CN109710745A - Information processing method, device and electronic equipment - Google Patents
Information processing method, device and electronic equipment Download PDFInfo
- Publication number
- CN109710745A CN109710745A CN201811627054.5A CN201811627054A CN109710745A CN 109710745 A CN109710745 A CN 109710745A CN 201811627054 A CN201811627054 A CN 201811627054A CN 109710745 A CN109710745 A CN 109710745A
- Authority
- CN
- China
- Prior art keywords
- vocabulary
- information
- enterprise
- exclusive
- lexicon
- 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
Abstract
The embodiment of the present invention provides a kind of information processing method, device and electronic equipment, this method comprises: obtaining the input information that smart machine receives, inputting includes the mark of enterprise belonging to smart machine in message；It obtains enterprise and identifies corresponding exclusive lexicon；And according to the vocabulary in exclusive lexicon, identifying processing is carried out to input information；Further according to identifying processing as a result, control smart machine executes corresponding operation.Information processing method, device and electronic equipment provided in an embodiment of the present invention, improve the accuracy of recognition result.
Description
Technical field
The present embodiments relate to field of terminal technology more particularly to a kind of information processing methods, device and electronic equipment.
Background technique
In the prior art, when terminal collects the voice messaging of user, which is reported into server and is carried out
Processing, server identifies the voice messaging by ASR model, corresponding text information is obtained, by NLU model to this
Text information is handled, to carry out semantic understanding, thus according to semantic understanding as a result, determining the corresponding sound of the voice messaging
Information is answered, and controls smart machine and exports the response message.When control smart machine exports the response message, mainly NLU
Model is based on pre-stored inquiry request in inquiry request library and carries out semantic understanding to the corresponding text information of voice messaging.
However, being carried out when being identified by ASR model to voice messaging, or through NLP model to text information
When semantic understanding (can be understood as the identification to text information), the vocabulary trained can only be accurately identified, and instructed
When white silk, it is impossible to cover all vocabulary of all spectra, therefore, for untrained vocabulary, can not carry out accurately
Identification, so that the accuracy of recognition result is not high.
Summary of the invention
The embodiment of the present invention provides a kind of information processing method, device and electronic equipment, to improve the accurate of recognition result
Degree.
In a first aspect, the embodiment of the present invention provides a kind of information processing method, which may include:
The input information that smart machine receives is obtained, includes enterprise belonging to the smart machine in the input message
Mark；
It obtains the enterprise and identifies corresponding exclusive lexicon；
According to the vocabulary in the exclusive lexicon, identifying processing is carried out to the input information；
According to identifying processing as a result, controlling the smart machine executes corresponding operation.
In one possible implementation, the vocabulary in the exclusive lexicon includes the exclusive word of enterprise and/or described
The corresponding synonym of the exclusive word of enterprise.
In one possible implementation, described according to the vocabulary when the input information is voice messaging, it is right
The input information carries out identifying processing, comprising:
The input information is identified by automatic speech recognition ASR model, it is corresponding to obtain the input information
Text information；
If including target word corresponding with any vocabulary in the exclusive lexicon in the text information, by the text
Target word in this information replaces with any vocabulary, the text information after obtaining error correction；
It is described according to identifying processing as a result, controlling the smart machine executes corresponding operation, comprising:
According to the text information after the error correction, controls the smart machine and execute corresponding operation.
In one possible implementation, the text information according to after the error correction, controls the smart machine
Execute corresponding operation, comprising:
Text information after controlling the smart machine output error correction.
In one possible implementation, described according to the vocabulary, identifying processing, packet are carried out to the input information
It includes:
If comprising any vocabulary in the exclusive lexicon in the input information, according to any vocabulary, to institute
It states input information and carries out word segmentation processing；
Based on after word segmentation processing as a result, obtaining semantic parsing result to the semantic parsing of input information progress；
It is described according to identifying processing as a result, controlling the smart machine executes corresponding operation, comprising:
According to the semantic parsing result, controls the smart machine and execute corresponding operation.
In one possible implementation, the method also includes:
The first lexical information of input is obtained, first lexical information includes the first vocabulary and enterprise's mark of input；
At least one corresponding first vocabulary is identified according to each enterprise mark and the enterprise, establishes each enterprise's mark pair
The exclusive lexicon answered.
In one possible implementation, it establishes after the corresponding exclusive lexicon of each enterprise's mark, the method
Further include:
Obtain the second lexical information of input；Wherein, second lexical information includes at least one second vocabulary and enterprise
Industry mark；
It, will if second vocabulary identifies any vocabulary in corresponding exclusive lexicon with the enterprise and is all different
Second vocabulary is added to the enterprise and identifies in corresponding exclusive lexicon.
In one possible implementation, the method also includes:
If second vocabulary is identified with the enterprise, any vocabulary in corresponding exclusive lexicon is identical, and output mentions
Show that information, the prompt information have included second vocabulary for prompting the enterprise to identify in corresponding exclusive lexicon.
Second aspect, the embodiment of the present invention provide a kind of information processing unit, which may include:
Acquiring unit, the input information received for obtaining smart machine include the intelligence in the input message
The mark of enterprise belonging to equipment；And it obtains the enterprise and identifies corresponding exclusive lexicon；
Processing unit, for carrying out identifying processing to the input information according to the vocabulary in the exclusive lexicon；
Control unit, for according to identifying processing as a result, controlling the smart machine executes corresponding operation.
In one possible implementation, the vocabulary in the exclusive lexicon includes the exclusive word of enterprise and/or described
The corresponding synonym of the exclusive word of enterprise.
In one possible implementation, when the input information is voice messaging,
The processing unit is obtained specifically for being identified by automatic speech recognition ASR model to the input information
To the corresponding text information of the input information；If including and any vocabulary in the exclusive lexicon in the text information
Target word in the text information is replaced with any vocabulary, the text information after obtaining error correction by corresponding target word；
Described control unit, specifically for controlling the smart machine execution pair according to the text information after the error correction
The operation answered.
In one possible implementation, described control unit is specifically used for controlling the smart machine output error correction
Text information afterwards.
In one possible implementation, the processing unit, if being specifically used in the input information comprising described
Any vocabulary in exclusive lexicon carries out word segmentation processing to the input information according to any vocabulary；At participle
It is after reason as a result, carry out semantic parsing to the input information, obtain semantic parsing result；
Described control unit is specifically used for that it is corresponding to control the smart machine execution according to the semantic parsing result
Operation.
In one possible implementation, described device further includes establishing unit；
The acquiring unit, is also used to obtain the first lexical information of input, and first lexical information includes input
First vocabulary and enterprise's mark；
It is described to establish unit, for identifying at least one corresponding first vocabulary according to each enterprise mark and the enterprise,
It establishes each enterprise and identifies corresponding exclusive lexicon.
In one possible implementation, the acquiring unit is also used to obtain the second lexical information of input；Its
In, second lexical information includes at least one second vocabulary and enterprise's mark；
It is described to establish unit, appointing in corresponding exclusive lexicon is identified with the enterprise if being also used to second vocabulary
Meaning vocabulary is all different, then second vocabulary is added to the enterprise and identified in corresponding exclusive lexicon.
In one possible implementation, described control unit, if being also used to second vocabulary and enterprise mark
Any vocabulary known in corresponding exclusive lexicon is identical, then exports prompt information, the prompt information is for prompting the enterprise
It has included second vocabulary that industry, which identifies in corresponding exclusive lexicon,.
The third aspect, the embodiment of the present invention provide a kind of electronic equipment, including processor and memory, wherein described to deposit
Reservoir is for storing program instruction；
The processor is used to read the program instruction in the memory, and according to the program instruction in the memory
Execute information processing method described in first aspect.
Fourth aspect, the embodiment of the present invention provide a kind of computer readable storage medium, the computer-readable storage medium
Matter is stored with computer program, when the computer program is executed by processor, executes information processing described in first aspect
Method.
Information processing method, device and electronic equipment provided in an embodiment of the present invention are obtaining what smart machine received
After inputting information, inputting in message includes the mark of enterprise belonging to smart machine；It obtains enterprise and identifies corresponding exclusive lexicon；
And according to the vocabulary in exclusive lexicon, identifying processing is carried out to input information；Further according to identifying processing as a result, control intelligence
Equipment executes corresponding operation.It can be seen that information processing method provided in an embodiment of the present invention, knows to input information
It is according to including that the vocabulary that identifies in corresponding exclusive lexicon of enterprise belonging to smart machine is defeated to this in input message when other
Enter information and carry out identifying processing, to improve the accuracy of recognition result.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is this hair
Bright some embodiments for those of ordinary skill in the art without any creative labor, can be with
It obtains other drawings based on these drawings.
Fig. 1 is a kind of flow diagram of information processing method provided in an embodiment of the present invention；
Fig. 2 is the flow diagram of another information processing method provided in an embodiment of the present invention；
Fig. 3 is the flow diagram of another information processing method provided in an embodiment of the present invention；
Fig. 4 is a kind of structural schematic diagram of information processing unit provided in an embodiment of the present invention；
Fig. 5 is the structural schematic diagram of another information processing unit provided in an embodiment of the present invention；
Fig. 6 is the structural schematic diagram of a kind of electronic equipment provided in an embodiment of the present invention.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art
Every other embodiment obtained without creative efforts, shall fall within the protection scope of the present invention.
Description and claims of this specification and term " first ", " second ", " third " " in above-mentioned attached drawing
The (if present)s such as four " are to be used to distinguish similar objects, without for describing exclusive sequence or precedence.It should manage
The data that solution uses in this way are interchangeable under appropriate circumstances, so that the embodiment of the present invention described herein for example can be to remove
Sequence other than those of illustrating or describe herein is implemented.In addition, term " includes " and " having " and theirs is any
Deformation, it is intended that cover it is non-exclusive include, for example, containing the process, method of a series of steps or units, system, production
Product or equipment those of are not necessarily limited to be clearly listed step or unit, but may include be not clearly listed or for this
A little process, methods, the other step or units of product or equipment inherently.
In the prior art, when being identified by ASR model to voice messaging, or by NLP model to text envelope
Breath, which carries out semantic understanding, (can be understood as the identification to text information semanteme, i.e. field (domain), intention (intent) and slot
The identification that position (slot) is extracted) when, the vocabulary trained can only be accurately identified, it, can not for untrained vocabulary
It is accurately identified, so that the accuracy of recognition result is not high.In order to improve the accuracy of recognition result, the present invention is real
It applies example and provides a kind of information processing method, after obtaining the input information that smart machine receives, inputting includes intelligence in message
Enterprise's mark belonging to energy equipment；It obtains enterprise and identifies corresponding exclusive lexicon；It is right and according to the vocabulary in exclusive lexicon
It inputs information and carries out identifying processing；Further according to identifying processing as a result, control smart machine executes corresponding operation.Thus may be used
See, information processing method provided in an embodiment of the present invention, is to include according in input message when being identified to input information
Enterprise belonging to smart machine identifies the vocabulary in corresponding exclusive lexicon and carries out identifying processing to the input information, to mention
The high accuracy of recognition result.
