27 | 2022-10-18 | 天翼电子商务有限公司 | A method of time text extraction and calculation in multi-round interaction of intelligent outbound calls |
| CN115394293A (en) * | 2022-08-08 | 2022-11-25 | 湖北星纪时代科技有限公司 | Dialogue system and method for implementing dialogues |
| CN116501592A (en) * | 2023-06-19 | 2023-07-28 | 阿里巴巴（中国）有限公司 | Man-machine interaction data processing method and server |
| CN117496973A (en) * | 2024-01-02 | 2024-02-02 | 四川蜀天信息技术有限公司 | Method, device, equipment and medium for improving man-machine conversation interaction experience |
| CN118627631A (en) * | 2024-08-14 | 2024-09-10 | 中电数据产业集团有限公司 | Multi-round dialogue method, device, equipment and storage medium based on DST |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160163311A1 (en) * | 2014-12-09 | 2016-06-09 | Microsoft Technology Licensing, Llc | Communication system |
| CN106649694A (en) * | 2016-12-19 | 2017-05-10 | 北京云知声信息技术有限公司 | Method and device for identifying user's intention in voice interaction |
| CN107704612A (en) * | 2017-10-23 | 2018-02-16 | 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 |
-
2018
- 2018-11-29 CN CN201811446940.8A patent/CN109616108B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160163311A1 (en) * | 2014-12-09 | 2016-06-09 | Microsoft Technology Licensing, Llc | Communication system |
| CN106649694A (en) * | 2016-12-19 | 2017-05-10 | 北京云知声信息技术有限公司 | Method and device for identifying user's intention in voice interaction |
| CN107704612A (en) * | 2017-10-23 | 2018-02-16 | 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 |
Cited By (61)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110032633A (en) * | 2019-04-17 | 2019-07-19 | 腾讯科技（深圳）有限公司 | More wheel dialog process method, apparatus and equipment |
| CN110032633B (en) * | 2019-04-17 | 2021-03-23 | 腾讯科技（深圳）有限公司 | Multi-turn dialogue processing method, device and equipment |
| CN110111795B (en) * | 2019-04-23 | 2021-08-27 | 维沃移动通信有限公司 | Voice processing method and terminal equipment |
| CN110111795A (en) * | 2019-04-23 | 2019-08-09 | 维沃移动通信有限公司 | A kind of method of speech processing and terminal device |
| WO2020216064A1 (en) * | 2019-04-24 | 2020-10-29 | 京东方科技集团股份有限公司 | Speech emotion recognition method, semantic recognition method, question-answering method, computer device and computer-readable storage medium |
| CN112911073A (en) * | 2019-04-30 | 2021-06-04 | 五竹科技(北京)有限公司 | Intelligent knowledge graph construction method and device for outbound process conversation content |
| CN110175226A (en) * | 2019-05-09 | 2019-08-27 | 厦门邑通软件科技有限公司 | A kind of dialogue decision-making technique based on various dimensions scene analysis |
| CN110222162A (en) * | 2019-05-10 | 2019-09-10 | 天津中科智能识别产业技术研究院有限公司 | A kind of intelligent answer method based on natural language processing and knowledge mapping |
| WO2020238341A1 (en) * | 2019-05-31 | 2020-12-03 | 华为技术有限公司 | Speech recognition method, apparatus and device, and computer-readable storage medium |
| US12087289B2 (en) | 2019-05-31 | 2024-09-10 | Huawei Technologies Co., Ltd. | Speech recognition method, apparatus, and device, and computer-readable storage medium |
| CN112215010A (en) * | 2019-07-10 | 2021-01-12 | 北京猎户星空科技有限公司 | Semantic recognition method and equipment |
| CN110347816B (en) * | 2019-07-15 | 2023-08-04 | 腾讯科技（上海）有限公司 | Information recommendation method and device |
| CN110347816A (en) * | 2019-07-15 | 2019-10-18 | 腾讯科技（上海）有限公司 | A kind of information recommendation method and device |
| CN114144773B (en) * | 2019-08-01 | 2022-10-28 | 国际商业机器公司 | Adjusting conversational flow based on behavior in human-machine cognitive interactions |
| CN114144773A (en) * | 2019-08-01 | 2022-03-04 | 国际商业机器公司 | Adjusting conversational flow based on behavior in human-machine cognitive interactions |
| CN112445902B (en) * | 2019-09-04 | 2024-12-03 | 深圳Tcl数字技术有限公司 | A method for identifying user intention in multi-round dialogue and related equipment |
| CN112445902A (en) * | 2019-09-04 | 2021-03-05 | 深圳Tcl数字技术有限公司 | Method for identifying user intention in multi-turn conversation and related equipment |
| CN110543556A (en) * | 2019-09-05 | 2019-12-06 | 出门问问信息科技有限公司 | Dialogue configuration method, storage medium and electronic equipment |
| CN112487137B (en) * | 2019-09-12 | 2023-11-03 | 甲骨文国际公司 | Streamlining dialog processing using integrated shared resources |
| CN112487137A (en) * | 2019-09-12 | 2021-03-12 | 甲骨文国际公司 | Streamlining conversation processing using integrated shared resources |
