CN111985491A - Similar information merging method, device, equipment and medium based on deep learning - Google Patents
Similar information merging method, device, equipment and medium based on deep learning Download PDFInfo
- Publication number
- CN111985491A CN111985491A CN202010917729.0A CN202010917729A CN111985491A CN 111985491 A CN111985491 A CN 111985491A CN 202010917729 A CN202010917729 A CN 202010917729A CN 111985491 A CN111985491 A CN 111985491A
- Authority
- CN
- China
- Prior art keywords
- word vector
- image set
- text
- output text
- key value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 50
- 238000013135 deep learning Methods 0.000 title claims description 37
- 239000013598 vector Substances 0.000 claims abstract description 157
- 238000007781 pre-processing Methods 0.000 claims abstract description 15
- 238000002372 labelling Methods 0.000 claims description 22
- 238000012545 processing Methods 0.000 claims description 18
- 230000006870 function Effects 0.000 claims description 17
- 230000008569 process Effects 0.000 claims description 10
- 238000004590 computer program Methods 0.000 claims description 7
- 238000001914 filtration Methods 0.000 claims description 6
- 238000004422 calculation algorithm Methods 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 4
- 239000000126 substance Substances 0.000 claims description 2
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 238000005516 engineering process Methods 0.000 abstract description 2
- 238000005070 sampling Methods 0.000 description 26
- 238000007726 management method Methods 0.000 description 5
- PCTMTFRHKVHKIS-BMFZQQSSSA-N (1s,3r,4e,6e,8e,10e,12e,14e,16e,18s,19r,20r,21s,25r,27r,30r,31r,33s,35r,37s,38r)-3-[(2r,3s,4s,5s,6r)-4-amino-3,5-dihydroxy-6-methyloxan-2-yl]oxy-19,25,27,30,31,33,35,37-octahydroxy-18,20,21-trimethyl-23-oxo-22,39-dioxabicyclo[33.3.1]nonatriaconta-4,6,8,10 Chemical compound C1C=C2C[C@@H](OS(O)(=O)=O)CC[C@]2(C)[C@@H]2[C@@H]1[C@@H]1CC[C@H]([C@H](C)CCCC(C)C)[C@@]1(C)CC2.O[C@H]1[C@@H](N)[C@H](O)[C@@H](C)O[C@H]1O[C@H]1/C=C/C=C/C=C/C=C/C=C/C=C/C=C/[C@H](C)[C@@H](O)[C@@H](C)[C@H](C)OC(=O)C[C@H](O)C[C@H](O)CC[C@@H](O)[C@H](O)C[C@H](O)C[C@](O)(C[C@H](O)[C@H]2C(O)=O)O[C@H]2C1 PCTMTFRHKVHKIS-BMFZQQSSSA-N 0.000 description 4
- 230000003321 amplification Effects 0.000 description 4
- 238000006243 chemical reaction Methods 0.000 description 4
- 238000012937 correction Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000003199 nucleic acid amplification method Methods 0.000 description 4
- 238000004891 communication Methods 0.000 description 2
- 238000009499 grossing Methods 0.000 description 2
- 239000004973 liquid crystal related substance Substances 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 239000002699 waste material Substances 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/14—Image acquisition
- G06V30/148—Segmentation of character regions
- G06V30/153—Segmentation of character regions using recognition of characters or words
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/10—Text processing
- G06F40/12—Use of codes for handling textual entities
- G06F40/126—Character encoding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/10—Text processing
- G06F40/166—Editing, e.g. inserting or deleting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
- G06F40/295—Named entity recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
Abstract
The invention relates to an artificial intelligence technology, and discloses a similar information merging method, which comprises the following steps: preprocessing a plurality of acquired original image sets to obtain an image set to be merged; respectively performing text recognition on the image sets to be combined by using a preset text recognition model to obtain word vectors and encode the word vectors, generating key values and corresponding result values, and respectively converting the image sets to be combined into output texts; calculating an editing distance between the output texts by using the key values; if the editing distance is larger than a preset threshold value, directly outputting the output text; otherwise, merging the output texts to obtain and output a merged data set. Furthermore, the invention also relates to a blockchain technique, and the output text can be stored in blockchain nodes. The invention also provides a similar information merging device, electronic equipment and a storage medium. The invention can intelligently combine the similar information and reduce manual intervention.
