CN109726266A - Note signature processing method, equipment and computer readable storage medium - Google Patents
Note signature processing method, equipment and computer readable storage medium Download PDFInfo
- Publication number
- CN109726266A CN109726266A CN201811574935.5A CN201811574935A CN109726266A CN 109726266 A CN109726266 A CN 109726266A CN 201811574935 A CN201811574935 A CN 201811574935A CN 109726266 A CN109726266 A CN 109726266A
- Authority
- CN
- China
- Prior art keywords
- enterprise
- name
- information
- prediction result
- black
- 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
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses the signing messages that a kind of note signature processing method, equipment and computer readable storage medium, the method pass through acquisition short message；The corresponding black and white lists prediction result of the signing messages is obtained by the black and white lists network model pre-established according to the signing messages, it is tentatively whether effective to signing messages is distinguished；When the black and white lists prediction result is white list, the corresponding enterprise's name prediction result of the signing messages is obtained by the enterprise's name network model pre-established according to the signing messages；Enterprise name prediction result is successively matched with the enterprise information for prestoring in the database, to obtain from the database and the matched enterprise's name information of enterprise's name prediction result and its corresponding trade classification information, go out enterprise's alias, trade classification by note signature information excavating, improves the utilization rate of note signature information.
Description
Technical field
The present invention relates to short breath process field more particularly to a kind of note signature processing methods, equipment and computer-readable
Storage medium.
Background technique
Short message erects effective channel as the carrier for transmitting information between magnanimity client, for communication to each other.Along with
Short message is widely used, and how more efficiently to utilize, excavates short breath data, becomes the research neck popular for a comparison
Domain.
Currently, for enterprise-class short message, it generally can be in the beginning of short message content or ending often with some enterprise's names
The identification information about the enterprise such as title, i.e., so-called note signature information.But since the note signature data of enterprise-class are deposited
The features such as information content is fewer, signature contents itself are more chaotic, cause signing messages utilization rate not high.
Summary of the invention
It can in view of the above-mentioned problems, the purpose of the present invention is to provide a kind of note signature processing method, equipment and computers
Storage medium is read, enterprise's alias, trade classification can be identified by note signature information, improves the benefit of note signature information
With rate.
In a first aspect, the embodiment of the invention provides a kind of note signature processing methods, comprising the following steps:
Obtain the signing messages of short message；
It is corresponding to be obtained by the black and white lists network model pre-established for the signing messages according to the signing messages
Black and white lists prediction result；
When the black and white lists prediction result is white list, according to the signing messages, pass through the enterprise pre-established
Name network model obtains the corresponding enterprise's name prediction result of the signing messages；
Enterprise name prediction result is successively matched with the enterprise information for prestoring in the database, with from described
It is obtained in database and the matched enterprise's name information of enterprise's name prediction result and its corresponding trade classification information.
Preferably, described successively to carry out enterprise name prediction result with the enterprise's name information for prestoring in the database
Match, to obtain from the database and the matched enterprise's name information of enterprise's name prediction result and its corresponding trade classification
Information specifically includes:
Judge whether enterprise name prediction result matches with the enterprise information for prestoring in the database；
If so, obtaining from the database and the matched enterprise information of enterprise name prediction result and its corresponding
Trade classification information；
If it is not, enterprise's name prediction result and the enterprise's name information prestored in the database are successively carried out similarity meter
It calculates, from the database acquisition and the highest enterprise's name information of enterprise's name prediction result similarity, and according to the label
Name data obtain trade classification information by the trade classification network model pre-established.
Preferably, the method also includes:
When enterprise's name prediction result is less than preset threshold with the similarity for prestoring enterprise's name information in the database
When value, the signing messages is ended processing.
Preferably, the method also includes following black and white lists network model construction steps:
Dictionary is constructed to signature sample gathered in advance；Wherein, the signature sample include original signature information, with it is described
The corresponding black and white lists classification information of original signature information, enterprise's name information, trade classification coding；The dictionary includes the original
Mapping relations in beginning signing messages between each word and number；
Black and white lists classified dictionary is constructed to signature sample gathered in advance；Wherein, the black and white lists classified dictionary packet
Include the mapping relations between the corresponding black and white lists classification information of the original signature information and number；
According to the dictionary, the black and white lists classified dictionary, by the original signature information and its corresponding black and white name
Single classification information is converted to Serial No., as the first Serial No.；
First Serial No. is trained by convolutional neural networks model, establishes the black and white lists network mould
Type.
