CN114266240A - Multi-intention identification method and device based on robot - Google Patents
Multi-intention identification method and device based on robot Download PDFInfo
- Publication number
- CN114266240A CN114266240A CN202111651001.9A CN202111651001A CN114266240A CN 114266240 A CN114266240 A CN 114266240A CN 202111651001 A CN202111651001 A CN 202111651001A CN 114266240 A CN114266240 A CN 114266240A
- Authority
- CN
- China
- Prior art keywords
- text
- intention
- intent
- recognized
- robot
- 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
Images
Landscapes
- Machine Translation (AREA)
Abstract
The application provides a robot-based multi-intention identification method and a robot-based multi-intention identification device, wherein the robot-based multi-intention identification method comprises the following steps: receiving a text to be recognized; carrying out whole sentence intention type prediction on the text to be recognized to obtain a plurality of prediction probabilities; judging whether the text to be identified is a multi-purpose text or not according to the plurality of prediction probabilities; when the text to be recognized is the multi-intention text, performing clause recognition on the text to be recognized based on a sentence structure to obtain a plurality of clauses; and identifying the intentions of the plurality of clauses to obtain a plurality of clause intentions. As can be seen, by implementing this embodiment, the robot can improve the processing capability for complex multi-intent input and improve the response effect for complex multi-intent.
Description
Technical Field
The application relates to the field of robots, in particular to a multi-purpose recognition method and device based on a robot.
Background
At present, most of intelligent service robots based on natural language dialogue/question-and-answer technology in the industry only recognize and respond to single intentions, but pay less attention to recognition and response of multiple intentions. Therefore, the existing intelligent service robot has poor processing capability and response effect on complex multi-purpose input.
Disclosure of Invention
An object of the embodiments of the present application is to provide a method and an apparatus for identifying multiple intents based on a robot, which can improve the processing capability of the robot for inputting complex multiple intents and improve the response effect for the complex multiple intents.
The embodiment of the application provides a robot-based multi-intention identification method in a first aspect, which comprises the following steps:
receiving a text to be recognized;
carrying out whole sentence intention type prediction on the text to be recognized to obtain a plurality of prediction probabilities;
judging whether the text to be identified is a multi-purpose text or not according to the plurality of prediction probabilities;
when the text to be recognized is the multi-intention text, performing clause recognition on the text to be recognized based on a sentence structure to obtain a plurality of clauses;
and identifying the intentions of the plurality of clauses to obtain a plurality of clause intentions.
In the implementation process, the method can judge whether the input text is the multi-intention text, and clause recognition and intention recognition are carried out when the input text is the multi-intention text, so that multi-intention recognition processing is realized, the processing capability of the robot on multi-intention input is improved, and the robot can be more intelligent.
Further, after the step of performing intent recognition on the plurality of clauses to obtain a plurality of clause intentions, the method further includes:
carrying out consent graph merging processing according to the plurality of clause intentions to obtain a multi-intention result; the multi-intent result includes a plurality of different intents;
and performing multi-intention response according to the multi-intention result.
In the implementation process, the method can also perform corresponding response according to the multi-intention analysis result and the multi-intention type, so that the response effect of the robot can be improved, and the robot can more intelligently complete corresponding tasks.
Further, the step of judging whether the text to be recognized is a multi-purpose text according to the plurality of prediction probabilities includes:
calculating according to the plurality of prediction probabilities to obtain an entropy value, a maximum confidence value and a confidence difference value;
when the entropy value is larger than a preset entropy, the maximum confidence value is smaller than a preset confidence value, and the confidence difference value is larger than a preset difference value, judging whether punctuation exists in the text to be recognized or not;
and when the punctuation exists in the text to be recognized, confirming that the text to be recognized is the multi-intention text.
Further, the formula for calculating the entropy value is
The maximum confidence value is calculated according to the formula
The confidence coefficient difference value is calculated by the formula
Wherein, Pθ(yi| x) is the predicted prediction probability;
i is used to represent the predicted ith intention;
n is used to represent the predicted total number of intents;
Further, the step of making a multi-intent response according to the multi-intent result comprises:
extracting an intent type of each intent in the multi-intent result;
and performing multi-intention response according to the plurality of different intentions and the intention types.