Description
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to a method and a device for merging similar information based on deep learning, electronic equipment and a computer readable storage medium.
Background
With the increasing perfection of legal system, currently, a judge can acquire different case information from different channels in the case handling process, the case information formats acquired from different channels are different, and the cases acquired at the same time can be similar or repeated. How to merge similar information becomes an increasingly important requirement.
The mainstream information merging method in the market at present is to merge information manually. However, this method is too dependent on manual work, and the efficiency is low, so that it is not possible to achieve efficient and personalized information merging.
Disclosure of Invention
The invention provides a method and a device for merging similar information based on deep learning, electronic equipment and a computer readable storage medium, and mainly aims to merge the similar information and reduce manual intervention.
In order to achieve the above object, the present invention provides a method for merging similar information based on deep learning, which includes:
acquiring a first original image set and a second original image set, and preprocessing the first original image set and the second original image set to obtain a first image set to be combined and a second image set to be combined;
performing text recognition on the first image set to be combined and the second image set to be combined by using a pre-trained text recognition model to obtain a first word vector and a second word vector, encoding the first word vector and the second word vector to generate a first key value, a first result value and a second key value corresponding to the first key value, and a second result value corresponding to the second key value, and converting the first image set to be combined and the second image set to be combined into a first output text and a second output text according to the first key value, the first result value, the second key value and the second result value;
calculating the editing distance between the first output text and the second output text by using the key value;
comparing the editing distance with a preset threshold value;
if the editing distance is smaller than or equal to a preset threshold value, merging the first output text and the second output text to obtain and output a merged data set;
and if the editing distance is larger than a preset threshold value, directly outputting the first output text and the second output text.
Optionally, said encoding said first and second word vectors comprises:
obtaining a unique coding function of the word vector;
compiling the unique coding function of the word vector into a code generation statement by using a compiler;
and encoding the first word vector and the second word vector by using the encoding generation statement.
Optionally, the text recognition model comprises:
the word vector layer is used for converting words and characters in the text contained in the first image set to be combined into a first word vector and converting words and characters in the text contained in the second image set to be combined into a second word vector;
the Bi-LSTM layer is used for dividing the first word vector and the second word vector, coding the divided first word vector and the divided second word vector to obtain a first coding representation of the first word vector and a second coding representation of the second word vector, and labeling the divided first word vector and the divided second word vector by using the first coding representation and the second coding representation to obtain a first key value and a corresponding first result value and a corresponding second key value and a corresponding second result value;
and the CRF layer is used for splicing the first key value and the corresponding first result value and the second key value and the result value of the same type in the corresponding second result value to generate an output text.
Optionally, the labeling the segmented first word vector and the segmented second word vector by using the first coding representation and the second coding representation includes:
setting a labeling queue task;
and labeling the first word vector and the second word vector according to the sequence of the labeling queue task.
Optionally, the splicing the key values and the result values of the same type to generate an output text includes:
splicing key values and result values of the same type;
and decoding the spliced text according to the inverse process of the encoding to generate the first output text and the second output text.
Optionally, the calculating an edit distance of the first output text and the second output text by using the key value includes:
calculating the edit distance Sim using an edit distance algorithmtopic:
Simtopic=Pearson(R,S)
Wherein, R is the key value of the first output text, S is the key value of the second output text, and Pearson is the edit distance operation.
Optionally, the pre-processing comprises:
amplifying the image signals in the first original image set and the image signals in the second original image set to obtain a first amplified image signal and a second amplified image signal;
and filtering the first amplified image signal and the second amplified image signal to obtain the first image set to be combined and the second image set to be combined.