Preferably, described that the label are obtained by the black and white lists network model pre-established according to the signing messages
The corresponding black and white lists prediction result of name information, specifically includes:
According to the dictionary, the signing messages is converted into Serial No., as the second Serial No.；
Using second Serial No. as the input value of the black and white lists network model, to obtain the signing messages
Corresponding black and white lists prediction result.
Preferably, the method also includes following enterprise's name network model construction steps:
To signature sample construction enterprise name dictionary；Wherein, enterprise's name dictionary includes the original signature information
Mapping relations between corresponding enterprise's name information and number；
According to the dictionary, enterprise name dictionary, the original signature information and its corresponding enterprise name information are turned
It is changed to Serial No., as third Serial No.；
The third Serial No. is trained by neural translation model, establishes enterprise's name network model.
Preferably, described that enterprise name prediction result is successively subjected to phase with the enterprise information for prestoring in the database
It is calculated like degree, from the database acquisition and the highest enterprise's name information of enterprise's name prediction result similarity, and according to
The signed data is obtained trade classification information, is specifically included by the trade classification network model pre-established:
Each word in enterprise name prediction result, enterprise name information is converted into word vector respectively, and according to
The frequency of each word calculates enterprise name prediction result, the corresponding word weight of each word in enterprise name information；
According to the word vector of enterprise's name prediction result and its word vector of corresponding word weight, enterprise name information
And its corresponding word weight, sentence vector, the enterprise name information of enterprise's name prediction result are calculated separately using SIF algorithm
Sentence vector；
The sentence vector of sentence vector and enterprise name information to enterprise's name prediction result carries out cosine similarity meter
It calculates, from the database acquisition and the highest enterprise's name information of enterprise's name prediction result cosine similarity, and according to institute
Signed data is stated, by the trade classification network model pre-established, obtains trade classification information.
Preferably, the method also includes following trade classification network model construction steps:
Trade classification dictionary is constructed to signature sample gathered in advance；The trade classification dictionary includes the trade classification
Mapping relations between coding and number；
According to the dictionary, the trade classification dictionary, the original signature information and its corresponding trade classification are compiled
Code is converted to Serial No., as the 4th Serial No.；
The 4th Serial No. is trained by convolutional neural networks model, establishes the trade classification network mould
Type.
Second aspect, the embodiment of the invention provides a kind of note signature processing equipment, including processor, memory and
The computer program executed by the processor is stored in the memory and is configured as, the processor executes the meter
The note signature processing method as described in any one of first aspect is realized when calculation machine program.
The third aspect, the embodiment of the invention provides a kind of computer readable storage medium, the computer-readable storage
Medium includes the computer program of storage, wherein controls the computer-readable storage medium in computer program operation
Equipment executes the note signature processing method as described in any one of first aspect where matter.
Above embodiments have the following beneficial effects:
By the signing messages for obtaining short message；According to the signing messages, pass through the black and white lists network mould pre-established
Whether type obtains the corresponding black and white lists prediction result of the signing messages, tentatively effective to signing messages is distinguished；When described
When black and white lists prediction result is white list, obtained according to the signing messages by the enterprise's name network model pre-established
The corresponding enterprise's name prediction result of the signing messages；By enterprise name prediction result with prestore enterprise's name in the database
Information is successively matched, with obtain from the database with the matched enterprise's name information of enterprise name prediction result and its
Corresponding trade classification information identifies enterprise's alias, trade classification by note signature information, improves note signature information
Utilization rate.
Detailed description of the invention
In order to illustrate more clearly of technical solution of the present invention, attached drawing needed in embodiment will be made below
Simply introduce, it should be apparent that, the accompanying drawings in the following description is only some embodiments of the present invention, general for this field
For logical technical staff, without creative efforts, it is also possible to obtain other drawings based on these drawings.
Fig. 1 is the flow diagram for the note signature processing method that first embodiment of the invention provides.