In the following, how will be solved with specific embodiment to technical solution of the present invention and technical solution of the present invention
Technical problem is stated to be described in detail.These specific embodiments can be combined with each other below, for the same or similar general
It reads or process repeats no more in certain embodiments.Below in conjunction with attached drawing, the embodiment of the present invention is described.
Fig. 1 is a kind of flow diagram of information processing method provided in an embodiment of the present invention, and this method can be by servicing
Device executes, and can also be executed by the controller of smart machine.Exemplary, shown in Figure 1, which can wrap
It includes:
S101, the input information that smart machine receives is obtained.
Wherein, inputting in message includes the mark of enterprise belonging to smart machine.Inputting information can be voice messaging, can also
Think text information.Exemplary, which can be " I wants to see fine work film ", and certainly, input information may be " to ask
Next week X " is introduced, " all X " here is the leader or employee of some enterprise.It is of course also possible to be other information, here, this hair
Bright embodiment is only illustrated so that both input information as an example, but is not represented the embodiment of the present invention and be limited only to this.
For an enterprise, multiple smart machines (such as robot) can be set, each smart machine meeting
Know enterprise's mark of oneself owned enterprise.Exemplary, enterprise's mark can be the title of enterprise, or the volume of enterprise
Number, as long as can not be limited the specific implementation form of enterprise's mark in the embodiment of the present invention with the unique identification enterprise
It is fixed.
Optionally, when obtaining input information, the input information of user can be acquired by the microphone of smart machine,
The input information that other equipment obtain user can be received by modes such as network or bluetooths, it is of course also possible to pass through intelligence
The screen of equipment obtains the input information of user, here, the embodiment of the present invention is only obtained with these three possible implementations
It is illustrated for the input information of user, but does not represent the embodiment of the present invention and be limited only to this.
After getting the input information including enterprise's mark, so that it may execute following S102:
S102, the corresponding exclusive lexicon of enterprise's mark is obtained.
Optionally, the vocabulary in exclusive lexicon includes the exclusive word of enterprise.The exclusive word of enterprise is primarily referred to as the enterprise (or should
Field or industry etc. where enterprise) distinctive or exclusive vocabulary.For different enterprises, corresponding exclusive word may not
Together.
It further, can also include the corresponding synonym of the exclusive word of enterprise in the exclusive lexicon.In other words, exclusive
Vocabulary in lexicon can only include the exclusive word of enterprise, and the vocabulary in the exclusive lexicon can also be with the exclusive word of enterprise and enterprise
The corresponding synonym of exclusive word.The corresponding synonym of the exclusive word of some enterprise can be understood as the substitute of the exclusive word of the enterprise
Or extensive word, that is, the vocabulary of same or similar semanteme is indicated, by the way that the corresponding synonym of the exclusive word of enterprise is stored in exclusive word
It converges in library, allows to according to the exclusive word of the enterprise and the corresponding synonym of the exclusive word of enterprise, to the more of the same input information
Kind way to put questions is preferably identified, to improve the accuracy of input information identification.It is exemplary, for mobile-phone manufacturers,
When the exclusive word of enterprise is " apple ", semanteme is mobile phone brand, and corresponding synonym can be " iPhone ",
" iphone " or " iphone mobile phone " etc..And for fruit retailer, it is semantic when the exclusive word of enterprise is " apple "
For fruit classification, corresponding synonym can be " Fuji apple ", " Guoguang apple " etc..It should be noted that different enterprises
The exclusive vocabulary of enterprise may be the same or different.
The exclusive lexicon of enterprise can be associated with enterprise's question and answer library (including at least one question and answer to), for example, for hand
For machine manufacturer, the exclusive word of enterprise " apple ", corresponding question and answer to can be, " function of introducing iPhone " and its
Reply message, " iPhone how much " and its reply message etc..
In order to improve the accuracy of input information identification, therefore, each enterprise can be with the respective exclusive word of correspondence establishment
Remittance library, and respective exclusive lexicon is identified by enterprise's mark, receiving the input including enterprise's mark in this way
After information, so that it may its corresponding exclusive lexicon is found according to enterprise's mark, thus according to the word in the exclusive lexicon
It converges and identifying processing is carried out to input information, that is, execute following S103:
S103, according to the vocabulary in exclusive lexicon, identifying processing is carried out to input information.
It should be noted that in embodiments of the present invention, when carrying out identifying processing to input information, when input information
When input mode difference, different identifying processing modes can be corresponded to, it is exemplary, it may include at least two possible realization sides
Formula, specifically:
In one possible implementation, when inputting information is voice messaging, which is carried out at identification
Reason identifies input information it is to be understood that first passing through automatic speech recognition ASR model, obtains the corresponding text of input information
This information；Further according to the vocabulary in exclusive lexicon, identifying processing is carried out to input information.
It is exemplary, to input information as " I wants to see fine work film ", input information is being identified by ASR model
When, which may be identified as " I wants to see that fine work point reflects " to get to text information be " I wants to see that fine work point reflects ",
After obtaining text information " I wants to see that fine work point reflects ", so that it may right according to the vocabulary " fine work film " in the exclusive lexicon
The corresponding text information of input information " I wants to see that fine work point reflects " carries out correction process, obtains the corresponding text envelope of input information
It ceases " I wants to see fine work film ", to complete the speech recognition to the input information.
In alternatively possible implementation, when inputting information is text information, which is identified
Processing carries out word segmentation processing to input information it is to be understood that first according to the vocabulary in exclusive lexicon；It is based on word segmentation processing again
Afterwards as a result, carry out semantic parsing to input information, obtain semantic parsing result.
It is exemplary, to input information for " next week X please be introduce ", due to including the exclusive vocabulary of enterprise " week in the input information
X " therefore can carry out word segmentation processing to input information " please introduce next week X " according to the exclusive vocabulary of the enterprise " all X ", obtain defeated
Enter the corresponding word segmentation processing result of information " next week X please be introduce " " ask-introduce-under-week X ", rather than word segmentation processing is at " asking-be situated between
Continue-next week X ", and then is carried out according to word segmentation processing result " ask-introduce-under-week X " to input information " please introduce next week X "
Semanteme parsing, obtains semantic parsing result, to complete to handle the semantics recognition of the input information.
It should be noted that in the first possible implementation, when input is voice messaging, according to exclusive word
Vocabulary in remittance library carries out identifying processing to input information, after obtaining identifying processing result, can also further execute second
The possible implementation of kind, i.e., continue word segmentation processing to identifying processing result, then based on after word segmentation processing as a result, to defeated
Enter information and carry out semantic parsing, obtains semantic parsing result.
According to the vocabulary in exclusive lexicon, to input information carry out identifying processing obtain identifying processing result it
Afterwards, so that it may execute following S104:
S104, according to identifying processing as a result, control smart machine executes corresponding operation.
Wherein, it may include at least two that control smart machine, which executes corresponding operation, and a kind of operate sets for control intelligence
The result of standby output identifying processing；A kind of operate exports the corresponding reply message of identifying processing result for control smart machine.
It is alternatively possible in conjunction with the description in above-mentioned S103, in above-mentioned S103 in the first possible implementation, when
When input information is voice messaging, according to the vocabulary in exclusive lexicon, identifying processing is carried out to input information and is identified
After processing result, the result of smart machine output identifying processing can control；Second of possible realization in above-mentioned S103
In mode, when inputting information is text information, first according to the vocabulary in exclusive lexicon, input information is carried out at participle
Reason；Again based on after word segmentation processing as a result, after obtaining semantic parsing result, can be controlled to the semantic parsing of input information progress
Smart machine processed exports the corresponding reply message of semantic parsing result.
Information processing method provided in an embodiment of the present invention, after obtaining the input information that smart machine receives, input
It include the mark of enterprise belonging to smart machine in message；It obtains enterprise and identifies corresponding exclusive lexicon；And according to exclusive vocabulary
Vocabulary in library carries out identifying processing to input information；Further according to identifying processing as a result, control smart machine execution is corresponding
Operation.It can be seen that information processing method provided in an embodiment of the present invention is according to defeated when identifying to input information
Enter the vocabulary identified in corresponding exclusive lexicon in message including enterprise belonging to smart machine to know the input information
Other places reason, to improve the accuracy of recognition result.
It is possible it is possible to further two kinds described in combination above-mentioned S103 and S104 based on embodiment shown in FIG. 1
S103 and S104 is described in detail in implementation, according to the vocabulary in exclusive lexicon, carries out identifying processing to input information；
And according to identifying processing as a result, control smart machine executes corresponding operation.
In one possible implementation, exemplary when inputting information is voice messaging, reference can be made to shown in Fig. 2, figure
2 be the flow diagram of another information processing method provided in an embodiment of the present invention, and this method may include:
S201, input information is identified by automatic speech recognition ASR model, obtains the corresponding text of input information
Information.
Exemplary, input information can be " I wants to see fine work film ", identified by ASR model to input information
When, which may be identified as to " I wants to see that fine work point reflects ", " I thinks to obtain the corresponding text information of input information
See that fine work point reflects ".
If including target word corresponding with any vocabulary in exclusive lexicon in S202, text information, by text information
In target word replace with any vocabulary, the text information after obtaining error correction.
Wherein, target word can be understood as in exclusive lexicon terminology match degree or similarity-rough set it is high, but
Not exactly the same vocabulary.Exemplary, if the vocabulary in exclusive lexicon includes " fine work film ", " I wants to see text information
Fine work point reflects " in " fine work point reflects " can be understood as the target word of the vocabulary " fine work film " in exclusive lexicon.
It is exemplary, after obtaining text information " I wants to see that fine work point reflects " by above-mentioned S201, it can be determined that the text
Whether including the corresponding target word of any vocabulary in exclusive lexicon in information " I wants to see that fine work point reflects ", since the text is believed
The target word " fine work point reflects " including the vocabulary " fine work film " in exclusive lexicon in " I wants to see that fine work point reflects " is ceased, then is illustrated
When carrying out speech recognition to input information, mistake can be identified, it therefore, can be in text information " I wants to see that fine work point reflects "
The vocabulary " fine work film " that " fine work point reflects " is revised as in exclusive lexicon is to get to the text information after error correction, " I wants to see essence
Product film ".
After obtaining the text information after error correction, so that it may control smart machine and execute corresponding operation, that is, under executing
State S203.
It should be noted that if comprising any vocabulary in exclusive lexicon in text information, then without executing at error correction
Reason is directly entered and handles in next step；If in text information not only do not include exclusive lexicon in any vocabulary, but also do not include with it is special
Belong to the corresponding target word of any vocabulary in lexicon, without correction process is executed, is directly entered and handles in next step.
S203, according to the text information after the error correction, control the smart machine and execute corresponding operation.
Optionally, in embodiments of the present invention, it according to the text information after the error correction, controls the smart machine and executes
Corresponding operation may include: the text information after the control smart machine output error correction, it is of course also possible to being entangled
It text information after mistake and then further dissection process is executed to the text information after the error correction operates, further according to parsing
The result control smart machine of processing executes corresponding operation, that is, executes following technical sides when inputting information and being text information
Case.
It is exemplary, after obtaining the text information after error correction " I wants to see fine work film ", so that it may control smart machine
Text information " I wants to see fine work film " after exporting the error correction, so that user can directly be obtained by the screen of smart machine
To the corresponding speech recognition result of input information.
It can be seen that information processing method provided in an embodiment of the present invention, to input information " I wants to see fine work film "
When being identified, be according to input message " I wants to see fine work film " in include smart machine belonging to enterprise mark it is corresponding specially
The vocabulary " fine work film " belonged in lexicon carries out identifying processing to the input information " I wants to see fine work film ", to improve
The accuracy of recognition result.
It is exemplary when inputting information is text information in alternatively possible implementation, reference can be made to shown in Fig. 3,
Fig. 3 is the flow diagram of another information processing method provided in an embodiment of the present invention, and this method may include:
If comprising any vocabulary in exclusive lexicon in S301, input information, according to any vocabulary, to input information into
Row word segmentation processing.
Exemplary, input information can be " please introduce next week X ", if the vocabulary in exclusive lexicon includes " all X ", due to
Include in the input information " next week X please be introduce " the exclusive vocabulary of enterprise " all X ", then it can be according to the vocabulary in the exclusive lexicon
" all X " carries out word segmentation processing to the input information " please introduce next week X ", obtains corresponding point of input information " please introduce next week X "
Word processing result " ask-introduce-under-week X ", rather than word segmentation processing is at " asking-introduce-next week X ".
It should be noted that if not including any vocabulary in exclusive lexicon in input information, then it is based on generic way,
Word segmentation processing is carried out to input information.
S302, based on after word segmentation processing as a result, carry out semantic parsing to input information, obtain semantic parsing result.
It is exemplary, after obtaining word segmentation processing result " ask-introduce-under-week X " by above-mentioned S301, so that it may according to
The word segmentation processing result " ask-introduce-under-week X " and semantic parsing is carried out to input information " please introduce next week X ", obtain semantic solution
Analyse result.
After obtaining the corresponding semantic parsing result of input information, so that it may according to the semantic parsing result, control
The smart machine executes corresponding operation, that is, executes following S303:
S303, according to the semantic parsing result, control the smart machine and execute corresponding operation.
After obtaining the corresponding semantic parsing result of input information by above-mentioned S302, so that it may be parsed according to the semanteme
As a result the corresponding reply message of input information is obtained, so that controlling smart machine exports the corresponding answer letter of semantic parsing result
Breath.
It is exemplary, obtaining the corresponding semantic parsing result of input information " next week X please be introduce " " ask-introduce-under-week X "
Later, so that it may it is corresponding that input information " please introduce next week X " be obtained according to the semanteme parsing result " ask-introduce-under-week X "
Reply message, so that controlling smart machine exports the corresponding reply message of semantic parsing result.
It can be seen that information processing method provided in an embodiment of the present invention, carries out to input information " please introduce next week X "
It is according to including that enterprise belonging to smart machine identifies corresponding exclusive vocabulary in input message " next week X please be introduce " when identification
Vocabulary " all X " in library carries out word segmentation processing to the input information " please introduce next week X ", and obtaining input information " please introduce next week
The corresponding word segmentation processing result of X " " ask-introduce-under-week X ", and according to the word segmentation processing result " ask-introduce-under-week X " to defeated
Enter information " next week X please be introduce " and carries out semantic parsing, to obtain the corresponding reply message of input information " next week X please be introduce ", from
And control smart machine and export the corresponding reply message of semantic parsing result, to improve input information " next week X please be introduce "
The accuracy of recognition result.
Optionally, in the embodiment of the present invention, before executing technical solution shown in FIG. 1, the of input can first be obtained
One lexical information, the first lexical information include the first vocabulary and enterprise's mark of input；And it is marked according to each enterprise mark and enterprise
Know at least one corresponding first vocabulary, establishes each enterprise and identify corresponding exclusive lexicon.It is exemplary, obtaining input
When the first lexical information, the first lexical information of input can be obtained by the enterprise platform of each enterprise, an enterprise can wrap
Include at least one enterprise platform.
It should be noted that in embodiments of the present invention, all being needed before executing technical solution shown in FIG. 1 every time
The first lexical information of input is obtained, the first lexical information includes the first vocabulary and enterprise's mark of input；And according to each enterprise
Industry mark and enterprise identify at least one corresponding first vocabulary, establish each enterprise and identify corresponding exclusive lexicon, but
Before executing technical solution shown in FIG. 1 for the first time, the first lexical information of input is obtained, the first lexical information includes input
The first vocabulary and enterprise mark；And at least one corresponding first vocabulary is identified according to each enterprise mark and enterprise, it establishes every
A enterprise identifies corresponding exclusive lexicon, can be directly special according to this in this way in subsequent execution information processing method
Belong to lexicon and identifying processing is carried out to input information.
In addition it is also necessary to which explanation identifies corresponding exclusive vocabulary establishing each enterprise for enterprise customer
After library, when there is new exclusive vocabulary to generate, enterprise customer can send the second lexical information in respective enterprise platform,
In, which includes at least one second vocabulary and enterprise's mark；So that get second lexical information it
Afterwards, the enterprise first can be found according to the enterprise's mark for including in second lexical information and identifies corresponding exclusive lexicon, then
At least one second vocabulary for including in second lexical information is identified with enterprise to the word in corresponding exclusive lexicon respectively
Remittance is matched, and is all different if the second vocabulary identifies any vocabulary in corresponding exclusive lexicon with enterprise, is illustrated to work as
There is no second vocabulary in preceding exclusive lexicon, the second vocabulary can be added to enterprise at this time and identify corresponding exclusive lexicon
In, to realize the update to the exclusive lexicon.Opposite, if in the second vocabulary exclusive lexicon corresponding with enterprise's mark
Any vocabulary it is identical, then illustrate that in current exclusive lexicon included second vocabulary, can be mentioned at this time to user's output
Show information, which has included the second vocabulary for prompting enterprise to identify in corresponding exclusive lexicon, to prompt to use
Family second vocabulary has stored in exclusive lexicon.Exemplary, which can be for " you are good, and the vocabulary is
It is stored in exclusive lexicon " or first prompt information may be " you are good, please re-enter ".
Fig. 4 is a kind of structural schematic diagram of information processing unit 40 provided in an embodiment of the present invention, the information processing unit
40 can be located at smart machine, exemplary, refer to Fig. 4, which may include:
Acquiring unit 401, the input information received for obtaining smart machine include the intelligence in the input message
Enterprise's mark belonging to energy equipment；And it obtains the enterprise and identifies corresponding exclusive lexicon.
Processing unit 402, for being carried out at identification to the input information according to the vocabulary in the exclusive lexicon
Reason.
Control unit 403, for according to identifying processing as a result, controlling the smart machine executes corresponding operation.
Optionally, the vocabulary in the exclusive lexicon includes that the exclusive word of enterprise and/or the exclusive word of the enterprise are corresponding
Synonym.
Optionally, when the input information is voice messaging, the processing unit 402 is specifically used for passing through automatic language
Sound identification ASR model identifies the input information, obtains the corresponding text information of the input information；If the text
Include target word corresponding with any vocabulary in the exclusive lexicon in information, the target word in the text information is replaced
It is changed to any vocabulary, the text information after obtaining error correction
Described control unit 403, specifically for controlling the smart machine and executing according to the text information after the error correction
Corresponding operation.
Optionally, described control unit 403, specifically for controlling the text information after the smart machine exports error correction.
Optionally, the processing unit 402, if specifically for including in the exclusive lexicon in the input information
Any vocabulary carries out word segmentation processing to the input information according to any vocabulary；Based on after word segmentation processing as a result, right
The input information carries out semantic parsing, obtains semantic parsing result.
Described control unit 403 is specifically used for controlling the smart machine according to the semantic parsing result and executing correspondence
Operation.
Optionally, which can also include establishing unit 404, exemplary, shown in Figure 5, Fig. 5
For the structural schematic diagram of another information processing unit 40 provided in an embodiment of the present invention.
The acquiring unit 401, is also used to obtain the first lexical information of input, and first lexical information includes input
The first vocabulary and enterprise mark.
It is described to establish unit 404, for identifying at least one corresponding first word according to each enterprise mark and the enterprise
It converges, establishes each enterprise and identify corresponding exclusive lexicon.
Optionally, the acquiring unit 401 is also used to obtain the second lexical information of input；Wherein, second vocabulary
Information includes at least one second vocabulary and enterprise's mark.
It is described to establish unit 404, if being also used in second vocabulary exclusive lexicon corresponding with enterprise mark
Any vocabulary be all different, then second vocabulary is added to the enterprise and identified in corresponding exclusive lexicon.
Optionally, described control unit 403, if being also used to second vocabulary exclusive word corresponding with enterprise mark
Any vocabulary in remittance library is identical, then exports prompt information, and the prompt information is corresponding specially for prompting enterprise's mark
Belonging in lexicon has included second vocabulary.
Information processing unit 40 shown in the embodiment of the present invention can execute information processing shown in any of the above-described embodiment
The realization principle and beneficial effect class of the technical solution of method, realization principle and beneficial effect and information processing method
Seemingly, it is no longer repeated herein.
Fig. 6 is the structural schematic diagram of a kind of electronic equipment provided in an embodiment of the present invention, which can be intelligence
The controller of equipment can also be server, not be defined to the specific implementation of electronic equipment in the embodiment of the present invention.Example
, shown in Figure 6, which may include processor 601 and memory 602, wherein
Memory 602 is for storing program instruction.
Processor 601 is used to read the program instruction in memory 602, and is held according to the program instruction in memory 602
Information processing method shown in any of the above-described embodiment of row.
Electronic equipment 60 shown in the embodiment of the present invention can execute information processing method shown in any of the above-described embodiment
Technical solution, realization principle and beneficial effect are similar with the realization principle of information processing method and beneficial effect, this
Place is no longer repeated.
The embodiment of the present invention also provides a kind of computer readable storage medium, is stored on the computer readable storage medium
Computer program can execute information processing side shown in any of the above-described embodiment when computer program is executed by processor
The technical solution of method, realization principle and beneficial effect are similar with the realization principle of information processing method and beneficial effect,
It is no longer repeated herein.
Processor can be general processor, digital signal processor (digital signal in above-described embodiment
Processor, DSP), it is specific integrated circuit (application specific integrated circuit, ASIC), existing
At programmable gate array (field programmable gate array, FPGA) or other programmable logic device, discrete
Door or transistor logic, discrete hardware components.It may be implemented or execute the disclosed each side in the embodiment of the present invention
Method, step and logic diagram.General processor can be microprocessor or the processor is also possible to any conventional processing
Device etc..The step of method in conjunction with disclosed in the embodiment of the present invention, can be embodied directly in hardware decoding processor and execute completion,
Or in decoding processor hardware and software module combination execute completion.Software module can be located at random access memory
(random access memory, RAM), flash memory, may be programmed read-only deposit at read-only memory (read-only memory, ROM)
In the storage medium of this fields such as reservoir or electrically erasable programmable memory, register maturation.The storage medium, which is located at, to be deposited
The step of reservoir, processor reads the instruction in memory, completes the above method in conjunction with its hardware.