In order to solve the above problem, the present invention further provides a similar information merging device based on deep learning, including:
the image merging module is used for merging the first original image set and the second original image set into a second image set to be merged;
a text recognition module, configured to perform text recognition on the first image set to be merged and the second image set to be merged by using a pre-trained text recognition model to obtain a first word vector and a second word vector, encode the first word vector and the second word vector to generate a first key value, a corresponding first result value, a corresponding second key value, and a corresponding second result value, and convert the first image set to be merged and the second image set to be merged into a first output text and a second output text according to the first key value, the corresponding first result value, the second key value, and the corresponding second result value;
the editing distance calculation module is used for calculating the editing distance between the first output text and the second output text by using the key value;
and the merging processing module is used for comparing the size between the editing distance and a preset threshold value, merging the first output text and the second output text to obtain and output a merged data set if the editing distance is smaller than or equal to the preset threshold value, and directly outputting the first output text and the second output text if the editing distance is larger than the preset threshold value.
In order to solve the above problem, the present invention also provides an electronic device, including:
at least one processor; and the number of the first and second groups,
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores instructions executable by the at least one processor to cause the at least one processor to implement the deep learning based similarity information merging method described above.
In order to solve the above problem, the present invention also provides a computer-readable storage medium including a storage data area storing created data and a storage program area storing a computer program; wherein, the computer program realizes the similar information merging method based on deep learning when being executed by a processor.
The method comprises the steps of obtaining a plurality of original image sets, preprocessing the original image sets respectively to obtain image sets to be combined, performing text recognition on the image sets to be combined by utilizing a text recognition model trained in advance to obtain an output text, calculating the editing distance of the output text, judging the editing distance of the output text and a preset threshold value, and determining whether the text is combined or directly output. Therefore, the method, the device and the computer readable storage medium for merging the similar information based on deep learning provided by the invention can intelligently merge the similar information and reduce manual intervention.
Drawings
Fig. 1 is a schematic flowchart of a method for merging similar information based on deep learning according to an embodiment of the present invention;
fig. 2 is a block diagram of a deep learning-based similarity information merging apparatus according to an embodiment of the present invention;
fig. 3 is a schematic diagram of an internal structure of an electronic device implementing a deep learning-based similar information merging method according to an embodiment of the present invention;
the implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
The execution subject of the deep learning based similar information merging method provided by the embodiment of the present application includes, but is not limited to, at least one of electronic devices such as a server and a terminal that can be configured to execute the method provided by the embodiment of the present application. In other words, the deep learning based similar information merging method may be performed by software or hardware installed in a terminal device or a server device, and the software may be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, and the like.
The invention provides a similar information merging method based on deep learning. Fig. 1 is a schematic flow chart of a deep learning-based similar information merging method according to an embodiment of the present invention. In this embodiment, the method for merging similar information based on deep learning includes:
s1, acquiring a first original image set and a second original image set, and preprocessing the first original image set and the second original image set to obtain a first image set to be merged and a second image set to be merged.
In an embodiment of the present invention, the first original image set and the second original image set are obtained by two-dimensionally scanning a paper document, such as an image, a certificate paper file, and the like.
Wherein the paper documents include, but are not limited to, asset information obtained from different channels, such as: the method comprises the steps of obtaining account opening information and account asset information of a user from a commercial bank and a people bank, and obtaining real estate information of a party from a real estate bureau and a real estate registration center.
In order to remove interference factors such as noise in the first original image set and the second original image set obtained by two-dimensional scanning, the embodiment of the present invention first performs the following preprocessing on the first original image set and the second original image set, including:
amplifying the image signals in the first original image set and the image signals in the second original image set to obtain a first amplified image signal and a second amplified image signal;
and filtering the first amplified image signal and the second amplified image signal to obtain the first image set to be combined and the second image set to be combined.
In detail, the amplification process includes:
sampling image signals in the first original image set and image signals in the second original image set to obtain a first sampling data set and a second sampling data set;
and performing smooth data processing on the first sampling data set and the second sampling data set to obtain a first amplified image signal and a second amplified image signal.
Further, the sampling the image signals in the first original image set and the image signals in the second original image set to obtain a first sampling data set and a second sampling data set includes:
mapping the image signals in the first original image set and the image signals in the first original image set to a rectangular coordinate system;
and sampling the image signals in the first original image set and the image signals in the first original image set according to a preset sampling interval by taking the x axis of the rectangular coordinate system as the direction to obtain a first sampling data set and a second sampling data set.