| CN110909159A (en) * | 2019-09-29 | 2020-03-24 | 珠海格力电器股份有限公司 | User intention identification method, device, terminal and storage medium |
| CN110706702B (en) * | 2019-10-15 | 2021-11-09 | 芜湖盟博科技有限公司 | Infinite stage multi-turn dialogue method for speech recognition |
| CN110706702A (en) * | 2019-10-15 | 2020-01-17 | 芜湖盟博科技有限公司 | Infinite stage multi-turn dialogue method for speech recognition |
| CN110737765A (en) * | 2019-10-25 | 2020-01-31 | 上海喜马拉雅科技有限公司 | Dialogue data processing method for multi-turn dialogue and related device |
| CN110827831A (en) * | 2019-11-15 | 2020-02-21 | 广州洪荒智能科技有限公司 | Voice information processing method, device, equipment and medium based on man-machine interaction |
| CN111178940B (en) * | 2019-12-11 | 2023-09-01 | 北京明略软件系统有限公司 | Method and system for automatically generating sales call operation map |
| CN111178940A (en) * | 2019-12-11 | 2020-05-19 | 北京明略软件系统有限公司 | Method and system for automatically generating sales call technology map |
| CN111026886B (en) * | 2019-12-26 | 2023-05-02 | 成都航天科工大数据研究院有限公司 | Multi-round dialogue processing method for professional scene |
| CN111026886A (en) * | 2019-12-26 | 2020-04-17 | 成都航天科工大数据研究院有限公司 | Multi-round dialogue processing method for professional scene |
| CN111222323B (en) * | 2019-12-30 | 2024-05-03 | 深圳市优必选科技股份有限公司 | Word slot extraction method, word slot extraction device and electronic equipment |
| CN111222323A (en) * | 2019-12-30 | 2020-06-02 | 深圳市优必选科技股份有限公司 | Word slot extraction method, word slot extraction device and electronic equipment |
| CN111177358B (en) * | 2019-12-31 | 2023-05-12 | 华为技术有限公司 | Intent recognition method, server and storage medium |
| CN111177358A (en) * | 2019-12-31 | 2020-05-19 | 华为技术有限公司 | Intent recognition method, server and storage medium |
| CN111341308B (en) * | 2020-02-12 | 2023-07-21 | 百度在线网络技术（北京）有限公司 | Method and device for outputting information |
| CN111341308A (en) * | 2020-02-12 | 2020-06-26 | 百度在线网络技术（北京）有限公司 | Method and apparatus for outputting information |
| CN111339283B (en) * | 2020-05-15 | 2020-08-21 | 支付宝(杭州)信息技术有限公司 | Method and device for providing customer service answers aiming at user questions |
| CN111339283A (en) * | 2020-05-15 | 2020-06-26 | 支付宝(杭州)信息技术有限公司 | Method and device for providing customer service answers aiming at user questions |
| CN113868428B (en) * | 2020-06-30 | 2025-05-13 | 中国移动通信有限公司研究院 | Business recommendation method, device, electronic device and computer-readable storage medium |
| CN113868428A (en) * | 2020-06-30 | 2021-12-31 | 中国移动通信有限公司研究院 | Service recommendation method and device, electronic equipment and computer readable storage medium |
| CN111651615B (en) * | 2020-07-24 | 2023-06-13 | 上海茂声智能科技有限公司 | Method and system for human-computer interaction based on knowledge graph |
| CN111651615A (en) * | 2020-07-24 | 2020-09-11 | 上海茂声智能科技有限公司 | A method and system for human-computer interaction based on knowledge graph |
| CN111966803A (en) * | 2020-08-03 | 2020-11-20 | 深圳市欢太科技有限公司 | Dialogue simulation method, dialogue simulation device, storage medium and electronic equipment |
| CN111966803B (en) * | 2020-08-03 | 2024-04-12 | 深圳市欢太科技有限公司 | Dialogue simulation method, device, storage medium and electronic device |
| CN112164401B (en) * | 2020-09-18 | 2022-03-18 | 广州小鹏汽车科技有限公司 | Voice interaction method, server and computer-readable storage medium |
| CN112164401A (en) * | 2020-09-18 | 2021-01-01 | 广州小鹏汽车科技有限公司 | Voice interaction method, server and computer-readable storage medium |
| CN112463939B (en) * | 2020-11-12 | 2024-05-24 | 深圳市欢太科技有限公司 | Human-computer dialogue method, system, service equipment and computer storage medium |
| CN112463939A (en) * | 2020-11-12 | 2021-03-09 | 深圳市欢太科技有限公司 | Man-machine conversation method, system, service device and computer storage medium |
| CN112581955B (en) * | 2020-11-30 | 2024-03-08 | 广州橙行智动汽车科技有限公司 | Voice control method, server, voice control system, and readable storage medium |
| CN112581955A (en) * | 2020-11-30 | 2021-03-30 | 广州橙行智动汽车科技有限公司 | Voice control method, server, voice control system and readable storage medium |
| CN112364147A (en) * | 2020-12-01 | 2021-02-12 | 四川长虹电器股份有限公司 | Cross-domain multi-turn dialogue method based on knowledge graph and implementation system |
| CN115081454A (en) * | 2021-03-11 | 2022-09-20 | 阿里巴巴新加坡控股有限公司 | Dialogue understanding method and device, electronic equipment and storage medium |