In several embodiments provided by the present invention, it should be understood that disclosed device and method can pass through it
Its mode is realized.For example, the apparatus embodiments described above are merely exemplary, for example, the division of the unit, only
Only a kind of logical function partition, there may be another division manner in actual implementation, such as multiple units or components can be tied
Another system is closed or is desirably integrated into, or some features can be ignored or not executed.Another point, it is shown or discussed
Mutual coupling, direct-coupling or communication connection can be through some interfaces, the INDIRECT COUPLING or logical of device or unit
Letter connection can be electrical property, mechanical or other forms.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme
's.In addition, the functional units in various embodiments of the present invention may be integrated into one processing unit, it is also possible to each
Unit physically exists alone, and can also be integrated in one unit with two or more units.Above-mentioned integrated unit both may be used
To use formal implementation of hardware, can also be realized in the form of hardware adds SFU software functional unit.
Those skilled in the art after considering the specification and implementing the invention disclosed here, will readily occur to its of the disclosure
Its embodiment.The present invention is directed to cover any variations, uses, or adaptations of the disclosure, these modifications, purposes or
Person's adaptive change follows the general principles of this disclosure and including the undocumented common knowledge in the art of the disclosure
Or conventional techniques.The description and examples are only to be considered as illustrative, and the true scope and spirit of the disclosure are by following
Claims are pointed out.
It should be understood that the present disclosure is not limited to the precise structures that have been described above and shown in the drawings, and
And various modifications and changes may be made without departing from the scope thereof.The scope of the present disclosure is only limited by appended claims
System.
Claims (10)
1. a kind of information processing method characterized by comprising
The input information that smart machine receives is obtained, includes the mark of enterprise belonging to the smart machine in the input message
Know；
It obtains the enterprise and identifies corresponding exclusive lexicon；
According to the vocabulary in the exclusive lexicon, identifying processing is carried out to the input information；
According to identifying processing as a result, controlling the smart machine executes corresponding operation.
2. the method according to claim 1, wherein the vocabulary in the exclusive lexicon includes the exclusive word of enterprise
And/or the corresponding synonym of the exclusive word of enterprise.
3. the method according to claim 1, wherein when the input information be voice messaging when, the basis
The vocabulary carries out identifying processing to the input information, comprising:
The input information is identified by automatic speech recognition ASR model, obtains the corresponding text of the input information
Information；
If including target word corresponding with any vocabulary in the exclusive lexicon in the text information, by the text envelope
Target word in breath replaces with any vocabulary, the text information after obtaining error correction；
It is described according to identifying processing as a result, controlling the smart machine executes corresponding operation, comprising:
According to the text information after the error correction, controls the smart machine and execute corresponding operation.
4. according to the method described in claim 3, it is characterized in that, the text information according to after the error correction, controls institute
It states smart machine and executes corresponding operation, comprising:
Text information after controlling the smart machine output error correction.
5. method according to claim 1-4, which is characterized in that it is described according to the vocabulary, to the input
Information carries out identifying processing, comprising:
If comprising any vocabulary in the exclusive lexicon in the input information, according to any vocabulary, to described defeated
Enter information and carries out word segmentation processing；
Based on after word segmentation processing as a result, obtaining semantic parsing result to the semantic parsing of input information progress；
It is described according to identifying processing as a result, controlling the smart machine executes corresponding operation, comprising:
According to the semantic parsing result, controls the smart machine and execute corresponding operation.
6. method according to claim 1-4, which is characterized in that the method also includes:
The first lexical information of input is obtained, first lexical information includes the first vocabulary and enterprise's mark of input；
At least one corresponding first vocabulary is identified according to each enterprise mark and the enterprise, it is corresponding to establish each enterprise's mark
Exclusive lexicon.
7. according to the method described in claim 6, it is characterized in that, establish each enterprise identify corresponding exclusive lexicon it
Afterwards, the method also includes:
Obtain the second lexical information of input；Wherein, second lexical information includes at least one second vocabulary and enterprise's mark
Know；
It, will be described if second vocabulary identifies any vocabulary in corresponding exclusive lexicon with the enterprise and is all different
Second vocabulary is added to the enterprise and identifies in corresponding exclusive lexicon.
8. the method according to the description of claim 7 is characterized in that the method also includes:
If it is identical that second vocabulary and the enterprise identify any vocabulary in corresponding exclusive lexicon, prompt letter is exported
Breath, the prompt information have included second vocabulary for prompting the enterprise to identify in corresponding exclusive lexicon.
9. a kind of information processing unit characterized by comprising
Acquiring unit, the input information received for obtaining smart machine include the smart machine in the input message
Affiliated enterprise's mark；And it obtains the enterprise and identifies corresponding exclusive lexicon；
Processing unit, for carrying out identifying processing to the input information according to the vocabulary in the exclusive lexicon；
Control unit, for according to identifying processing as a result, controlling the smart machine executes corresponding operation.