Specifically, the performing smooth data processing on the first sampled data set and the second sampled data set to obtain a first amplified image signal and a second amplified image signal includes:
Y1(n)=aX1(n)+(1-α)Y1(n-1)
Y2(n)=aX2(n)+(1-α)Y2(n-1)
where Y1(n) is the first amplified image signal, X1(n) is the first sampled data set, Y1(n-1) is the image signal in the first original image set, Y2(n) is the second amplified image signal, X2(n) is the second sampled data set, Y2(n-1) is the image signal in the second original image set, and α is the smoothing coefficient.
In the embodiment of the present invention, the first original image set and the second original image set are subjected to the above-mentioned amplification, sampling, and filtering, so that interference factors such as noise in the first original image set and the second original image set are removed, the first image set to be merged and the second image set to be merged are obtained, and accuracy of subsequent text error correction is ensured.
It should be noted that, step S1 is to remove interference factors such as noise in the image, and when there is no noise in the first original image set and the second original image set or there is relatively little noise, step S1 may be omitted.
S2, performing text recognition on the first image set to be combined and the second image set to be combined by using a pre-trained text recognition model to obtain a first word vector and a second word vector, encoding the first word vector and the second word vector to generate a first key value, a corresponding first result value, a corresponding second key value and a corresponding second result value, and converting the first image set to be combined and the second image set to be combined into a first output text and a second output text according to the first key value, the corresponding first result value, the corresponding second key value and the corresponding second result value.
Preferably, in this embodiment of the present invention, the pre-trained text Recognition model may be a pre-trained NER (Named Entity Recognition) model.
Preferably, the NER model adopts a Bi-LSTM-CRF structure, and comprises the following components:
the word vector layer is used for converting words and characters in the text contained in the first image set to be combined into a first word vector and converting words and characters in the text contained in the second image set to be combined into a second word vector;
the Bi-LSTM layer is used for dividing the first word vector and the second word vector, coding the divided first word vector and the divided second word vector to obtain a first coding representation of the first word vector and a second coding representation of the second word vector, and labeling the divided first word vector and the divided second word vector by using the first coding representation and the second coding representation to obtain a first key value and a corresponding first result value and a corresponding second key value and a corresponding second result value;
and the CRF layer is used for splicing the first key value and the corresponding first result value and the second key value and the result value of the same type in the corresponding second result value to generate an output text.
The word vector layer converts words and characters in the text contained in the first image set to be merged into a first word vector and converts words and characters in the text contained in the second image set to be merged into a second word vector by using the trained word vector as an initialization parameter. The trained word vectors are a set of standard conversion rules summarized in the past when the first word vectors and the second word vectors are converted.
Because the first image set to be merged and the second image set to be merged may contain more texts and the sentences in the texts may be longer, and text viscosity may occur if only character conversion is performed, which is not beneficial to subsequent text error correction, the embodiment of the present invention may use the Bi-LSTM layer to segment the first word vector and the second word vector.
Preferably, the Bi-LSTM layer may use java language to divide the first word vector and the second word vector, and encode the divided first word vector and second word vector to obtain a first encoded representation of the first word vector and a second encoded representation of the second word vector. The coding representation comprises six types of labeling types of Key-B, Value-B, Key-I, Value-I, Other-B and Other-I. Wherein, Key is Key Value, Value is result Value, Other is Value, B and I are parameters.
Further, said encoding the first word vector and the second word vector comprises:
obtaining a unique coding function of the word vector;
compiling the unique coding function of the word vector into a code generation statement by using a compiler;
and executing the coding generation statement to generate and obtain a first coding representation of the first word vector and a second coding representation of the second word vector.
The unique coding function of the word vector enables the word vector to have unique identification degree, and the word vector does not need to be identified in other modes.
Specifically, labeling the segmented first word vector and the segmented second word vector by using the first coding representation and the second coding representation includes:
setting a labeling queue task;
and labeling the first word vector and the second word vector according to the sequence of the labeling queue task.
Preferably, the tagging queue task is implemented by using a subscriber notification Message Queue (MQ), and specifically, multiple word vectors to be tagged are processed in batches by setting an interval threshold of time, so as to ensure that a previous word vector is sent to the end and then processed continuously.