A second aspect of embodiments of the present application provides a robot-based multi-intent recognition apparatus, including:
the receiving unit is used for receiving the text to be recognized;
the prediction unit is used for carrying out whole sentence intention type prediction on the text to be recognized to obtain a plurality of prediction probabilities;
the judging unit is used for judging whether the text to be identified is a multi-purpose text or not according to the plurality of prediction probabilities;
the recognition unit is used for carrying out clause recognition on the text to be recognized based on a sentence structure to obtain a plurality of clauses when the text to be recognized is the multi-intention text; and identifying the intentions of the plurality of clauses to obtain a plurality of clause intentions.
In the implementation process, the device can automatically realize the recognition of the multi-intention text and the splitting of the multi-clause through the combination among the units, so that the robot carrying the device can realize the effect of multi-intention recognition, and the robot can effectively respond to the multi-intention contents correspondingly.
Further, the multiple intention recognition apparatus further includes:
the processing unit is used for carrying out consent graph merging processing according to the plurality of clause intentions to obtain a multi-intention result; the multi-intent result includes a plurality of different intents;
and the response unit is used for carrying out multi-intention response according to the multi-intention result.
Further, the judging unit includes:
the calculating subunit is used for calculating according to the plurality of prediction probabilities to obtain an entropy value, a maximum confidence value and a confidence difference value;
the judging subunit is used for judging whether punctuations exist in the text to be recognized or not when the entropy value is larger than a preset entropy, the maximum confidence value is smaller than a preset confidence value and the confidence difference value is larger than a preset difference value;
and the determining subunit is used for determining that the text to be recognized is the multi-intention text when the punctuation exists in the text to be recognized.
A third aspect of embodiments of the present application provides an electronic device, including a memory and a processor, where the memory is used to store a computer program, and the processor runs the computer program to cause the electronic device to execute the robot-based multiple intention identification method according to any one of the first aspect of embodiments of the present application.
A fourth aspect of the embodiments of the present application provides a computer-readable storage medium, which stores computer program instructions, and when the computer program instructions are read and executed by a processor, the computer program instructions perform the robot-based multi-intent recognition method according to any one of the first aspect of the embodiments of the present application.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments of the present application will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and that those skilled in the art can also obtain other related drawings based on the drawings without inventive efforts.
Fig. 1 is a schematic flowchart of a robot-based multi-intent recognition method according to an embodiment of the present disclosure;
FIG. 2 is a schematic flowchart of another robot-based multi-intent recognition method according to an embodiment of the present disclosure;
fig. 3 is a schematic structural diagram of a robot-based multi-intent recognition apparatus according to an embodiment of the present disclosure;
FIG. 4 is a diagram illustrating an example of syntactic analysis provided in an embodiment of the present application;
FIG. 5 is a diagram of sentence connectivity provided by an embodiment of the present application;
FIG. 6 is a plurality of sub-graphs of a communication provided in an embodiment of the present application; the upper graph is a segmentation result graph, the lower left graph is a segmented connected graph, and the lower right graph is another segmented connected graph.
Detailed Description
The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application.
It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures. Meanwhile, in the description of the present application, the terms "first", "second", and the like are used only for distinguishing the description, and are not to be construed as indicating or implying relative importance.
Referring to fig. 1, fig. 1 is a flowchart illustrating a robot-based multi-intent recognition method according to an embodiment of the present disclosure. The robot-based multi-intention identification method comprises the following steps:
s101, receiving a text to be recognized.
S102, performing whole sentence intention type prediction on the text to be recognized to obtain a plurality of prediction probabilities.
S103, judging whether the text to be identified is a multi-purpose text or not according to the plurality of prediction probabilities, and if so, executing steps S104-S105; if not, the flow is ended.
And S104, performing clause recognition on the text to be recognized based on the sentence structure to obtain a plurality of clauses.
And S105, identifying the intentions of the clauses to obtain a plurality of clause intentions.
It can be seen that, by implementing the multi-intent recognition method based on the robot described in this embodiment, it can be determined whether the input text is a multi-intent text, and clause recognition and intent recognition are performed when the input text is the multi-intent text, so that multi-intent recognition processing is implemented, and further, the robot can increase understanding and response capabilities under the condition of multi-intent input, and effectively enhance interactive experience and effect.
Referring to fig. 2, fig. 2 is a schematic flowchart illustrating a robot-based multi-intent recognition method according to an embodiment of the present application. The robot-based multi-intention identification method comprises the following steps:
s201, receiving a text to be recognized.
In this embodiment, the method may receive the text to be recognized in various ways.