| CN113470625A (en) * | 2021-06-30 | 2021-10-01 | 西安力传智能技术有限公司 | Voice conversation processing method, device, equipment and storage medium |
| CN115204188A (en) * | 2022-05-27 | 2022-10-18 | 天翼电子商务有限公司 | A method of time text extraction and calculation in multi-round interaction of intelligent outbound calls |
| CN115186147B (en) * | 2022-05-31 | 2023-08-25 | 华院计算技术(上海)股份有限公司 | Dialog content generation method and device, storage medium, and terminal |
| CN115186147A (en) * | 2022-05-31 | 2022-10-14 | 华院计算技术(上海)股份有限公司 | Method and device for generating conversation content, storage medium and terminal |
| CN115394293A (en) * | 2022-08-08 | 2022-11-25 | 湖北星纪时代科技有限公司 | Dialogue system and method for implementing dialogues |
| CN116501592B (en) * | 2023-06-19 | 2023-09-19 | 阿里巴巴（中国）有限公司 | Man-machine interaction data processing method and server |
| CN116501592A (en) * | 2023-06-19 | 2023-07-28 | 阿里巴巴（中国）有限公司 | Man-machine interaction data processing method and server |
| CN117496973A (en) * | 2024-01-02 | 2024-02-02 | 四川蜀天信息技术有限公司 | Method, device, equipment and medium for improving man-machine conversation interaction experience |
| CN117496973B (en) * | 2024-01-02 | 2024-03-19 | 四川蜀天信息技术有限公司 | Method, device, equipment and medium for improving man-machine conversation interaction experience |
| CN118627631A (en) * | 2024-08-14 | 2024-09-10 | 中电数据产业集团有限公司 | Multi-round dialogue method, device, equipment and storage medium based on DST |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109616108B (en) | 2022-05-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109616108B (en) | Multi-turn dialogue interaction processing method and device, electronic equipment and storage medium | |
| CN107943998B (en) | Man-machine conversation control system and method based on knowledge graph | |
| US11676575B2 (en) | On-device learning in a hybrid speech processing system | |
| US11600291B1 (en) | Device selection from audio data | |
| US10878808B1 (en) | Speech processing dialog management | |
| US20200301954A1 (en) | Reply information obtaining method and apparatus | |
| JP7066349B2 (en) | Translation method, translation equipment and computer program | |
| US20190279618A1 (en) | System and method for language model personalization | |
| JP2020537777A (en) | Methods and devices for identifying the user's intent of speech | |
| US11250857B1 (en) | Polling with a natural language interface | |
| US11200885B1 (en) | Goal-oriented dialog system | |
| CN109243468B (en) | Voice recognition method and device, electronic equipment and storage medium | |
| CN111081220A (en) | Vehicle-mounted voice interaction method, full-duplex dialogue system, server and storage medium | |
| US11605376B1 (en) | Processing orchestration for systems including machine-learned components | |
| JP2016122183A (en) | Disambiguating heteronyms in speech synthesis | |
| CN105512228A (en) | Bidirectional question-answer data processing method and system based on intelligent robot | |
| CN115952272A (en) | Method, device and equipment for generating dialogue information and readable storage medium | |
| US20170018268A1 (en) | Systems and methods for updating a language model based on user input | |
| US11646035B1 (en) | Dialog management system | |
| US20210210088A1 (en) | Speech interaction method and apparatus, device and storage medium | |
| CN110175242B (en) | Human-computer interaction association method, device and medium based on knowledge graph | |
| CN113807080A (en) | Text correction method, device and storage medium | |
| US10929601B1 (en) | Question answering for a multi-modal system | |
| CN111625629A (en) | Task-based conversational robot response method, device, robot and storage medium | |
| CN111078848A (en) | Input prompting method and device for conversation robot |
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 | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20211119 Address after: 210038 8th floor, building D11, Hongfeng science and Technology Park, Nanjing Economic and Technological Development Zone, Jiangsu Province Applicant after: New Technology Co.,Ltd. Applicant after: Volkswagen (China) Investment Co., Ltd Address before: Room 501, floor 5, No. 67, North Fourth Ring West Road, Fengtai District, Beijing 100080 Applicant before: Beijing Yufanzhi Information Technology Co.,Ltd. |
|
| TA01 | Transfer of patent application right | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |