CN117010378A - Semantic conversion method and device, storage medium and electronic device - Google Patents
Semantic conversion method and device, storage medium and electronic device Download PDFInfo
- Publication number
- CN117010378A CN117010378A CN202210467403.1A CN202210467403A CN117010378A CN 117010378 A CN117010378 A CN 117010378A CN 202210467403 A CN202210467403 A CN 202210467403A CN 117010378 A CN117010378 A CN 117010378A
- Authority
- CN
- China
- Prior art keywords
- operator
- operators
- control instruction
- semantic
- type
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 64
- 238000006243 chemical reaction Methods 0.000 title claims abstract description 42
- 230000009471 action Effects 0.000 claims description 36
- 238000000605 extraction Methods 0.000 claims description 13
- 238000004590 computer program Methods 0.000 claims description 9
- 230000015654 memory Effects 0.000 claims description 7
- 238000010586 diagram Methods 0.000 description 10
- 230000001276 controlling effect Effects 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 5
- 238000004378 air conditioning Methods 0.000 description 4
- 230000011218 segmentation Effects 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000013135 deep learning Methods 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 238000012706 support-vector machine Methods 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000007477 logistic regression Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000001105 regulatory effect Effects 0.000 description 2
- 230000002123 temporal effect Effects 0.000 description 2
- 238000005406 washing Methods 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000007635 classification algorithm Methods 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000010411 cooking Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 239000000779 smoke Substances 0.000 description 1
- 238000010408 sweeping Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Machine Translation (AREA)
- Toys (AREA)
- User Interface Of Digital Computer (AREA)
Abstract
The invention discloses a semantic conversion method and device, a storage medium and an electronic device, and relates to the technical field of smart families, wherein the semantic conversion method comprises the following steps: under the condition that a first control instruction sent by a first object is received, carrying out semantic recognition on a plurality of operators corresponding to the first control instruction through a defined operator model to obtain semantics of the operators, wherein the first control instruction is used for controlling target equipment, and the defined operator model comprises definitions of the operators; the semantics of the plurality of operators are converted into a second control instruction that the target device is allowed to execute. By adopting the technical scheme, in the prior art, when a control instruction of a user to equipment is converted into an execution instruction of the equipment, the problem of overlarge error of semantic recognition of the control instruction of the user is solved.
Description
Technical Field
The invention relates to the technical field of smart families, in particular to a semantic conversion method and device, a storage medium and an electronic device.
Background
At present, more intelligent devices begin to use device semantic understanding technology. Device semantic understanding technology is a technology based on Natural Language Understanding (NLU) that converts user control instructions of a device into device-understandable semantics. The existing natural language understanding technology mainly adopts Pipeline (Chinese name is Pipeline) frame type model or method for analysis. The Pipeline frame model mainly comprises two steps of intention recognition and slot filling, wherein the intention recognition is a classification problem, and is generally classified by adopting a machine learning and deep learning method, and also classified by adopting a similarity calculation method. Machine learning mostly adopts a Support Vector Machine (SVM) (generally called Support Vector Machine), logistic regression (generally called Logistic Regression) and other classification methods, and deep learning mostly adopts: text classification algorithm TextCNN (full name: convolutional Neural Network), long-Short Term Memory network LSMT (full name Long Short-Term Memory), etc., and slot filling is generally identified by mechanical dictionary matching, fuzzy matching, entity identification, etc.
The intent classification in the Pipeline frame-based nlu system depends on manual data marking, and the quality of the manual data marking directly influences the recognition effect of the model. However, errors exist in the manual data annotation, which results in uncontrollable process of converting the user instruction into the semantic meaning understood by the device; on the other hand, the intention of the user is identified by the machine learning or the deep learning, and the unknown corpus is classified into a known class, so that the classification result of the unknown corpus has huge error.
Aiming at the problem that in the prior art, when a control instruction of a user to equipment is converted into an execution instruction of the equipment, the error of semantic recognition of the control instruction of the user is overlarge, no effective solution is proposed yet.
Disclosure of Invention
The embodiment of the invention provides a semantic conversion method and device, a storage medium and an electronic device, which at least solve the problem that in the prior art, when a control instruction of a user to equipment is converted into an execution instruction of the equipment, the error of semantic recognition of the control instruction of the user is overlarge.
According to an embodiment of the present invention, there is provided a semantic conversion method including: under the condition that a first control instruction sent by a first object is received, carrying out semantic recognition on a plurality of operators corresponding to the first control instruction through a defined operator model to obtain semantics of the operators, wherein the first control instruction is used for controlling target equipment, and the defined operator model comprises definitions of the operators; the semantics of the plurality of operators are converted into a second control instruction which the target device is allowed to execute.
In an exemplary embodiment, in a case of receiving a first control instruction sent by a first object, before performing semantic recognition on a plurality of operators corresponding to the first control instruction through a defined operator model, the method further includes: determining a definition of each of all operators in the operator model, wherein the definition of each operator comprises: the operator type and operator attribute of each operator.
In one exemplary embodiment, determining the definition of each of all operators in the operator model includes: receiving a definition operation of a second object, and determining an operator type and an operator attribute of each operator in all operators in the operator model according to the definition operation, wherein the operator type comprises: action word type, entity word type, parameter word type; the operator attributes include: the meaning of each operator, whether each operator refers to other operators, whether each operator participates in operator combination, and whether each operator participates in operator output.
In an exemplary embodiment, performing semantic recognition on a plurality of operators corresponding to the first control instruction through the defined operator model includes: combining the operators according to the definition of each operator in the plurality of operators corresponding to the first control instruction through the defined operator model to obtain an operator combination of the operators; combining the operators as semantics of the plurality of operators.
In an exemplary embodiment, combining the plurality of operators according to the definition of each of the plurality of operators corresponding to the first control instruction through the defined operator model includes: determining whether an operator type of a first operator of the plurality of operators is an action word type and whether the first operator participates in operator combination, determining whether an operator type of a second operator of the plurality of operators is a device word type and whether the second operator participates in operator combination, and determining whether an operator of a parameter word type exists in the plurality of operators; and combining the first operator and the second operator through the defined operator model under the condition that the operator type of the first operator is an action word type and the first operator participates in operator combination, the operator type of the second operator is a device word type and the second operator participates in operator combination, and no operator of the parameter word type exists in the operators.