Fig. 2 is the structural schematic diagram for the note signature processing equipment that second embodiment of the invention provides.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
Fig. 1 is please referred to, first embodiment of the invention provides a kind of note signature processing method, can be by note signature
Equipment is managed to execute, and the following steps are included:
S11 obtains the signing messages of short message；
In embodiments of the present invention, the note signature processing equipment can be computer, mobile phone, tablet computer, notebook electricity
Brain or server etc. calculate equipment, the note signature processing method can be used as one of functional module it is integrated with it is described short
In letter signature processing equipment, executed by the note signature processing equipment.
S12 obtains the signing messages by the black and white lists network model pre-established according to the signing messages
Corresponding black and white lists prediction result；
Since the signing messages of short message is disorderly and unsystematic, by the black and white lists network model that pre-establishes to the signature of short message
Information carries out black and white lists classification prediction, distinguishes effective signing messages and invalid signature information, effective signing messages is pre-
It surveys and is predicted to be blacklist for white list, invalid signing messages.When the black and white lists prediction result is blacklist, terminate
Handle signing messages.
S13, according to the signing messages, passes through what is pre-established when the black and white lists prediction result is white list
Enterprise's name network model obtains the corresponding enterprise's name prediction result of the signing messages；
S14 successively matches enterprise name prediction result with the enterprise information for prestoring in the database, with from
It is obtained in the database and the matched enterprise's name information of enterprise's name prediction result and its corresponding trade classification information.
The embodiment of the present invention is few to information content, chaotic note signature information carries out depth excavation, identify enterprise's alias,
Trade classification information improves the utilization rate of note signature information.
In an alternative embodiment, it is described by enterprise name prediction result with prestore enterprise's name in the database
Information is successively matched, with obtain from the database with the matched enterprise's name information of enterprise name prediction result and its
Corresponding trade classification information, specifically includes:
Judge whether enterprise name prediction result matches with the enterprise information for prestoring in the database；
If so, obtaining from the database and the matched enterprise information of enterprise name prediction result and its corresponding
Trade classification information；
If it is not, enterprise's name prediction result and the enterprise's name information prestored in the database are successively carried out similarity meter
It calculates, from the database acquisition and the highest enterprise's name information of enterprise's name prediction result similarity, and according to the label
Name data obtain trade classification information by the trade classification network model pre-established.
In embodiments of the present invention, it for enterprise's name prediction result, needs and enterprise's name information pair existing in database
Than；If identical as enterprise's name information in database, enterprise's name information and its associated industry point in output database
Category information；If different from enterprise's name information in database, similarity calculation is carried out with enterprise's name information in database, it is defeated
Out with enterprise's name prediction result similarity highest enterprise name information and according to the signed data in database, by preparatory
The trade classification network model of foundation exports trade classification information.The embodiment of the present invention can simplify the data mining of signing messages
Process improves the treatment effeciency of signing messages.
In an alternative embodiment, the method also includes:
When enterprise's name prediction result is less than preset threshold with the similarity for prestoring enterprise's name information in the database
When value, the signing messages is ended processing.
In an alternative embodiment, the method also includes following black and white lists network model construction steps:
Dictionary is constructed to signature sample gathered in advance；Wherein, the signature sample include original signature information, with it is described
The corresponding black and white lists classification information of original signature information, enterprise's name information, trade classification coding；The dictionary includes the original
Mapping relations in beginning signing messages between each word and number；
Black and white lists classified dictionary is constructed to signature sample gathered in advance；Wherein, the black and white lists classified dictionary packet
Include the mapping relations between the corresponding black and white lists classification information of the original signature information and number；
According to the dictionary, the black and white lists classified dictionary, by the original signature information and its corresponding black and white name
Single classification information is converted to Serial No., as the first Serial No.；
First Serial No. is trained by convolutional neural networks model, establishes the black and white lists network mould
Type.
In embodiments of the present invention, for example, signature sample for acquisition, " you " is mapped as 101 in the dictionary,
" good " is mapped as 205；White list is mapped as number 0 by the black and white lists classified dictionary, and blacklist is mapped as number 1, according to
The dictionary, the black and white lists classified dictionary turn the original signature information and its corresponding black and white lists classification information
It is changed to Serial No., and feeds convolutional neural networks (CNN) model and carries out model training, later to unknown signing messages, together
Sample is converted to Serial No. according to the dictionary, the black and white lists classified dictionary, and feeds the black and white lists network after training
Model obtains the prediction result that it is blacklist or white list, carries out black and white lists prediction by black and white lists network model
The identification accuracy that black and white lists can be greatlyd improve, avoids effective signing messages misjudged.