The subscriber informs the message queue to reduce the occupation of computing resources, cuts a large amount of data and pushes the data in batches, and avoids the occupation and waste of the computing resources caused by data congestion.
In detail, the generating an output text by splicing the key values and the result values of the same type includes:
splicing key values and result values of the same type;
and decoding the spliced text according to the inverse process of the encoding to generate the first output text and the second output text.
For example, the code representation Key-B and the code representation Key-I belong to the same type, the code representation Value-B and the code representation Value-I belong to the same type, and the code representation Other-B and the code representation Other-I belong to the same type.
Further, in the embodiment of the present invention, the first image set to be merged and the second image set to be merged are converted into the first output text and the second output text according to the key value and the corresponding result value.
For example, in the embodiment of the present invention, the first to-be-merged image set has a text "pay 2.00 m (cash payment) classification is negative 0.00 m", and after being processed by the NER model, the following output text is generated:
key: { Payment, classification self-burden } |
Value: {2.00 yuan, 0.00 yuan } |
Other (Cash payment) |
And S3, calculating the edit distance of the first output text and the second output text by using the key value.
In the embodiment of the present invention, the edit distance refers to the minimum number of edit operations required to convert one character string into another character string between two character strings, wherein the smaller the edit distance is, the greater the text similarity between two texts is.
For example, the edit distance between the first output text ABBD and the second output text ABCD is calculated, and since only the third character is different between the first output text ABBD and the second output text ABCD, the minimum number of edit operations is calculated to be 1 by the above method, that is, the 'B' character is replaced by the 'C' character.
In detail, the calculating an edit distance between the first output text and the second output text by using the key value in the embodiment of the present invention includes:
calculating the edit distance Sim using an edit distance algorithmtopic:
Simtopic=Pearson(R,S)
Wherein, R is the key value of the first output text, S is the key value of the second output text, and Pearson is the edit distance operation.
And S4, comparing the editing distance with a preset threshold value.
And if the editing distance is smaller than or equal to a preset threshold value, executing S5, and merging the first output text and the second output text to obtain and output a merged data set.
Because the edit distance is less than or equal to the preset threshold, it indicates that the similarity between the first output text and the second output text is high, and therefore, subsequent merging operation is required, and the data redundancy is reduced to the lowest, so that the computer is more efficient in the subsequent data processing process.
Specifically, if the edit distance is less than or equal to the preset threshold, it may be determined that the similarity between the first output text and the second output text is high.
If the edit distance is greater than the preset threshold, it may be determined that the similarity between the first output text and the second output text is not high, and it is not necessary to merge the two texts, and S6 is performed to directly output the first output text and the second output text.
Fig. 2 is a schematic block diagram of a similar information merging device based on deep learning according to the present invention.
The deep learning based similar information merging apparatus 100 of the present invention may be installed in an electronic device. According to the implemented functions, the deep learning based similarity information combining device 100 may include a preprocessing module 101, a text recognition module 102, an edit distance calculation module 103, and a combining processing module 104. A module according to the present invention, which may also be referred to as a unit, refers to a series of computer program segments that can be executed by a processor of an electronic device and that can perform a fixed function, and that are stored in a memory of the electronic device.
In the present embodiment, the functions regarding the respective modules/units are as follows:
the preprocessing module 101 is configured to obtain a first original image set and a second original image set, and preprocess the first original image set and the second original image set to obtain a first image set to be merged and a second image set to be merged;
the text recognition module 102 is configured to perform text recognition on the first image set to be merged and the second image set to be merged by using a pre-trained text recognition model to obtain a first word vector and a second word vector, encode the first word vector and the second word vector to generate a first key value, a corresponding first result value, a corresponding second key value, and a corresponding second result value, and convert the first image set to be merged and the second image set to be merged into a first output text and a second output text according to the first key value, the corresponding first result value, the second key value, and the corresponding second result value;
the edit distance calculating module 103 is configured to calculate an edit distance between the first output text and the second output text by using the key value;
the merging processing module 104 is configured to compare the editing distance with a preset threshold, merge the first output text and the second output text to obtain and output a merged data set if the editing distance is smaller than or equal to the preset threshold, and directly output the first output text and the second output text if the editing distance is greater than the preset threshold.