In an exemplary embodiment, combining the plurality of operators according to the definition of each of the plurality of operators corresponding to the first control instruction through the defined operator model includes: determining whether an operator type of a first operator of the plurality of operators is an action word type and whether the first operator participates in an operator combination, and determining whether an operator type of a second operator of the plurality of operators is a device word type and whether the second operator participates in an operator combination, and determining whether an operator type of a third operator of the plurality of operators is a parameter word type and whether the third operator participates in an operator combination; and combining the first operator and the third operator through the defined operator model under the condition that the first operator is of an action word type and the first operator participates in operator combination, the second operator is of a device word type and the second operator participates in operator combination, and the third operator is of a parameter word type and the third operator participates in operator combination.
In an exemplary embodiment, before performing semantic recognition on the plurality of operators corresponding to the first control instruction through the defined operator model, the method further includes: operator extraction is carried out on the first control instruction so as to obtain all operators corresponding to the first control instruction; determining operators with ambiguity in all operators; and taking other operators except the ambiguous operator in all operators as a plurality of operators of the first control instruction.
In one exemplary embodiment, converting semantics of the plurality of operators into a second control instruction that the target device is permitted to execute, includes: identifying semantic intentions of the operators in a preset intention identification mode to obtain semantic identification results of the semantics of the operators; filling the slots in the semantic recognition result to obtain a slot filling result; and taking the slot filling result as a second control instruction which is allowed to be executed by the target equipment.
According to another embodiment of the present invention, there is also provided a semantic conversion apparatus including: the identification module is used for carrying out semantic identification on a plurality of operators corresponding to a first control instruction through a defined operator model under the condition that the first control instruction sent by a first object is received, so as to obtain the semantics of the plurality of operators, wherein the first control instruction is used for controlling target equipment, and the defined operator model comprises definitions of the plurality of operators; and the conversion module is used for converting the semantics of the operators into a second control instruction which is allowed to be executed by the target equipment.
According to a further aspect of embodiments of the present invention, there is also provided a computer readable storage medium having a computer program stored therein, wherein the computer program is arranged to perform the above-described semantic conversion method when run.
According to still another aspect of the embodiments of the present invention, there is further provided an electronic device including a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the semantic conversion method described above through the computer program.
In the embodiment of the invention, under the condition that a first control instruction for controlling target equipment sent by a first object is received, semantic recognition is carried out on a plurality of operators corresponding to the first control instruction through a defined operator model, so as to obtain the semantics of the operators, wherein the defined operator model comprises the definitions of the operators; and further converting semantics of the plurality of operators into a second control instruction that the target device is permitted to execute. Because the operators in the operator model are limited and the model in the operators can be defined by defining the operator model, the operators in the operator model are controllable, and the object of semantic recognition is the operator corresponding to the first control instruction instead of the first control instruction, so that the accuracy of semantic recognition is higher than that of direct semantic recognition of the control instruction.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description, serve to explain the principles of the invention.
In order to more clearly illustrate the embodiments of the invention or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be briefly described, and it will be obvious to a person skilled in the art that other drawings can be obtained from these drawings without inventive effort.
FIG. 1 is a schematic diagram of a hardware environment of a semantic conversion method according to an embodiment of the present invention;
FIG. 2 is a flow chart of an alternative semantic conversion method according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of an alternative operator definition method according to an embodiment of the invention;
FIG. 4 is a block diagram of an alternative semantic conversion method according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of an alternative semantic conversion method according to an embodiment of the present invention;
fig. 6 is a block diagram of an alternative semantic conversion device according to an embodiment of the present invention.
Detailed Description
In order that those skilled in the art will better understand the present invention, a technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, shall fall within the scope of the present invention.
It should be noted that the terms "first," "second," and the like in the description and the claims of the present invention and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the invention described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
According to an aspect of an embodiment of the present invention, there is provided a semantic conversion method. The semantic conversion method is widely applied to full-house intelligent digital control application scenes such as intelligent Home (Smart Home), intelligent Home equipment ecology, intelligent Home (Intelligence House) ecology and the like. Alternatively, in the present embodiment, the above-described semantic conversion method may be applied to a hardware environment constituted by the terminal device 102 and the server 104 as shown in fig. 1. As shown in fig. 1, the server 104 is connected to the terminal device 102 through a network, and may be used to provide services (such as application services and the like) for a terminal or a client installed on the terminal, a database may be set on the server or independent of the server, for providing data storage services for the server 104, and cloud computing and/or edge computing services may be configured on the server or independent of the server, for providing data computing services for the server 104.
The network may include, but is not limited to, at least one of: wired network, wireless network. The wired network may include, but is not limited to, at least one of: a wide area network, a metropolitan area network, a local area network, and the wireless network may include, but is not limited to, at least one of: WIFI (Wireless Fidelity ), bluetooth. The terminal device 102 may not be limited to a PC, a mobile phone, a tablet computer, an intelligent air conditioner, an intelligent smoke machine, an intelligent refrigerator, an intelligent oven, an intelligent cooking range, an intelligent washing machine, an intelligent water heater, an intelligent washing device, an intelligent dish washer, an intelligent projection device, an intelligent television, an intelligent clothes hanger, an intelligent curtain, an intelligent video, an intelligent socket, an intelligent sound box, an intelligent fresh air device, an intelligent kitchen and toilet device, an intelligent bathroom device, an intelligent sweeping robot, an intelligent window cleaning robot, an intelligent mopping robot, an intelligent air purifying device, an intelligent steam box, an intelligent microwave oven, an intelligent kitchen appliance, an intelligent purifier, an intelligent water dispenser, an intelligent door lock, and the like.
In this embodiment, a semantic conversion method is provided and applied to the terminal device, and fig. 2 is a flowchart of an alternative semantic conversion method according to an embodiment of the present invention, where the flowchart includes the following steps:
step S202, under the condition that a first control instruction sent by a first object is received, carrying out semantic recognition on a plurality of operators corresponding to the first control instruction through a defined operator model to obtain semantics of the operators, wherein the first control instruction is used for controlling target equipment, and the defined operator model comprises definitions of the operators;
the present invention should be described that, the plurality of operators corresponding to the first control instruction may be understood that the first control instruction may be converted into a plurality of operators, for example, in the first control instruction: and adjusting the temperature of the refrigerator by 20 degrees, wherein an operator corresponding to the first control instruction is as follows: a refrigerator; a temperature; adjusting; 20 degrees. Further, the first control instruction may be converted into a plurality of operators, because the control instruction of the device may be limited by the user, such as the control instruction of the air conditioner: open, close, open the temperature to 30 degrees, set as limited instructions such as holiday mode.
Step S204, converting the semantics of the plurality of operators into a second control instruction which is allowed to be executed by the target device.
It should be noted that, the above embodiment may be understood as converting semantics of the plurality of operators into a second control instruction that is understandable by the device and that allows execution, where the second control instruction is related to the first control instruction, and the second control instruction may be understood as a control instruction that is understandable by the device corresponding to the first control instruction.
Under the condition that a first control instruction for controlling target equipment sent by a first object is received, carrying out semantic recognition on a plurality of operators corresponding to the first control instruction through a defined operator model to obtain semantics of the operators, wherein the defined operator model comprises definitions of the operators; and further converting semantics of the plurality of operators into a second control instruction that the target device is permitted to execute. Because the operators in the operator model are limited and the model in the operators can be defined by defining the operator model, the operators in the operator model are controllable, and the object of semantic recognition is the operator corresponding to the first control instruction instead of the first control instruction, so that the accuracy of semantic recognition is higher than that of direct semantic recognition of the control instruction.
In an exemplary embodiment, in a case of receiving a first control instruction sent by a first object, before performing semantic recognition on a plurality of operators corresponding to the first control instruction through a defined operator model, determining a definition of each operator in all operators in the operator model, where the definition of each operator includes: the operator type and operator attribute of each operator.
Optionally, in this embodiment, before determining the definition of each of all operators in the operator model, the method further includes: acquiring historical control data of the target device, the historical control data comprising: and determining that the result of the semantic recognition of the control instruction by the device is matched with the first control instruction sent by the user under the condition that the execution result of the device is not adjusted and storing the corresponding relation between the control instruction and the result of the semantic recognition. The result of this semantic recognition is divided into a plurality of words by word segmentation techniques and the plurality of words are used as a plurality of operators in an operator model. It should be further noted that the duration of the above-mentioned preset period needs to be sufficiently short, and may be 1 minute, 5 minutes, or the like.
To help understand the above embodiment, for example, a control instruction sent by a user to an air conditioner within one month is obtained, where the instruction may be a voice instruction or a text instruction input on a terminal, and a result of semantic recognition of the control instruction by the device is obtained, and it is determined whether an execution result of the device is adjusted within one minute after the device finishes executing the control instruction, for example, the semantic instruction of the user is: the temperature of the air conditioner is too high, and the semantic recognition result of the air conditioner is that: the air conditioner temperature is lowered, and thus, the air conditioner lowers the temperature. At this time, if the user does not adjust the air conditioner, the semantic recognition result of the air conditioner accords with the user requirement, the semantic recognition result is matched with the control instruction, and the 'temperature of the air conditioner is reduced' by the word segmentation technology to be split into three operators: regulating down, refrigerator and temperature.
In one exemplary embodiment, a defining operation of a second object is received, and an operator type and an operator attribute of each of all operators in the operator model are determined according to the defining operation, wherein the operator type includes: action word type, entity word type, parameter word type; the operator attributes include: the meaning of each operator, whether each operator refers to other operators, whether each operator participates in operator combination, and whether each operator participates in operator output.
To assist in understanding the above embodiments, control is an action word type, a device word type, and 37 degrees is a parameter word type, for example. It should be further noted that the parameter word types include controllable parameters such as temperature, humidity, volume, etc.
Optionally, in this embodiment, the definition of each operator may further include: description of operators. The description of the operator may be a Chinese description annotation of the operator, as shown in FIG. 3, FIG. 3 is a schematic diagram of an alternative operator definition method according to an embodiment of the present invention, and in FIG. 3, the description of the operator open is; open, i.e. the chinese name of open, open can be referenced, i.e. an operator referring to open is described as present, it being noted that this reference relation is that the two operators established in the presence of the reference relation are similar, e.g. both operator start (chinese meaning: open) and operator begin (chinese meaning: open) can be referenced to each other, since the meanings of both are similar. Whether an operator is a real word is defined because in the case where an operator is a real word, the operator's output may be involved, e.g., the operator's "not involved" output is an imaginary word having no temporal meaning.
In an exemplary embodiment, before performing semantic recognition on a plurality of operators corresponding to the first control instruction through the defined operator model, performing operator extraction on the first control instruction to obtain all operators corresponding to the first control instruction; determining operators with ambiguity in all operators; and taking other operators except the ambiguous operator in all operators as a plurality of operators of the first control instruction.
Optionally, in this embodiment, a method for performing operator extraction on the first control instruction includes, but is not limited to: the operator extraction is carried out by adopting a model, wherein the model is not an operator model, but a translation model, a common entity recognition model and the like can recognize the first control instruction as the model of the operator; operator extraction is carried out through the regular expression, namely operators corresponding to the first control instruction are matched through the regular expression; and extracting operators by a dictionary matching method, namely matching operators corresponding to the first control instruction by a preset dictionary.
Optionally, in this embodiment, the method for determining an operator with ambiguity in all operators includes, but is not limited to: disambiguation rules are preset, disambiguation is performed based on preset disambiguation rules, for example, the following rules are preset: when the meaning of the word of the action word type and the meaning of the word of the parameter word type are repeated, eliminating the word of the action word type, and based on the rule, extracting all operators as follows: open, reduce, mode, operator to determine ambiguity is: the delete, therefore, only retains the two operators open and mode.
In an exemplary embodiment, combining the plurality of operators according to the definition of each operator in the plurality of operators corresponding to the first control instruction through the defined operator model to obtain an operator combination of the plurality of operators; combining the operators as semantics of the plurality of operators. To assist in understanding the above embodiments, the above operators may be, for example: the operator 'adjusting', the operator 'air conditioning', and the operator '35 DEG', the three operators can be combined, and the air conditioning temperature is adjusted to 35 deg.
It should be noted that, the first control instruction corresponds to a plurality of operators, but the operators are split, and a single operator cannot generally explain the intention of the first control instruction of the user, for example, operator "open", and cannot know which device needs to be opened by the user. The operators have a certain logic relation, and the operators can be combined through the logic relation to obtain the actual intention of the user. For example, by combining open and bridge, it is possible to obtain that the intention of the user is to open the refrigerator, further, the combination may be performed by using language custom logic between operators, for example, the combination of action word type and equipment word type, or the combination rule may be preset by the first object, or the retrospective inference may be performed, that is, by retrospecting the first control instruction, the arrangement sequence of the operators is determined, and then the combination is performed according to the arrangement sequence.
In one exemplary embodiment, determining whether an operator type of a first operator of the plurality of operators is an action word type and the first operator participates in an operator combination, and determining whether an operator type of a second operator of the plurality of operators is a device word type and the second operator participates in an operator combination, and determining whether there is an operator of a parameter word type in the plurality of operators; and combining the first operator and the second operator through the defined operator model under the condition that the operator type of the first operator is an action word type and the first operator participates in operator combination, the operator type of the second operator is a device word type and the second operator participates in operator combination, and no operator of the parameter word type exists in the operators. And combining the first operator and the third operator through the defined operator model under the condition that the first operator is of an action word type and the first operator participates in operator combination, the second operator is of a device word type and the second operator participates in operator combination, and the third operator is of a parameter word type and the third operator participates in operator combination.
It should be noted that, in the above embodiment, in the case where the first operator is of an action word type and the first operator may participate in the operator combination, the first operator may be combined with the operator of the parameter word type and the operator of the device word type, and it is also a precondition that the operator of the parameter word type and the operator of the device word type combined with the first operator may participate in the operator combination. Such as: the operator open of the action word type and the operator refrigerator of the device word type can be combined into an openbridge in case both are defined as being participatable in the operator combination. Further, the operator of the parameter word type and the operator of the equipment word type are combined with each other by the combination priority, the operator of the action word type and the operator of the parameter word type are combined with each other by the combination priority, and the operator of the action word type and the operator of the equipment word type are combined under the condition that the operator of the parameter word type is not present or the operator of the parameter word type is present but the operator of the parameter word type is not participated in the combination of the operators. For example, an operator open of action word type, an operator bridge of device word type, an operator holidy of parameter word type (mode), all of which are defined as being able to participate in the operator combination, the result of the operator combination is openMode.
In one exemplary embodiment, converting semantics of the plurality of operators into a second control instruction that the target device is permitted to execute, includes: identifying semantic intentions of the operators in a preset intention identification mode to obtain semantic identification results of the semantics of the operators; filling the slots in the semantic recognition result to obtain a slot filling result; and taking the slot filling result as a second control instruction which is allowed to be executed by the target equipment.
The intention recognition identifies the semantic device semantics of the plurality of operators to be converted into the understandable intention of the device, and if the intention recognition result is a direct intention, for example, the intention recognition result is that ice is opened, the box directly outputs the intention recognition result, and if the intention recognition result is that some word is intended, the slot filling is needed. For example, the first control instruction of the user is to open the holiday mode of the refrigerator, and the result obtained after the operator identification and the semantic identification is that: after intention recognition, the intention of the user obtained by the user terminal is as follows: the setMode is not able to obtain the direct intention of the user, and thus a slot filling is required, and the result of the slot filling is setmode=hotday device=frame.
In addition, when slot filling is performed, some slots are not extracted from the semantics of multiple operators, and default slots need to be filled in during configuration. Such as: today there is a slot corresponding to rain: time=today type=weather, wherein time is corpus extraction slot, and type is default slot set in advance.
Optionally, in this embodiment, converting semantics of the plurality of operators into a second control instruction that the target device is allowed to execute includes: step 2, constructing a multi-intention recognition model based on cluster pre-analysis, and recognizing a plurality of intentions corresponding to the semantics of a plurality of operators; constructing a BiLSTM-CRF semantic Slot filling model based on a Slot-connected association gate mechanism, and guiding the filling of the semantic slots by using the result of intention recognition; and optimizing and training a joint model formed by the BERT model, the multi-intention recognition model and the semantic slot filling model, and recognizing by utilizing the joint model completed by optimizing and training. It should be noted that, in the embodiment, the connection between the intent recognition and the semantic slot filling is fully considered, and a joint modeling method is provided, the two tasks of the intent recognition and the slot filling are combined into one task, and the accuracy of the multi-intent recognition is improved, and the accuracy of the semantic slot filling is also improved, so that the accuracy of converting the semantics of a plurality of operators into the second control instruction is improved.
In order to better understand the process of the semantic conversion method, the following describes the implementation method flow of the semantic conversion in combination with the alternative embodiment, but is not used for limiting the technical scheme of the embodiment of the invention.
In order to better understand the process of the semantic conversion method, the foregoing embodiment will be described with reference to fig. 4, where fig. 4 is a frame diagram of an alternative semantic conversion method according to an embodiment of the present invention, in fig. 4, definitions of all operators are determined according to a definition operation of a first object and stored in an operator configuration layer, when a first control instruction of the first object is acquired, operator identification is performed first, all operators corresponding to the first control instruction are identified, ambiguous operators in the operators are eliminated, and other operators except for the ambiguous operators in all operators are used as a plurality of operators corresponding to the first control instruction. And after the semantics of the operators are subjected to intention recognition, slot filling is carried out on the result of intention recognition, and the obtained slot filling result is a corresponding second control instruction and the second control instruction is output.