In this embodiment, the method may listen to the voice information input by the user through the voice listening device of the robot, and then perform text conversion on the voice information to obtain the text to be recognized.
In this embodiment, the method may also directly input the text to be recognized.
S202, performing whole sentence intention type prediction on the text to be recognized to obtain a plurality of prediction probabilities.
In this embodiment, the method may perform intent recognition on the text to be recognized. Wherein the method may treat the process of intent recognition as a process of text classification. Specifically, the text to be recognized is input into a Bert classification model, so that the Bert classification model predicts the complete text intention category of the text to be recognized, and a plurality of prediction probabilities are obtained.
In this embodiment, the prediction probability is used to represent a single intention probability.
In this embodiment, whether the text to be recognized is the multi-intent text is determined according to the plurality of prediction probabilities, and the specific process executes steps S203 to S204.
S203, calculating according to the plurality of prediction probabilities to obtain an entropy value, a maximum confidence value and a confidence difference value.
As an alternative embodiment, the formula for calculating the entropy value is
The maximum confidence value is calculated by the formula
The confidence difference is calculated by
Wherein, Pθ(yi| x) is the predicted prediction probability;
i is used to represent the predicted ith intention;
n is used to represent the predicted total number of intents;
S204, when the entropy value is larger than the preset entropy, the maximum confidence value is smaller than the preset confidence value and the confidence difference value is larger than the preset difference value, judging whether punctuation exists in the text to be recognized, if so, executing the steps S205-S209; if not, the flow is ended.
In this embodiment, the method may perform the calculation of the entropy value, the maximum confidence value, and the confidence difference value according to the intention classification result on the whole sentence. The entropy value H, the maximum confidence value C, the confidence difference value M and whether the punctuation exists are used for judging whether the text is really the multi-intention text.
In the embodiment, if H > a, C < b, M > C (where a is a preset entropy, b is a preset confidence value, and C is a preset difference value, where a, b, and C are all floating point type thresholds set empirically) are satisfied at the same time, and the text to be recognized is determined to be the multi-intent text when punctuation marks are included in the sentence.
And S205, performing clause recognition on the text to be recognized based on the sentence structure to obtain a plurality of clauses.
In this embodiment, the method may analyze the text to be recognized by using dependency syntax.
As an alternative implementation, step S205 may include:
identifying a text to be identified based on a sentence structure to obtain text words and word dependence relations;
constructing a connected whole graph by taking the text words as nodes and the word dependency relationship as edges;
segmenting the whole connected graph according to the word dependency relationship to obtain a plurality of connected subgraphs;
and converting the plurality of connected subgraphs into a plurality of clauses.
For example, the text to be recognized is: turning on the television and adjusting the fan to the third gear.
Referring to fig. 4, fig. 4 shows an exemplary diagram of a syntactic analysis. In the figure, the method can identify an "IC" structure in the sentence "turn on the television, turn the fan to the third file", so that each word is divided, label the corresponding part of speech for each word, and then build an "IC" dependency relationship based on the verb relationship, so that the sentence can be divided into a plurality of text words "turn on", "television", "/" turn "," third ", and" file ", and obtain the word dependency relationship among the text words. Wherein v represents a verb, n represents a noun, w represents a punctuation, p represents a preposition, m is a numerator, and q is a quantifier. Meanwhile, the dependency relationship between text words includes: VOB is the relation between verb and its object, PUN is the relation of birelation, POB is the relation between preposition and object, ADV is verb modification relation, CMP move word combination relation, NUM is the relation between number word and quantifier.
Specifically, the method may represent each text word in the sentence as a node, and the dependency relationship is used as an edge, so that the whole sentence may be represented as a graph structure as follows, which is a connected graph at this time. This communication diagram is shown in fig. 5. Wherein "1" corresponds to "on", "2" corresponds to "tv", "3" corresponds to ",", "4" corresponds to "handle", "5" corresponds to "fan", "6" corresponds to "tune", "7" corresponds to "," 8 "corresponds to" three ", and" 9 "corresponds to" gear ".
At this time, if the "IC" dependency relationship is removed, the sentence is divided into several connected subgraphs, which are shown in fig. 6. The upper diagram in fig. 6 is a division result diagram, the lower left diagram in fig. 6 is a connection diagram corresponding to "turn on the television", and the lower right diagram in fig. 6 is a connection diagram corresponding to "shift the fan to third gear".