In detail, the similar information merging apparatus 100 based on deep learning can implement the similar information merging method described in the following steps through modules:
firstly, the preprocessing module 101 acquires a first original image set and a second original image set, and preprocesses the first original image set and the second original image set to obtain a first image set to be merged and a second image set to be merged.
In an embodiment of the present invention, the first original image set and the second original image set are obtained by two-dimensionally scanning a paper document, such as an image, a certificate paper file, and the like.
Wherein the paper documents include, but are not limited to, asset information obtained from different channels, such as: the method comprises the steps of obtaining account opening information and account asset information of a user from a commercial bank and a people bank, and obtaining real estate information of a party from a real estate bureau and a real estate registration center.
In order to remove interference factors such as noise in the first original image set and the second original image set obtained by two-dimensional scanning, the preprocessing module 101 according to the embodiment of the present invention first performs the following preprocessing on the first original image set and the second original image set, including:
amplifying the image signals in the first original image set and the image signals in the second original image set to obtain a first amplified image signal and a second amplified image signal;
and filtering the first amplified image signal and the second amplified image signal to obtain the first image set to be combined and the second image set to be combined.
In detail, the amplification process includes:
sampling image signals in the first original image set and image signals in the second original image set to obtain a first sampling data set and a second sampling data set;
and performing smooth data processing on the first sampling data set and the second sampling data set to obtain a first amplified image signal and a second amplified image signal.
Further, the sampling the image signals in the first original image set and the image signals in the second original image set to obtain a first sampling data set and a second sampling data set includes:
mapping the image signals in the first original image set and the image signals in the first original image set to a rectangular coordinate system;
and sampling the image signals in the first original image set and the image signals in the first original image set according to a preset sampling interval by taking the x axis of the rectangular coordinate system as the direction to obtain a first sampling data set and a second sampling data set.
Specifically, the performing smooth data processing on the first sampled data set and the second sampled data set to obtain a first amplified image signal and a second amplified image signal includes:
Y1(n)=aX1(n)+(1-α)Y1(n-1)
Y2(n)=aX2(n)+(1-α)Y2(n-1)
where Y1(n) is the first amplified image signal, X1(n) is the first sampled data set, Y1(n-1) is the image signal in the first original image set, Y2(n) is the second amplified image signal, X2(n) is the second sampled data set, Y2(n-1) is the image signal in the second original image set, and α is the smoothing coefficient.
In the embodiment of the present invention, the first original image set and the second original image set are subjected to the above-mentioned amplification, sampling, and filtering, so that interference factors such as noise in the first original image set and the second original image set are removed, the first image set to be merged and the second image set to be merged are obtained, and accuracy of subsequent text error correction is ensured.
It should be noted that, in order to remove interference factors such as noise in the images, when there is no noise in the first original image set and the second original image set or there is relatively little noise, the first step may be omitted.
The text recognition module 102 performs text recognition on the first image set to be merged and the second image set to be merged by using a pre-trained text recognition model to obtain a first word vector and a second word vector, encodes the first word vector and the second word vector to generate a first key value and a corresponding first result value as well as a second key value and a corresponding second result value, and converts the first image set to be merged and the second image set to be merged into a first output text and a second output text according to the first key value and the corresponding first result value as well as the second key value and the corresponding second result value.
Preferably, in this embodiment of the present invention, the pre-trained text Recognition model may be a pre-trained NER (Named Entity Recognition) model.
Preferably, the NER model adopts a Bi-LSTM-CRF structure, and comprises the following components:
the word vector layer is used for converting words and characters in the text contained in the first image set to be combined into a first word vector and converting words and characters in the text contained in the second image set to be combined into a second word vector;
the Bi-LSTM layer is used for dividing the first word vector and the second word vector, coding the divided first word vector and the divided second word vector to obtain a first coding representation of the first word vector and a second coding representation of the second word vector, and labeling the divided first word vector and the divided second word vector by using the first coding representation and the second coding representation to obtain a first key value and a corresponding first result value and a corresponding second key value and a corresponding second result value;
and the CRF layer is used for splicing the first key value and the corresponding first result value and the second key value and the result value of the same type in the corresponding second result value to generate an output text.