In this embodiment, a semantic conversion method is provided, and fig. 5 is a schematic diagram of an alternative semantic conversion method according to an embodiment of the present invention, as shown in fig. 5, specifically including the following steps:
step S502: acquiring definition operation of a developer;
step S504: defining all operators of the operator model according to definition operation pairs;
step S506: receiving a first control instruction of a user to equipment;
step S508: all operators corresponding to the first control instruction are identified through operators, ambiguous operators in the operators are eliminated, and operators except the ambiguous operators in the operators are removed and used as a plurality of operators corresponding to the first control instruction;
step S510: combining the operators through a preset algorithm to obtain semantics of the operators;
step S512: semantic recognition is carried out on the semantics of the operators, slot filling is carried out against the result of the intention recognition, and a second control instruction corresponding to the semantics of the operators is obtained;
step S514: outputting a second control instruction.
Through the steps, the problem that in the prior art, when a control instruction of a user to equipment is converted into an execution instruction of the equipment, the error of semantic recognition of the control instruction of the user is overlarge is solved. In the scheme, on one hand, the limited operator set corresponding to the control instruction of the user is flexibly defined, the accuracy of semantic understanding of the control instruction of the equipment is improved, complex steps such as data labeling and model training are not needed, and if unknown words are encountered, corresponding operators and slots can be added, so that the difficulty that the new function of the new equipment cannot update the equipment semantics rapidly is solved; on the other hand, the semantic recognition of the operator is restricted by rules, and granularity adjustment can be performed according to service conditions, so that the semantic recognition of the control instruction has stronger controllability, and the result error of the semantic recognition is smaller.
From the description of the above embodiments, it will be clear to a person skilled in the art that the method according to the above embodiments may be implemented by means of software plus the necessary general hardware platform, but of course also by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (e.g. ROM/RAM, magnetic disk, optical disk) comprising several instructions for causing a terminal device (which may be a mobile phone, a computer, a server, or a network device, etc.) to perform the method of the various embodiments of the present invention.
In this embodiment, a semantic conversion device is further provided, and the semantic conversion device is used to implement the foregoing embodiments and preferred embodiments, and is not described in detail. As used below, the term "module" may be a combination of software and/or hardware that implements a predetermined function. While the means described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware, is also possible and contemplated.
FIG. 6 is a block diagram of an alternative semantic conversion device according to an embodiment of the present invention; as shown in fig. 6, includes:
the identifying module 62 is configured to, when a first control instruction sent by a first object is received, perform semantic identification on a plurality of operators corresponding to the first control instruction through a defined operator model, to obtain semantics of the plurality of operators, where the first control instruction is used to control a target device, and the defined operator model includes definitions of the plurality of operators;
a conversion module 64, configured to convert semantics of the plurality of operators into a second control instruction that the target device is allowed to execute.
It should be noted that, the above embodiment may be understood as converting semantics of the plurality of operators into a second control instruction that is understandable by the device and that allows execution, where the second control instruction is related to the first control instruction, and the second control instruction may be understood as a control instruction that is understandable by the device corresponding to the first control instruction.
Under the condition that a first control instruction for controlling target equipment sent by a first object is received, carrying out semantic recognition on a plurality of operators corresponding to the first control instruction through a defined operator model to obtain the semantics of the operators, wherein the defined operator model comprises the definitions of the operators; and further converting semantics of the plurality of operators into a second control instruction that the target device is permitted to execute. Because the operators in the operator model are limited and the model in the operators can be defined through the definition of the operator model, the operators in the operator model are controllable, and the object of semantic recognition is the operator corresponding to the first control instruction instead of the first control instruction, so that the accuracy of semantic recognition relative to the direct control instruction is higher.
In an exemplary embodiment, the identifying module 62 is further configured to determine, when the first control instruction sent by the first object is received, a definition of each operator in the operator model before performing semantic recognition on a plurality of operators corresponding to the first control instruction through the defined operator model, where the definition of each operator includes: the operator type and operator attribute of each operator.
Optionally, in this embodiment, the apparatus further includes: the acquisition module is used for acquiring historical control data of the target equipment, wherein the historical control data comprises: and determining that the result of the semantic recognition of the control instruction by the device is matched with the first control instruction sent by the user under the condition that the execution result of the device is not adjusted and storing the corresponding relation between the control instruction and the result of the semantic recognition. The result of this semantic recognition is divided into a plurality of words by word segmentation techniques and the plurality of words are used as a plurality of operators in an operator model. It should be further noted that the duration of the above-mentioned preset period needs to be sufficiently short, and may be 1 minute, 5 minutes, or the like.
To help understand the above embodiment, for example, a control instruction sent by a user to an air conditioner within one month is obtained, where the instruction may be a voice instruction or a text instruction input on a terminal, and a result of semantic recognition of the control instruction by the device is obtained, and it is determined whether an execution result of the device is adjusted within one minute after the device finishes executing the control instruction, for example, the semantic instruction of the user is: the temperature of the air conditioner is too high, and the semantic recognition result of the air conditioner is that: the air conditioner temperature is lowered, and thus, the air conditioner lowers the temperature. At this time, if the user does not adjust the air conditioner, the semantic recognition result of the air conditioner accords with the user requirement, the semantic recognition result is matched with the control instruction, and the 'temperature of the air conditioner is reduced' by the word segmentation technology to be split into three operators: regulating down, refrigerator and temperature.
In an exemplary embodiment, the identifying module 62 is further configured to receive a defining operation of the second object, and determine an operator type and an operator attribute of each of all operators in the operator model according to the defining operation, where the operator type includes: action word type, entity word type, parameter word type; the operator attributes include: the meaning of each operator, whether each operator refers to other operators, whether each operator participates in operator combination, and whether each operator participates in operator output.
To assist in understanding the above embodiments, control is an action word type, a device word type, and 37 degrees is a parameter word type, for example. It should be further noted that the parameter word types include controllable parameters such as temperature, humidity, volume, etc.