In this embodiment, the method can eliminate IC dependencies in the syntax structure, so that a sentence is split into several clauses as soon as the sentence is split into several dependency Token clusters. Such as two clusters in the example (on, tv), (handle, fan, tune, to, three, gear). Therefore, the sentence is divided into two clauses of "turn on the television", "turn the fan to the third gear". The formalized expression of the method is specifically a problem of solving connected subgraphs, wherein each graph is a clause.
And S206, identifying the intentions of the clauses to obtain a plurality of clause intentions.
In this embodiment, this step is applied to perform intent recognition on each of the analyzed clauses, so as to obtain a plurality of clause intentions. The above example corresponds to the clause "turn on tv" being recognized as the intention "tv-on"; the corresponding clause "adjust fan to third gear" is identified as the intent "fan-shift".
S207, carrying out consent graph merging processing according to the clause intentions to obtain a multi-intention result; the multi-intent result includes a plurality of different intents.
In this embodiment, if a plurality of clauses indicate the same intention, the intentions of the plurality of clauses need to be merged.
In this embodiment, after obtaining the multi-intent result, a multi-intent response is performed according to the multi-intent result, and the specific process is executed in S208-S209.
And S208, extracting the intention type of each intention in the multi-intention result.
In this embodiment, each intent in the multi-intent result has a corresponding intent type, which includes a direct answer type (A: FAQ) and a multiple round Dialog type (D: Dialog).
And S209, performing multi-intention response according to a plurality of different intentions and intention types.
In this embodiment, because each of the plurality of different intents has a corresponding intent type, the different intent types will constitute a multi-intent type combination as shown in the following table.
In this embodiment, for different combinations of multiple intention types, the method presets multiple response strategies corresponding to the combinations one by one, so that the method can directly respond to the multiple intention texts. Wherein, the specific content is shown in the following table:
in the embodiment of the present application, the execution subject of the method may be a computing device such as a computer and a server, and is not limited in this embodiment.
In this embodiment, an execution subject of the method may also be an intelligent device such as a smart phone and a tablet computer, which is not limited in this embodiment.
It can be seen that, by implementing the multi-intent recognition method based on the robot described in this embodiment, it can be determined whether the input text is a multi-intent text, and clause recognition and intent recognition are performed when the input text is the multi-intent text, so that multi-intent recognition processing is implemented, and further, the robot can increase understanding and response capabilities under the condition of multi-intent input, and effectively enhance interactive experience and effect.
Referring to fig. 3, fig. 3 is a schematic structural diagram of a robot-based multi-intent recognition apparatus according to an embodiment of the present application. As shown in fig. 3, the robot-based multi-intent recognition apparatus includes:
a receiving unit 310, configured to receive a text to be recognized;
the prediction unit 320 is configured to perform whole sentence intention category prediction on a text to be recognized to obtain multiple prediction probabilities;
a judging unit 330, configured to judge whether the text to be recognized is a multi-intent text according to a plurality of prediction probabilities;
the recognition unit 340 is configured to, when the text to be recognized is a multi-purpose text, perform clause recognition on the text to be recognized based on a sentence structure to obtain a plurality of clauses; and identifying the intentions of the plurality of clauses to obtain a plurality of clause intentions.
As an optional implementation, the multi-intent recognition apparatus further includes:
a processing unit 350, configured to perform consent graph merging processing according to the multiple clause intents to obtain a multiple intention result; the multi-intent result includes a plurality of different intents;
and a response unit 360 for performing multi-intent response according to the multi-intent result.
As an optional implementation, the determining unit 330 includes:
a calculating subunit 331, configured to perform calculation according to the multiple prediction probabilities to obtain an entropy value, a maximum confidence value, and a confidence difference value;
a judging subunit 332, configured to, when the entropy value is greater than the preset entropy, the maximum confidence value is smaller than the preset confidence value, and the confidence difference value is greater than the preset difference value, judge whether a punctuation exists in the text to be recognized;
the determining subunit 333 is configured to determine that the text to be recognized is the multi-intent text when the punctuation exists in the text to be recognized.