The word vector layer converts words and characters in the text contained in the first image set to be merged into a first word vector and converts words and characters in the text contained in the second image set to be merged into a second word vector by using the trained word vector as an initialization parameter. The trained word vectors are a set of standard conversion rules summarized in the past when the first word vectors and the second word vectors are converted.
Because there may be more texts contained in the first image set to be merged and the second image set to be merged, and sentences in the texts may be longer, if only character conversion is performed, a text sticky condition may occur, which is not favorable for subsequent text error correction, the text recognition module 102 in the embodiment of the present invention may use the Bi-LSTM layer to segment the first word vector and the second word vector.
Preferably, the Bi-LSTM layer may use java language to divide the first word vector and the second word vector, and encode the divided first word vector and second word vector to obtain a first encoded representation of the first word vector and a second encoded representation of the second word vector. The coding representation comprises six types of labeling types of Key-B, Value-B, Key-I, Value-I, Other-B and Other-I. Wherein, Key is Key Value, Value is result Value, Other is Value, B and I are parameters.
Further, the text recognition module 102 encodes the first word vector and the second word vector, including:
obtaining a unique coding function of the word vector;
compiling the unique coding function of the word vector into a code generation statement by using a compiler;
and executing the coding generation statement to generate and obtain a first coding representation of the first word vector and a second coding representation of the second word vector.
The unique coding function of the word vector enables the word vector to have unique identification degree, and the word vector does not need to be identified in other modes.
Specifically, the labeling, by the text recognition module 102, the segmented first word vector and the segmented second word vector by using the first encoding representation and the second encoding representation includes:
setting a labeling queue task;
and labeling the first word vector and the second word vector according to the sequence of the labeling queue task.
Preferably, the tagging queue task is implemented by using a subscriber notification Message Queue (MQ), and specifically, multiple word vectors to be tagged are processed in batches by setting an interval threshold of time, so as to ensure that a previous word vector is sent to the end and then processed continuously.
The subscriber informs the message queue to reduce the occupation of computing resources, cuts a large amount of data and pushes the data in batches, and avoids the occupation and waste of the computing resources caused by data congestion.
In detail, the CRF layer concatenates the first key value and the corresponding first result value and the same type of key value and result value in the second result value and the corresponding second result value to generate an output text, including:
splicing the first key value and the corresponding first result value and the same type of key values and result values in the second key value and the corresponding second result value;
and decoding the spliced text according to the inverse process of the encoding to generate the first output text and the second output text.
For example, the code representation Key-B and the code representation Key-I belong to the same type, the code representation Value-B and the code representation Value-I belong to the same type, and the code representation Other-B and the code representation Other-I belong to the same type.
Further, in the embodiment of the present invention, the first image set to be merged and the second image set to be merged are converted into the first output text and the second output text according to the key value and the corresponding result value.
For example, in the embodiment of the present invention, the first to-be-merged image set has a text "pay 2.00 m (cash payment) classification is negative 0.00 m", and after being processed by the NER model, the following output text is generated:
key: { Payment, classification self-burden } |
Value: {2.00 yuan, 0.00 yuan } |
Other (Cash payment) |
Step three, the edit distance calculation module 103 calculates the edit distance of the first output text and the second output text by using the key value.
In the embodiment of the present invention, the edit distance refers to the minimum number of edit operations required to convert one character string into another character string between two character strings, wherein the smaller the edit distance is, the greater the text similarity between two texts is.
For example, the edit distance calculating module 103 calculates the edit distance between the first output text ABBD and the second output text ABCD, and since only the third character is different between the first output text ABBD and the second output text ABCD, the minimum number of edit operations is calculated to be 1 by using the above method, that is, the 'B' character is replaced by the 'C' character.
In detail, the edit distance calculating module 103 according to the embodiment of the present invention calculates the edit distance between the first output text and the second output text by using the key value, and includes:
calculating the edit distance Sim using an edit distance algorithmtopic:
Simtopic=Pearson(R,S)
Wherein, R is the key value of the first output text, S is the key value of the second output text, and Pearson is the edit distance operation.