In an alternative embodiment, described according to the signing messages, pass through the black and white lists network pre-established
Model obtains the corresponding black and white lists prediction result of the signing messages, specifically includes:
According to the dictionary, the signing messages is converted into Serial No., as the second Serial No.；
Using second Serial No. as the input value of the black and white lists network model, to obtain the signing messages
Corresponding black and white lists prediction result.
In an alternative embodiment, the method also includes following enterprise's name network model construction steps:
To signature sample construction enterprise name dictionary；Wherein, enterprise's name dictionary includes the original signature information
Mapping relations between corresponding enterprise's name information and number；
According to the dictionary, enterprise name dictionary, the original signature information and its corresponding enterprise name information are turned
It is changed to Serial No., as third Serial No.；
The third Serial No. is trained by neural translation model, establishes enterprise's name network model.
In embodiments of the present invention, for example, signature sample for acquisition, " length " is mapped as in enterprise's name dictionary
0, " river " is mapped as 1, according to the dictionary, enterprise name dictionary, by the original signature information and its corresponding enterprise name
Information is converted to Serial No., and feeds nerve translation (NMT) model and carry out model training, later to unknown signing messages,
Also according to the dictionary, enterprise name dictionary, Serial No. is converted to, and feeds enterprise's name network model after training,
Its enterprise name prediction result is obtained, the corresponding enterprise's alias of signing messages can be recognized accurately by enterprise's name network model,
The content of depth excavation signing messages.
In an alternative embodiment, it is described by enterprise name prediction result with prestore enterprise's name in the database
Information successively carries out similarity calculation, obtains from the database and the highest enterprise of enterprise's name prediction result similarity
Name information, and trade classification information, tool are obtained by the trade classification network model pre-established according to the signed data
Body includes:
Each word in enterprise name prediction result, enterprise name information is converted into word vector respectively, and according to
The frequency of each word calculates enterprise name prediction result, the corresponding word weight of each word in enterprise name information；
According to the word vector of enterprise's name prediction result and its word vector of corresponding word weight, enterprise name information
And its corresponding word weight, sentence vector, the enterprise name information of enterprise's name prediction result are calculated separately using SIF algorithm
Sentence vector；
The sentence vector of sentence vector and enterprise name information to enterprise's name prediction result carries out cosine similarity meter
It calculates, from the database acquisition and the highest enterprise's name information of enterprise's name prediction result cosine similarity, and according to institute
Signed data is stated, by the trade classification network model pre-established, obtains trade classification information.
In embodiments of the present invention, first enterprise's name prediction result, enterprise name information are segmented, Zhi Houtong
Cross the CBOW algorithm in word2vec to after segmenting the enterprise name prediction result, the enterprise name information in each word turn
Be changed to word vector, and calculate word weight according to the frequency that each word occurs, by SIF algorithm to enterprise's name prediction result,
Each word vector does weighted average to obtain enterprise's name prediction result, enterprise name information in enterprise's name information
The sentence vector of entire sentence calculates the similarity between sentence vector by cosine similarity, and in output database with enterprise's name
Prediction result similarity highest enterprise name information and according to the signed data, passes through the trade classification network pre-established
Model exports trade classification information.
Calculating for sentence vector, is exemplified below:
(1) it is based on enterprise's name database library, the word frequency is counted by formula a/ (a+v/N)；Wherein a is fixed constant, example
If a takes 1-3 any number, v is word frequency, and N is total word frequency of all words；Reflecting for each word and its corresponding word weight is calculated
Firing table.
(2) enterprise's name information in enterprise's name prediction result and database is added together and is put into mapping table, pass through word
Enterprise's name prediction result, enterprise name information are converted to sequence vector respectively by vector；Meanwhile it being reflected by word and weight
Enterprise's name prediction result, enterprise name information are converted to weight sequence respectively by firing table.At this point, each enterprise's name is deposited
In two sequences.Such as: the sequence vector of " XY " this enterprise isWord weight sequence (0.1 0.2)；
So word weight matrix has just obtained one (0.05 0.04 0.05) multiplied by vector matrix, then divided by enterprise name length,
Obtain " XY " this enterprise weighted average sequence (0.025 0.02 0.025).
(3) after calculating the sequence of all enterprise's name information, the matrix A of a N*3 can be obtained by putting together；Then
Singular value matrix B is calculated, final Matrix C is A-B；The sentence vector of each enterprise is just certain a line in Matrix C.
In an alternative embodiment, the method also includes following trade classification network model construction steps:
Trade classification dictionary is constructed to signature sample gathered in advance；The trade classification dictionary includes the trade classification
Mapping relations between coding and number；
According to the dictionary, the trade classification dictionary, the original signature information and its corresponding trade classification are compiled
Code is converted to Serial No., as the 4th Serial No.；
The 4th Serial No. is trained by convolutional neural networks model, establishes the trade classification network mould
Type.
Compared with the existing technology, the beneficial effect that the present invention is implemented is:
1, relative to it is traditional based on the rule pre-established (such as in signing messages include certain keywords it is determined that
Blacklist) black and white lists classification is carried out, the embodiment of the present invention is based on CNN model and carries out black and white lists classification, greatly improves
The accuracy rate and coverage rate of black and white lists identification；
2, traditional to must manually perform and handle one by one by signature extraction enterprise's name and the corresponding trade classification of association
, and the embodiment of the present invention is based on CNN model and NMT model and carries out automatic processing, artificial only needs pair to signing messages
The enterprise's name information finally predicted and its corresponding trade classification information carry out review operations, substantially increase the place of signing messages
Efficiency is managed, while note signature information few to information content, chaotic carries out depth excavation, identifies enterprise's alias, trade classification
Information improves the utilization rate of note signature information.
It referring to fig. 2, is the schematic diagram for the note signature processing equipment that second embodiment of the invention provides.As shown in Fig. 2, should
Note signature processing equipment includes: at least one processor 11, such as CPU, at least one network interface 14 or other users
Interface 13, memory 15, at least one communication bus 12, communication bus 12 is for realizing the connection communication between these components.
Wherein, user interface 13 optionally may include USB interface and other standards interface, wireline interface.Network interface 14 is optional
May include Wi-Fi interface and other wireless interfaces.Memory 15 may include high speed RAM memory, it is also possible to also wrap
It includes non-labile memory (non-volatilememory), for example, at least a magnetic disk storage.Memory 15 is optional
It may include at least one storage device for being located remotely from aforementioned processor 11.
In some embodiments, memory 15 stores following element, executable modules or data structures, or
Their subset or their superset:
Operating system 151 includes various system programs, for realizing various basic businesses and hardware based of processing
Business；
Program 152.
Specifically, processor 11 executes short described in above-described embodiment for calling the program 152 stored in memory 15
Letter signature processing method, such as step S11 shown in FIG. 1.Alternatively, the processor is realized when executing the computer program
State the function of each module/unit in each Installation practice.
Illustratively, the computer program can be divided into one or more module/units, one or more
A module/unit is stored in the memory, and is executed by the processor, to complete the present invention.It is one or more
A module/unit can be the series of computation machine program instruction section that can complete specific function, and the instruction segment is for describing institute
State implementation procedure of the computer program in the note signature processing equipment.
The note signature processing equipment can be desktop PC, notebook, palm PC and cloud server etc.
Calculate equipment.The note signature processing equipment may include, but be not limited only to, processor, memory.Those skilled in the art can
To understand, the schematic diagram is only the example of note signature processing equipment, does not constitute the limit to note signature processing equipment
It is fixed, it may include perhaps combining certain components or different components than illustrating more or fewer components.
Alleged processor 11 can be central processing unit (Central Processing Unit, CPU), can also be
Other general processors, digital signal processor (Digital Signal Processor, DSP), specific integrated circuit
(Application Specific Integrated Circuit, ASIC), ready-made programmable gate array (Field-
Programmable Gate Array, FPGA) either other programmable logic device, discrete gate or transistor logic,
Discrete hardware components etc..General processor can be microprocessor or the processor is also possible to any conventional processor
Deng the processor 11 is the control centre of the note signature processing equipment, entirely short using various interfaces and connection
The various pieces of letter signature processing equipment.
The memory 15 can be used for storing the computer program and/or module, the processor 11 by operation or
Computer program and/or the module stored in the memory is executed, and calls the data being stored in memory, is realized
The various functions of the note signature processing equipment.The memory 15 can mainly include storing program area and storage data area,
Wherein, storing program area can application program needed for storage program area, at least one function (such as sound-playing function, figure
As playing function etc.) etc.；Storage data area, which can be stored, uses created data (such as audio data, phone according to mobile phone
This etc.) etc..In addition, memory 15 may include high-speed random access memory, it can also include nonvolatile memory, such as
Hard disk, memory, plug-in type hard disk, intelligent memory card (Smart Media Card, SMC), secure digital (Secure
Digital, SD) card, flash card (Flash Card), at least one disk memory, flush memory device or other volatibility are solid
State memory device.
Wherein, if module/unit that the note signature processing equipment integrates is realized in the form of SFU software functional unit
And when sold or used as an independent product, it can store in a computer readable storage medium.Based on such
Understand, the present invention realizes all or part of the process in above-described embodiment method, can also instruct phase by computer program
The hardware of pass is completed, and the computer program can be stored in a computer readable storage medium, which exists
When being executed by processor, it can be achieved that the step of above-mentioned each embodiment of the method.Wherein, the computer program includes computer journey
Sequence code, the computer program code can be source code form, object identification code form, executable file or certain intermediate shapes
Formula etc..The computer-readable medium may include: any entity or device, note that can carry the computer program code
Recording medium, USB flash disk, mobile hard disk, magnetic disk, CD, computer storage, read-only memory (ROM, Read-Only Memory),
Random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium
Deng.It should be noted that the content that the computer-readable medium includes can be real according to legislation in jurisdiction and patent
The requirement trampled carries out increase and decrease appropriate, such as in certain jurisdictions, according to legislation and patent practice, computer-readable medium
It does not include electric carrier signal and telecommunication signal.
Third embodiment of the invention provides a kind of computer readable storage medium, the computer readable storage medium packet
Include the computer program of storage, wherein where controlling the computer readable storage medium in computer program operation
Equipment executes note signature processing method as in the first embodiment.
The above is a preferred embodiment of the present invention, it is noted that for those skilled in the art
For, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications are also considered as
Protection scope of the present invention.
Claims (10)
1. a kind of note signature processing method characterized by comprising
Obtain the signing messages of short message；
It is corresponding black to be obtained by the black and white lists network model pre-established for the signing messages according to the signing messages
White list prediction result；
When the black and white lists prediction result is white list, according to the signing messages, pass through the enterprise's name net pre-established
Network model obtains the corresponding enterprise's name prediction result of the signing messages；
Enterprise name prediction result is successively matched with the enterprise information for prestoring in the database, with from the data
It is obtained in library and the matched enterprise's name information of enterprise's name prediction result and its corresponding trade classification information.
2. note signature processing method as described in claim 1, which is characterized in that it is described by enterprise name prediction result with
The enterprise's name information prestored in the database is successively matched, and is tied with obtaining from the database with enterprise's name prediction
The matched enterprise's name information of fruit and its corresponding trade classification information, specifically include:
Judge whether enterprise name prediction result matches with the enterprise information for prestoring in the database；
If so, obtaining from the database and the matched enterprise's name information of enterprise's name prediction result and its corresponding industry
Classification information；
If it is not, enterprise's name prediction result and the enterprise's name information prestored in the database are successively subjected to similarity calculation,
Acquisition and the highest enterprise's name information of enterprise's name prediction result similarity from the database, and according to the number of signature
According to, pass through the trade classification network model pre-established, obtain trade classification information.
3. note signature processing method as claimed in claim 2, which is characterized in that the method also includes:
When enterprise's name prediction result is less than preset threshold value with the similarity for prestoring enterprise's name information in the database,
End processing the signing messages.
4. note signature processing method as described in claim 1, which is characterized in that the method also includes following black and white lists
Network model construction step:
Dictionary is constructed to signature sample gathered in advance；Wherein, the signature sample include original signature information, with it is described original
The corresponding black and white lists classification information of signing messages, enterprise's name information, trade classification coding；The dictionary includes the original label
Mapping relations in name information between each word and number；
Black and white lists classified dictionary is constructed to signature sample gathered in advance；Wherein, the black and white lists classified dictionary includes institute
State the mapping relations between the corresponding black and white lists classification information of original signature information and number；
According to the dictionary, the black and white lists classified dictionary, by the original signature information and its corresponding black and white lists point
Category information is converted to Serial No., as the first Serial No.；
First Serial No. is trained by convolutional neural networks model, establishes the black and white lists network model.