As an alternative embodiment, the formula for calculating the entropy value is
The maximum confidence value is calculated by the formula
The confidence difference is calculated by
Wherein, Pθ(yi| x) is the predicted prediction probability;
i is used to represent the predicted ith intention;
n is used to represent the predicted total number of intents;
As an alternative embodiment, the identification unit 340 includes:
the recognition subunit 341 is configured to recognize the text to be recognized based on the sentence structure, so as to obtain a text word and a word dependency relationship;
the construction subunit 342 is configured to construct a connected whole graph by using the text word as a node and the word dependency relationship as an edge;
a dividing subunit 343, configured to divide the connected whole graph according to the word dependency relationship, to obtain a plurality of connected subgraphs;
a converting subunit 344, configured to convert the plurality of connected subgraphs into a plurality of clauses.
As an alternative embodiment, the response unit 360 includes:
an extracting subunit 361, for extracting an intention type of each intention in the multiple intention result;
a response subunit 362 for performing a multi-intent response according to a plurality of different intents and intent types.
In the embodiment of the present application, for the explanation of the multi-purpose recognition device based on a robot, reference may be made to the description in the method embodiment, and further description is not repeated in this embodiment.
It can be seen that, by implementing the multi-purpose recognition device based on the robot described in this embodiment, recognition of multi-purpose texts and splitting of multi-clauses can be automatically achieved through combination among multiple units, so that the robot carrying the device can achieve the effect of multi-purpose recognition, and further, the robot can perform corresponding effective response on multi-purpose contents.
The embodiment of the application provides an electronic device, which comprises a memory and a processor, wherein the memory is used for storing a computer program, and the processor runs the computer program to enable the electronic device to execute the robot-based multi-intention identification method in the embodiment of the application.
Embodiments of the present application provide a computer-readable storage medium, which stores computer program instructions, and when the computer program instructions are read and executed by a processor, the computer program instructions execute the robot-based multi-intent recognition method in the embodiments of the present application.
In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method can be implemented in other ways. The apparatus embodiments described above are merely illustrative, and for example, the flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
In addition, functional modules in the embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.
The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application or portions thereof that substantially contribute to the prior art may be embodied in the form of a software product stored in a storage medium and including instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
The above description is only an example of the present application and is not intended to limit the scope of the present application, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present application shall be included in the protection scope of the present application. It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures.
The above description is only for the specific embodiments of the present application, but the scope of the present application is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present application, and shall be covered by the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, 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 a process, method, article, or apparatus that comprises the element.
Claims (10)
1. A robot-based multi-intent recognition method, comprising:
receiving a text to be recognized;
carrying out whole sentence intention type prediction on the text to be recognized to obtain a plurality of prediction probabilities;
judging whether the text to be identified is a multi-purpose text or not according to the plurality of prediction probabilities;
when the text to be recognized is the multi-intention text, performing clause recognition on the text to be recognized based on a sentence structure to obtain a plurality of clauses;
and identifying the intentions of the plurality of clauses to obtain a plurality of clause intentions.
2. The robot-based multiple intent recognition method of claim 1, wherein after the step of performing intent recognition on the plurality of clauses to obtain a plurality of clause intents, the method further comprises:
carrying out consent graph merging processing according to the plurality of clause intentions to obtain a multi-intention result; the multi-intent result includes a plurality of different intents;
and performing multi-intention response according to the multi-intention result.
3. The robot-based multi-intent recognition method of claim 1, wherein the step of determining whether the text to be recognized is a multi-intent text according to the plurality of prediction probabilities comprises:
calculating according to the plurality of prediction probabilities to obtain an entropy value, a maximum confidence value and a confidence difference value;
when the entropy value is larger than a preset entropy, the maximum confidence value is smaller than a preset confidence value, and the confidence difference value is larger than a preset difference value, judging whether punctuation exists in the text to be recognized or not;
and when the punctuation exists in the text to be recognized, confirming that the text to be recognized is the multi-intention text.
4. A robot-based multi-intent recognition method according to claim 3, wherein the formula of calculating the entropy is
The maximum confidence value is calculated according to the formula
The confidence coefficient difference value is calculated by the formula
Wherein, Pθ(yi| x) is the predicted prediction probability;
i is used to represent the predicted ith intention;
n is used to represent the predicted total number of intents;
5. The robot-based multi-intent recognition method according to claim 1, wherein the step of performing clause recognition on the text to be recognized based on sentence structure to obtain a plurality of clauses comprises:
identifying a text to be identified based on a sentence structure to obtain text words and word dependence relations;
constructing a connected whole graph by taking the text words as nodes and the word dependency relationship as edges;
segmenting the connected whole graph according to the word dependency relationship to obtain a plurality of connected subgraphs;
converting the plurality of connected subgraphs into a plurality of clauses.