Optionally, in this embodiment, the definition of each operator may further include: description of operators. The description of the operator may be a Chinese description annotation of the operator, as shown in FIG. 3, FIG. 3 is a schematic diagram of an alternative operator definition method according to an embodiment of the present invention, and in FIG. 3, the description of the operator open is; open, i.e. the chinese name of open, open can be referenced, i.e. an operator referring to open is described as present, it being noted that this reference relation is that the two operators established in the presence of the reference relation are similar, e.g. both operator start (chinese meaning: open) and operator begin (chinese meaning: open) can be referenced to each other, since the meanings of both are similar. Whether an operator is a real word is defined because in the case where an operator is a real word, the operator's output may be involved, e.g., the operator's "not involved" output is an imaginary word having no temporal meaning.
In an exemplary embodiment, the recognition module 62 performs operator extraction on the first control instruction to obtain all operators corresponding to the first control instruction before performing semantic recognition on the plurality of operators corresponding to the first control instruction through the defined operator model; determining operators with ambiguity in all operators; and taking other operators except the ambiguous operator in all operators as a plurality of operators of the first control instruction.
Optionally, in this embodiment, the apparatus further includes: the extraction module is used for carrying out operator extraction on the first control instruction in the following manner: the operator extraction is carried out by adopting a model, wherein the model is not an operator model, but a translation model, a common entity recognition model and the like can recognize the first control instruction as the model of the operator; operator extraction is carried out through the regular expression, namely operators corresponding to the first control instruction are matched through the regular expression; and extracting operators by a dictionary matching method, namely matching operators corresponding to the first control instruction by a preset dictionary.
Optionally, in this embodiment, the apparatus further includes: the determining module is used for determining operators with ambiguity in all operators by the following modes: disambiguation rules are preset, disambiguation is performed based on preset disambiguation rules, for example, the following rules are preset: when the meaning of the word of the action word type and the meaning of the word of the parameter word type are repeated, eliminating the word of the action word type, and based on the rule, extracting all operators as follows: open, reduce, mode, operator to determine ambiguity is: the delete, therefore, only retains the two operators open and mode.
In an exemplary embodiment, the identifying module 62 is further configured to combine the plurality of operators according to the definition of each operator in the plurality of operators corresponding to the first control instruction by using the defined operator model, to obtain an operator combination of the plurality of operators; combining the operators as semantics of the plurality of operators. To assist in understanding the above embodiments, the above operators may be, for example: the operator 'adjusting', the operator 'air conditioning', and the operator '35 DEG', the three operators can be combined, and the air conditioning temperature is adjusted to 35 deg.
It should be noted that, the first control instruction corresponds to a plurality of operators, but the operators are split, and a single operator cannot generally explain the intention of the first control instruction of the user, for example, operator "open", and cannot know which device needs to be opened by the user. The operators have a certain logic relation, and the operators can be combined through the logic relation to obtain the actual intention of the user. For example, by combining open and bridge, it is possible to obtain that the intention of the user is to open the refrigerator, further, the combination may be performed by using language custom logic between operators, for example, the combination of action word type and equipment word type, or the combination rule may be preset by the first object, or the retrospective inference may be performed, that is, by retrospecting the first control instruction, the arrangement sequence of the operators is determined, and then the combination is performed according to the arrangement sequence.
In an exemplary embodiment, the identifying module 62 is further configured to determine whether an operator type of a first operator of the plurality of operators is an action word type and whether the first operator participates in an operator combination, and determine whether an operator type of a second operator of the plurality of operators is a device word type and whether the second operator participates in an operator combination, and determine whether there is an operator of a parameter word type in the plurality of operators; and combining the first operator and the second operator through the defined operator model under the condition that the operator type of the first operator is an action word type and the first operator participates in operator combination, the operator type of the second operator is a device word type and the second operator participates in operator combination, and no operator of the parameter word type exists in the operators. And combining the first operator and the third operator through the defined operator model under the condition that the first operator is of an action word type and the first operator participates in operator combination, the second operator is of a device word type and the second operator participates in operator combination, and the third operator is of a parameter word type and the third operator participates in operator combination.
It should be noted that, in the above embodiment, in the case where the first operator is of an action word type and the first operator may participate in the operator combination, the first operator may be combined with the operator of the parameter word type and the operator of the device word type, and it is also a precondition that the operator of the parameter word type and the operator of the device word type combined with the first operator may participate in the operator combination. Such as: the operator open of the action word type and the operator refrigerator of the device word type can be combined into an openbridge in case both are defined as being participatable in the operator combination. Further, the operator of the parameter word type and the operator of the equipment word type are combined with each other by the combination priority, the operator of the action word type and the operator of the parameter word type are combined with each other by the combination priority, and the operator of the action word type and the operator of the equipment word type are combined under the condition that the operator of the parameter word type is not present or the operator of the parameter word type is present but the operator of the parameter word type is not participated in the combination of the operators. For example, an operator open of action word type, an operator bridge of device word type, an operator holidy of parameter word type (mode), all of which are defined as being able to participate in the operator combination, the result of the operator combination is openMode.
In an exemplary embodiment, the conversion module 64 is further configured to identify semantic intents of the plurality of operators by using a preset intent identification manner, so as to obtain semantic identification results of the semantics of the plurality of operators; filling the slots in the semantic recognition result to obtain a slot filling result; and taking the slot filling result as a second control instruction which is allowed to be executed by the target equipment.
The intention recognition identifies the semantic device semantics of the plurality of operators to be converted into the understandable intention of the device, and if the intention recognition result is a direct intention, for example, the intention recognition result is that ice is opened, the box directly outputs the intention recognition result, and if the intention recognition result is that some word is intended, the slot filling is needed. For example, the first control instruction of the user is to open the holiday mode of the refrigerator, and the result obtained after the operator identification and the semantic identification is that: after intention recognition, the intention of the user obtained by the user terminal is as follows: the setMode is not able to obtain the direct intention of the user, and thus a slot filling is required, and the result of the slot filling is setmode=hotday device=frame.