5. note signature processing method as claimed in claim 4, which is characterized in that it is described according to the signing messages, pass through
The black and white lists network model pre-established obtains the corresponding black and white lists prediction result of the signing messages, specifically includes:
According to the dictionary, the signing messages is converted into Serial No., as the second Serial No.；
It is corresponding to obtain the signing messages using second Serial No. as the input value of the black and white lists network model
Black and white lists prediction result.
6. note signature processing method as claimed in claim 4, which is characterized in that the method also includes following enterprise's name nets
Network model construction step:
To signature sample construction enterprise name dictionary；Wherein, enterprise's name dictionary includes that the original signature information is corresponding
Enterprise name information and number between mapping relations；
According to the dictionary, enterprise name dictionary, the original signature information and its corresponding enterprise name information are converted to
Serial No., as third Serial No.；
The third Serial No. is trained by neural translation model, establishes enterprise's name network model.
7. note signature processing method as claimed in claim 2, which is characterized in that it is described by enterprise name prediction result with
The enterprise's name information prestored in the database successively carries out similarity calculation, obtains from the database pre- with enterprise's name
The highest enterprise's name information of result similarity is surveyed, and according to the signed data, passes through the trade classification network mould pre-established
Type obtains trade classification information, specifically includes:
Each word in enterprise's name prediction result, enterprise name information is converted into word vector respectively, and according to each
The frequency of word calculates enterprise name prediction result, the corresponding word weight of each word in enterprise name information；
According to the word vector of enterprise name prediction result and its corresponding word weight, enterprise name information word vector and its
Corresponding word weight calculates separately the sentence vector of enterprise's name prediction result, the sentence of enterprise name information using SIF algorithm
Vector；
The sentence vector of sentence vector and enterprise name information to enterprise's name prediction result carries out cosine similarity calculating, from
Acquisition and the highest enterprise's name information of enterprise's name prediction result cosine similarity in the database, and according to the signature
Data obtain trade classification information by the trade classification network model pre-established.
8. note signature processing method as claimed in claim 4, which is characterized in that the method also includes following trade classifications
Network model construction step:
Trade classification dictionary is constructed to signature sample gathered in advance；The trade classification dictionary includes the trade classification coding
Mapping relations between number；
According to the dictionary, the trade classification dictionary, the original signature information and its corresponding trade classification coding are turned
It is changed to Serial No., as the 4th Serial No.；
The 4th Serial No. is trained by convolutional neural networks model, establishes the trade classification network model.
9. a kind of note signature processing equipment, including processor, memory and storage in the memory and are configured as
The computer program executed by the processor, the processor realize such as claim 1 to 8 when executing the computer program
Any one of described in note signature processing method.
10. a kind of computer readable storage medium, which is characterized in that the computer readable storage medium includes the calculating of storage
Machine program, wherein equipment where controlling the computer readable storage medium in computer program operation is executed as weighed
Benefit require any one of 1 to 8 described in note signature processing method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811574935.5A CN109726266A (en) | 2018-12-21 | 2018-12-21 | Note signature processing method, equipment and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811574935.5A CN109726266A (en) | 2018-12-21 | 2018-12-21 | Note signature processing method, equipment and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109726266A true CN109726266A (en) | 2019-05-07 |
Family
ID=66297105
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811574935.5A Pending CN109726266A (en) | 2018-12-21 | 2018-12-21 | Note signature processing method, equipment and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109726266A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111966806A (en) * | 2020-08-17 | 2020-11-20 | 深圳市活力天汇科技股份有限公司 | Information extraction and repair method for hotel reservation short message |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101030856A (en) * | 2006-07-19 | 2007-09-05 | 王李琰 | Method for verifying SMS and transmitting reliability classification based on cipher technology mark |