10. a kind of electronic equipment, which is characterized in that including processor and memory, wherein
The memory is for storing program instruction；
The processor is used to read the program instruction in the memory, and is executed according to the program instruction in the memory
The described in any item information processing methods of claim 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811627054.5A CN109710745A (en) | 2018-12-28 | 2018-12-28 | Information processing method, device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811627054.5A CN109710745A (en) | 2018-12-28 | 2018-12-28 | Information processing method, device and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109710745A true CN109710745A (en) | 2019-05-03 |
Family
ID=66259274
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811627054.5A Pending CN109710745A (en) | 2018-12-28 | 2018-12-28 | Information processing method, device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109710745A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112215010A (en) * | 2019-07-10 | 2021-01-12 | 北京猎户星空科技有限公司 | Semantic recognition method and equipment |
| CN112232061A (en) * | 2019-06-28 | 2021-01-15 | 傲基科技股份有限公司 | Content processing method, electronic device, and computer-readable storage medium |
| CN112256947A (en) * | 2019-07-05 | 2021-01-22 | 北京猎户星空科技有限公司 | Method, device, system, equipment and medium for determining recommendation information |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103455535A (en) * | 2013-05-08 | 2013-12-18 | 深圳市明唐通信有限公司 | Method for establishing knowledge base based on historical consultation data |
| US8972445B2 (en) * | 2009-04-23 | 2015-03-03 | Deep Sky Concepts, Inc. | Systems and methods for storage of declarative knowledge accessible by natural language in a computer capable of appropriately responding |
| CN104598579A (en) * | 2015-01-14 | 2015-05-06 | 北京京东尚科信息技术有限公司 | Automatic question and answer method and system |
| CN104850539A (en) * | 2015-05-28 | 2015-08-19 | 宁波薄言信息技术有限公司 | Natural language understanding method and travel question-answering system based on same |
| CN105528437A (en) * | 2015-12-17 | 2016-04-27 | 浙江大学 | Question-answering system construction method based on structured text knowledge extraction |
-
2018
- 2018-12-28 CN CN201811627054.5A patent/CN109710745A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8972445B2 (en) * | 2009-04-23 | 2015-03-03 | Deep Sky Concepts, Inc. | Systems and methods for storage of declarative knowledge accessible by natural language in a computer capable of appropriately responding |
| CN103455535A (en) * | 2013-05-08 | 2013-12-18 | 深圳市明唐通信有限公司 | Method for establishing knowledge base based on historical consultation data |
| CN104598579A (en) * | 2015-01-14 | 2015-05-06 | 北京京东尚科信息技术有限公司 | Automatic question and answer method and system |
| CN104850539A (en) * | 2015-05-28 | 2015-08-19 | 宁波薄言信息技术有限公司 | Natural language understanding method and travel question-answering system based on same |
| CN105528437A (en) * | 2015-12-17 | 2016-04-27 | 浙江大学 | Question-answering system construction method based on structured text knowledge extraction |
Non-Patent Citations (1)
| Title |
|---|
| 袁津生等: "《搜索引擎原理与实践》", 31 December 2008 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112232061A (en) * | 2019-06-28 | 2021-01-15 | 傲基科技股份有限公司 | Content processing method, electronic device, and computer-readable storage medium |
| CN112256947A (en) * | 2019-07-05 | 2021-01-22 | 北京猎户星空科技有限公司 | Method, device, system, equipment and medium for determining recommendation information |
| CN112256947B (en) * | 2019-07-05 | 2024-01-26 | 北京猎户星空科技有限公司 | Recommendation information determining method, device, system, equipment and medium |
| CN112215010A (en) * | 2019-07-10 | 2021-01-12 | 北京猎户星空科技有限公司 | Semantic recognition method and equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Weisz et al. | Sample efficient deep reinforcement learning for dialogue systems with large action spaces | |
| CN102282609B (en) | System and method for recognizing proper names in dialog systems | |
| CN107644638B (en) | Audio recognition method, device, terminal and computer readable storage medium | |
| JP5695199B2 (en) | Thought tracking and action selection in dialogue systems | |
| CN109710745A (en) | Information processing method, device and electronic equipment | |
| CN109961780A (en) | A kind of man-machine interaction method, device, server and storage medium | |
| US10847151B2 (en) | Dialogue system and dialogue method | |
| CN110336723A (en) | Control method and device, the intelligent appliance equipment of intelligent appliance | |
| JP2018132754A (en) | Dialog system, dialog method, and method for adapting dialog system | |
| CN110347863A (en) | Talk about art recommended method and device and storage medium | |
| Wang et al. | Learning domain-independent dialogue policies via ontology parameterisation | |
| CN110955675A (en) | Robot dialogue method, device, equipment and computer readable storage medium | |
| CN110288995A (en) | Exchange method, device, storage medium and electronic equipment based on speech recognition | |
| JP2018025747A (en) | Dialogue action estimation method, dialogue action estimation device and program | |
| CN111178081B (en) | Semantic recognition method, server, electronic device and computer storage medium | |
| EP3422344A1 (en) | Electronic device for performing operation corresponding to voice input | |
| CN110489519A (en) | The session method and Related product of dialogue-based prediction model | |
| CN113821620A (en) | Multi-turn conversation task processing method and device and electronic equipment | |
| CN111400479B (en) | Method and device for identifying problems aiming at multi-round dialogue | |
| KR20200063886A (en) | Natural language processing based call center support system and method | |
| CN115083412B (en) | Voice interaction method and related device, electronic equipment and storage medium | |
| CN115658875B (en) | Data processing method based on chat service and related products | |
| CN112765981A (en) | Text information generation method and device | |
| CN116913274A (en) | Scene generation method, device and storage medium based on generation type large model | |
| CN110489517B (en) | Automatic learning method and system of virtual assistant |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190503 |