In addition, when slot filling is performed, some slots are not extracted from the semantics of multiple operators, and default slots need to be filled in during configuration. Such as: today there is a slot corresponding to rain: time=today type=weather, wherein time is corpus extraction slot, and type is default slot set in advance.
Optionally, in this embodiment, converting semantics of the plurality of operators into a second control instruction that the target device is allowed to execute includes: constructing a multi-intention recognition model based on cluster pre-analysis, and recognizing a plurality of intentions corresponding to the semantics of a plurality of operators; constructing a BiLSTM-CRF semantic Slot filling model based on a Slot-connected association gate mechanism, and guiding the filling of the semantic slots by using the result of intention recognition; and optimizing and training a joint model formed by the BERT model, the multi-intention recognition model and the semantic slot filling model, and recognizing by utilizing the joint model completed by optimizing and training. It should be noted that, in the embodiment, the connection between the intent recognition and the semantic slot filling is fully considered, and a joint modeling method is provided, the two tasks of the intent recognition and the slot filling are combined into one task, and the accuracy of the multi-intent recognition is improved, and the accuracy of the semantic slot filling is also improved, so that the accuracy of converting the semantics of a plurality of operators into the second control instruction is improved.
An embodiment of the present invention also provides a storage medium including a stored program, wherein the program executes the method of any one of the above.
Alternatively, in the present embodiment, the above-described storage medium may be configured to store program code for performing the steps of:
s1, under the condition that a first control instruction sent by a first object is received, carrying out semantic recognition on a plurality of operators corresponding to the first control instruction through a defined operator model to obtain semantics of the operators, wherein the first control instruction is used for controlling target equipment, and the defined operator model comprises definitions of the operators;
s2, converting the semantics of the operators into a second control instruction which is allowed to be executed by the target equipment.
An embodiment of the invention also provides an electronic device comprising a memory having stored therein a computer program and a processor arranged to run the computer program to perform the steps of any of the method embodiments described above.
Optionally, the electronic apparatus may further include a transmission device and an input/output device, where the transmission device is connected to the processor, and the input/output device is connected to the processor.
Alternatively, in the present embodiment, the above-described processor may be configured to execute the following steps by a computer program:
s1, under the condition that a first control instruction sent by a first object is received, carrying out semantic recognition on a plurality of operators corresponding to the first control instruction through a defined operator model to obtain semantics of the operators, wherein the first control instruction is used for controlling target equipment, and the defined operator model comprises definitions of the operators;
s2, converting the semantics of the operators into a second control instruction which is allowed to be executed by the target equipment.
Alternatively, in the present embodiment, the storage medium may include, but is not limited to: a U-disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a removable hard disk, a magnetic disk, or an optical disk, or other various media capable of storing program codes.
Alternatively, specific examples in this embodiment may refer to examples described in the foregoing embodiments and optional implementations, and this embodiment is not described herein.
It will be appreciated by those skilled in the art that the modules or steps of the invention described above may be implemented in a general purpose computing device, they may be concentrated on a single computing device, or distributed across a network of computing devices, they may alternatively be implemented in program code executable by computing devices, so that they may be stored in a memory device for execution by computing devices, and in some cases, the steps shown or described may be performed in a different order than that shown or described, or they may be separately fabricated into individual integrated circuit modules, or multiple modules or steps within them may be fabricated into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, but various modifications and variations can be made to the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the principle of the present invention should be included in the protection scope of the present invention.
Claims (11)
1. A semantic conversion method, comprising:
under the condition that a first control instruction sent by a first object is received, carrying out semantic recognition on a plurality of operators corresponding to the first control instruction through a defined operator model to obtain semantics of the operators, wherein the first control instruction is used for controlling target equipment, and the defined operator model comprises definitions of the operators;
the semantics of the plurality of operators are converted into a second control instruction which the target device is allowed to execute.
2. The semantic conversion method according to claim 1, wherein, in the case of receiving a first control instruction sent by a first object, before performing semantic recognition on a plurality of operators corresponding to the first control instruction by using a defined operator model, the method further comprises:
Determining a definition of each of all operators in the operator model, wherein the definition of each operator comprises: the operator type and operator attribute of each operator.
3. The semantic conversion method according to claim 2, wherein determining a definition of each of all operators in the operator model comprises:
receiving a definition operation of a second object, and determining an operator type and an operator attribute of each operator in all operators in the operator model according to the definition operation, wherein the operator type comprises: action word type, entity word type, parameter word type; the operator attributes include: the meaning of each operator, whether each operator refers to other operators, whether each operator participates in operator combination, and whether each operator participates in operator output.
4. The semantic conversion method according to claim 1, wherein performing semantic recognition on the plurality of operators corresponding to the first control instruction through the defined operator model includes:
combining the operators according to the definition of each operator in the plurality of operators corresponding to the first control instruction through the defined operator model to obtain an operator combination of the operators;
Combining the operators as semantics of the plurality of operators.
5. The semantic conversion method according to claim 4, wherein combining the plurality of operators according to the definition of each of the plurality of operators corresponding to the first control instruction by the defined operator model comprises:
determining whether an operator type of a first operator of the plurality of operators is an action word type and whether the first operator participates in operator combination, determining whether an operator type of a second operator of the plurality of operators is a device word type and whether the second operator participates in operator combination, and determining whether an operator of a parameter word type exists in the plurality of operators;
and combining the first operator and the second operator through the defined operator model under the condition that the operator type of the first operator is an action word type and the first operator participates in operator combination, the operator type of the second operator is a device word type and the second operator participates in operator combination, and no operator of the parameter word type exists in the operators.