| US20090234925A1 (en) * | 2008-03-14 | 2009-09-17 | International Business Machines Corporation | Dyanmic Domain Based Electronic Mail Signature Lines |
| CN104301532A (en) * | 2014-09-30 | 2015-01-21 | 小米科技有限责任公司 | Communication message identification method and device |
| CN106779467A (en) * | 2016-12-31 | 2017-05-31 | 成都数联铭品科技有限公司 | Enterprises ' industry categorizing system based on automatic information screening |
| CN106792589A (en) * | 2016-11-17 | 2017-05-31 | 中卓信（北京）科技有限公司 | Presentation of information updating menu method and apparatus |
| CN107797989A (en) * | 2017-10-16 | 2018-03-13 | 平安科技（深圳）有限公司 | Enterprise name recognition methods, electronic equipment and computer-readable recording medium |
| CN108763205A (en) * | 2018-05-21 | 2018-11-06 | 阿里巴巴集团控股有限公司 | A kind of brand alias recognition methods, device and electronic equipment |
-
2018
- 2018-12-21 CN CN201811574935.5A patent/CN109726266A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101030856A (en) * | 2006-07-19 | 2007-09-05 | 王李琰 | Method for verifying SMS and transmitting reliability classification based on cipher technology mark |
| US20090234925A1 (en) * | 2008-03-14 | 2009-09-17 | International Business Machines Corporation | Dyanmic Domain Based Electronic Mail Signature Lines |
| CN104301532A (en) * | 2014-09-30 | 2015-01-21 | 小米科技有限责任公司 | Communication message identification method and device |
| CN106792589A (en) * | 2016-11-17 | 2017-05-31 | 中卓信（北京）科技有限公司 | Presentation of information updating menu method and apparatus |
| CN106779467A (en) * | 2016-12-31 | 2017-05-31 | 成都数联铭品科技有限公司 | Enterprises ' industry categorizing system based on automatic information screening |
| CN107797989A (en) * | 2017-10-16 | 2018-03-13 | 平安科技（深圳）有限公司 | Enterprise name recognition methods, electronic equipment and computer-readable recording medium |
| CN108763205A (en) * | 2018-05-21 | 2018-11-06 | 阿里巴巴集团控股有限公司 | A kind of brand alias recognition methods, device and electronic equipment |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111966806A (en) * | 2020-08-17 | 2020-11-20 | 深圳市活力天汇科技股份有限公司 | Information extraction and repair method for hotel reservation short message |
| CN111966806B (en) * | 2020-08-17 | 2023-10-27 | 深圳市活力天汇科技股份有限公司 | Information extraction and repair method for hotel reservation short message |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11062089B2 (en) | Method and apparatus for generating information | |
| CN109271631A (en) | Segmenting method, device, equipment and storage medium | |
| CN109299458A (en) | Entity recognition method, device, equipment and storage medium | |
| CN107679144A (en) | News sentence clustering method, device and storage medium based on semantic similarity | |
| WO2022134421A1 (en) | Multi-knowledge graph based intelligent reply method and apparatus, computer device and storage medium | |
| US20160321254A1 (en) | Unsolicited bulk email detection using url tree hashes | |
| CN107368856A (en) | Clustering method and device, the computer installation and readable storage medium storing program for executing of Malware | |
| CN110347790A (en) | Text duplicate checking method, apparatus, equipment and storage medium based on attention mechanism | |
| CN105630767A (en) | Text similarity comparison method and device | |
| CN107644106B (en) | Method, terminal device and storage medium for automatically mining service middleman | |
| CN107908796A (en) | E-Government duplicate checking method, apparatus and computer-readable recording medium | |
| CN115941322B (en) | Attack detection method, device, equipment and storage medium based on artificial intelligence | |
| CN106648557A (en) | Application programming interface API sharing method and device | |
| CN110300191A (en) | Service system and data processing method | |
| CN114496083A (en) | Cell type determination method, device, equipment and storage medium | |
| CN110309293A (en) | Text recommended method and device | |
| CN114995791A (en) | API (application program interface) polymerization method and device, electronic equipment and storage medium | |
| CN109726266A (en) | Note signature processing method, equipment and computer readable storage medium | |
| CN114726823A (en) | Domain name generation method, device and equipment based on generation countermeasure network | |
| CN112905753A (en) | Method and device for distinguishing text information | |
| CN110147223A (en) | Generation method, device and the equipment of Component Gallery | |
| CN110503117A (en) | The method and apparatus of data clusters | |
| CN109754273A (en) | The method and apparatus for promoting any active ues quantity | |
| WO2021000411A1 (en) | Neural network-based document classification method and apparatus, and device and storage medium | |
| CN110895655A (en) | Method and device for extracting text core phrase |
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: 20190507 |