6. The robot-based multi-intent recognition method of claim 2, wherein the step of responding with multi-intent based on the multi-intent results comprises:
extracting an intent type of each intent in the multi-intent result;
and performing multi-intention response according to the plurality of different intentions and the intention types.
7. A robot-based multi-intent recognition apparatus, comprising:
the receiving unit is used for receiving the text to be recognized;
the prediction unit is used for carrying out whole sentence intention type prediction on the text to be recognized to obtain a plurality of prediction probabilities;
the judging unit is used for judging whether the text to be identified is a multi-purpose text or not according to the plurality of prediction probabilities;
the recognition unit is used for carrying out clause recognition on the text to be recognized based on a sentence structure to obtain a plurality of clauses when the text to be recognized is the multi-intention text; and identifying the intentions of the plurality of clauses to obtain a plurality of clause intentions.
8. The robot-based multi-intent recognition device of claim 7, further comprising:
the processing unit is used for carrying out consent graph merging processing according to the plurality of clause intentions to obtain a multi-intention result; the multi-intent result includes a plurality of different intents;
and the response unit is used for carrying out multi-intention response according to the multi-intention result.
9. An electronic device, comprising a memory for storing a computer program and a processor that runs the computer program to cause the electronic device to perform the robot-based multi-intent recognition method of any of claims 1-6.
10. A readable storage medium having stored therein computer program instructions which, when read and executed by a processor, perform the robot-based multi-intent recognition method of any of claims 1-6.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111651001.9A CN114266240A (en) | 2021-12-30 | 2021-12-30 | Multi-intention identification method and device based on robot |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111651001.9A CN114266240A (en) | 2021-12-30 | 2021-12-30 | Multi-intention identification method and device based on robot |
Publications (1)
Publication Number | Publication Date |
---|---|
CN114266240A true CN114266240A (en) | 2022-04-01 |
Family
ID=80831758
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111651001.9A Pending CN114266240A (en) | 2021-12-30 | 2021-12-30 | Multi-intention identification method and device based on robot |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114266240A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11930312B1 (en) | 2022-10-28 | 2024-03-12 | Shenzhen Shokz Co., Ltd. | Open earphones |
-
2021
- 2021-12-30 CN CN202111651001.9A patent/CN114266240A/en active Pending
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11930312B1 (en) | 2022-10-28 | 2024-03-12 | Shenzhen Shokz Co., Ltd. | Open earphones |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109635273B (en) | Text keyword extraction method, device, equipment and storage medium | |
CN108959246B (en) | Answer selection method and device based on improved attention mechanism and electronic equipment | |
US9633002B1 (en) | Systems and methods for coreference resolution using selective feature activation | |
CN116775847B (en) | Question answering method and system based on knowledge graph and large language model | |
EP3819785A1 (en) | Feature word determining method, apparatus, and server | |
JP6335898B2 (en) | Information classification based on product recognition | |
US8983826B2 (en) | Method and system for extracting shadow entities from emails | |
JP2020030408A (en) | Method, apparatus, device and medium for identifying key phrase in audio | |
CN107680588B (en) | Intelligent voice navigation method, device and storage medium | |
CN106997342B (en) | Intention identification method and device based on multi-round interaction | |
WO2017198031A1 (en) | Semantic parsing method and apparatus | |
CN116304748B (en) | Text similarity calculation method, system, equipment and medium | |
CN113821605A (en) | Event extraction method | |
CN107545505B (en) | Method and system for identifying insurance financing product information | |
CN112669842A (en) | Man-machine conversation control method, device, computer equipment and storage medium | |
CN108133014B (en) | Triple generation method and device based on syntactic analysis and clustering and user terminal | |
US10325024B2 (en) | Contextual analogy response | |
US20210082422A1 (en) | System and method to interpret natural language requests and handle natural language responses in conversation | |
CN114266240A (en) | Multi-intention identification method and device based on robot | |
US11625630B2 (en) | Identifying intent in dialog data through variant assessment | |
CN112487181B (en) | Keyword determination method and related equipment | |
CN115858776B (en) | Variant text classification recognition method, system, storage medium and electronic equipment | |
CN114818736B (en) | Text processing method, chain finger method and device for short text and storage medium | |
CN114647739B (en) | Entity chain finger method, device, electronic equipment and storage medium | |
CN114416923A (en) | News entity linking method and system based on rich text characteristics |
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 |