6. The semantic conversion method according to claim 4, wherein combining the plurality of operators according to the definition of each of the plurality of operators corresponding to the first control instruction by the defined operator model comprises:
Determining whether an operator type of a first operator of the plurality of operators is an action word type and whether the first operator participates in an operator combination, and determining whether an operator type of a second operator of the plurality of operators is a device word type and whether the second operator participates in an operator combination, and determining whether an operator type of a third operator of the plurality of operators is a parameter word type and whether the third operator participates in an operator combination;
and combining the first operator and the third operator through the defined operator model under the condition that the first operator is of an action word type and the first operator participates in operator combination, the second operator is of a device word type and the second operator participates in operator combination, and the third operator is of a parameter word type and the third operator participates in operator combination.
7. The semantic conversion method according to claim 1, wherein before performing semantic recognition on the plurality of operators corresponding to the first control instruction by using the defined operator model, the method further comprises:
operator extraction is carried out on the first control instruction so as to obtain all operators corresponding to the first control instruction;
Determining operators with ambiguity in all operators;
and taking other operators except the ambiguous operator in all operators as a plurality of operators of the first control instruction.
8. The semantic conversion method according to claim 1, wherein converting the semantics of the plurality of operators into the second control instruction that the target device is permitted to execute, comprises:
identifying semantic intentions of the operators in a preset intention identification mode to obtain semantic identification results of the semantics of the operators;
filling the slots in the semantic recognition result to obtain a slot filling result;
and taking the slot filling result as a second control instruction which is allowed to be executed by the target equipment.
9. A semantic conversion device, comprising:
the identification module is used for carrying out semantic identification on a plurality of operators corresponding to a first control instruction through a defined operator model under the condition that the first control instruction sent by a first object is received, so as to obtain the semantics of the plurality of operators, wherein the first control instruction is used for controlling target equipment, and the defined operator model comprises definitions of the plurality of operators;
And the conversion module is used for converting the semantics of the operators into a second control instruction which is allowed to be executed by the target equipment.
10. A computer readable storage medium, characterized in that the computer readable storage medium comprises a stored program, wherein the program when run performs the method of any one of claims 1 to 8.
11. An electronic device comprising a memory and a processor, characterized in that the memory has stored therein a computer program, the processor being arranged to execute the method according to any of the claims 1 to 8 by means of the computer program.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210467403.1A CN117010378A (en) | 2022-04-29 | 2022-04-29 | Semantic conversion method and device, storage medium and electronic device |
| PCT/CN2022/097583 WO2023206723A1 (en) | 2022-04-29 | 2022-06-08 | Semantic transformation method and apparatus, and storage medium and electronic apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210467403.1A CN117010378A (en) | 2022-04-29 | 2022-04-29 | Semantic conversion method and device, storage medium and electronic device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117010378A true CN117010378A (en) | 2023-11-07 |
Family
ID=88517122
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210467403.1A Pending CN117010378A (en) | 2022-04-29 | 2022-04-29 | Semantic conversion method and device, storage medium and electronic device |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN117010378A (en) |
| WO (1) | WO2023206723A1 (en) |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105206275A (en) * | 2015-08-31 | 2015-12-30 | 小米科技有限责任公司 | Device control method, apparatus and terminal |
| US10305826B1 (en) * | 2018-05-03 | 2019-05-28 | Progressive Casualty Insurance Company | Intelligent conversational systems |
| CN110866094B (en) * | 2018-08-13 | 2021-04-09 | 珠海格力电器股份有限公司 | Instruction recognition method, instruction recognition device, storage medium, and electronic device |
| CN110597082A (en) * | 2019-10-23 | 2019-12-20 | 北京声智科技有限公司 | Intelligent household equipment control method and device, computer equipment and storage medium |
| CN111048083A (en) * | 2019-12-12 | 2020-04-21 | 深圳康佳电子科技有限公司 | Voice control method, device and storage medium |
-
2022
- 2022-04-29 CN CN202210467403.1A patent/CN117010378A/en active Pending
- 2022-06-08 WO PCT/CN2022/097583 patent/WO2023206723A1/en unknown
Also Published As
| Publication number | Publication date |
|---|---|
| WO2023206723A1 (en) | 2023-11-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112051743A (en) | Device control method, conflict processing method, corresponding devices and electronic device | |
| CN110262273A (en) | A kind of home equipment control method, device, storage medium and smart home system | |
| CN109684456B (en) | Scene ability intelligent question-answering system based on Internet of things ability knowledge graph | |
| CN116229955B (en) | Interactive intention information determining method based on generated pre-training GPT model | |
| WO2023168838A1 (en) | Sentence text recognition method and apparatus, and storage medium and electronic apparatus | |
| US11902043B2 (en) | Self-learning home system and framework for autonomous home operation | |
| CN110414582B (en) | Model training method and device, computing equipment and storage medium | |
| WO2024001101A1 (en) | Text intention recognition method and apparatus, storage medium, and electronic apparatus | |
| CN110895936B (en) | Voice processing method and device based on household appliance | |
| WO2024001196A1 (en) | Household appliance control method and apparatus, storage medium, and electronic apparatus | |
| CN117010378A (en) | Semantic conversion method and device, storage medium and electronic device | |
| CN114925158A (en) | Sentence text intention recognition method and device, storage medium and electronic device | |
| CN114915514B (en) | Method and device for processing intention, storage medium and electronic device | |
| CN111128135B (en) | Voice communication method and device | |
| CN117689020B (en) | Method and device for constructing intelligent home body based on large model and electronic equipment | |
| CN117725995B (en) | Knowledge graph construction method, device and medium based on large model | |
| CN117725908B (en) | Time analysis method, device and equipment in natural language processing | |
| CN116913274A (en) | Scene generation method, device and storage medium based on generation type large model | |
| CN113378826B (en) | Data processing method, device, equipment and storage medium | |
| WO2023212993A1 (en) | Appliance control method, storage medium, and electronic device | |
| CN114911535B (en) | Application program component configuration method, storage medium and electronic device | |
| CN116090461A (en) | Intent recognition method of control instruction, storage medium and electronic device | |
| CN117689020A (en) | Method and device for constructing intelligent home body based on large model and electronic equipment | |
| CN115547331A (en) | Voice processing method, processing device, storage medium and electronic device | |
| CN116432658A (en) | Voice data processing method and device, storage medium and electronic device |
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 |