Step four, the merging processing module 104 compares the size between the edit distance and a preset threshold.
If the edit distance is smaller than or equal to a preset threshold, the merge processing module 104 merges the first output text and the second output text to obtain and output a merged data set.
Because the edit distance is less than or equal to the preset threshold, it indicates that the similarity between the first output text and the second output text is high, and therefore, subsequent merging operation is required, and the data redundancy is reduced to the lowest, so that the computer is more efficient in the subsequent data processing process.
Specifically, if the edit distance is less than or equal to the preset threshold, it may be determined that the similarity between the first output text and the second output text is high, and in the embodiment of the present invention, the merge processing module 104 performs merge processing on the first output text and the second output text, directly deletes any one of the output texts, and outputs another text as the merged text set.
If the edit distance is greater than the preset threshold, it may be determined that the similarity between the first output text and the second output text is not high, and it is not necessary to merge the two texts, and the merge processing module 104 directly outputs the first output text and the second output text.
Fig. 3 is a schematic structural diagram of an electronic device implementing a deep learning-based similar information merging method according to the present invention.
The electronic device 1 may comprise a processor 10, a memory 11 and a bus, and may further comprise a computer program, such as a deep learning based similar information merging program 12, stored in the memory 11 and executable on the processor 10.
The memory 11 includes at least one type of readable storage medium, which includes flash memory, removable hard disk, multimedia card, card-type memory (e.g., SD or DX memory, etc.), magnetic memory, magnetic disk, optical disk, etc. The memory 11 may in some embodiments be an internal storage unit of the electronic device 1, such as a removable hard disk of the electronic device 1. The memory 11 may also be an external storage device of the electronic device 1 in other embodiments, such as a plug-in mobile hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and the like, which are provided on the electronic device 1. Further, the memory 11 may also include both an internal storage unit and an external storage device of the electronic device 1. The memory 11 may be used not only to store application software installed in the electronic device 1 and various types of data, such as codes of the deep learning-based similar information merging program 12, but also to temporarily store data that has been output or is to be output.
The processor 10 may be composed of an integrated circuit in some embodiments, for example, a single packaged integrated circuit, or may be composed of a plurality of integrated circuits packaged with the same or different functions, including one or more Central Processing Units (CPUs), microprocessors, digital Processing chips, graphics processors, and combinations of various control chips. The processor 10 is a Control Unit (Control Unit) of the electronic device, connects various components of the whole electronic device by using various interfaces and lines, and executes various functions and processes data of the electronic device 1 by running or executing programs or modules stored in the memory 11 (for example, executing a similar information merging program based on deep learning, etc.), and calling data stored in the memory 11.
The bus may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. The bus is arranged to enable connection communication between the memory 11 and at least one processor 10 or the like.
Fig. 3 shows only an electronic device with components, and it will be understood by those skilled in the art that the structure shown in fig. 3 does not constitute a limitation of the electronic device 1, and may comprise fewer or more components than those shown, or some components may be combined, or a different arrangement of components.
For example, although not shown, the electronic device 1 may further include a power supply (such as a battery) for supplying power to each component, and preferably, the power supply may be logically connected to the at least one processor 10 through a power management device, so as to implement functions of charge management, discharge management, power consumption management, and the like through the power management device. The power supply may also include any component of one or more dc or ac power sources, recharging devices, power failure detection circuitry, power converters or inverters, power status indicators, and the like. The electronic device 1 may further include various sensors, a bluetooth module, a Wi-Fi module, and the like, which are not described herein again.
Further, the electronic device 1 may further include a network interface, and optionally, the network interface may include a wired interface and/or a wireless interface (such as a WI-FI interface, a bluetooth interface, etc.), which are generally used for establishing a communication connection between the electronic device 1 and other electronic devices.
Optionally, the electronic device 1 may further comprise a user interface, which may be a Display (Display), an input unit (such as a Keyboard), and optionally a standard wired interface, a wireless interface. Alternatively, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch device, or the like. The display, which may also be referred to as a display screen or display unit, is suitable for displaying information processed in the electronic device 1 and for displaying a visualized user interface, among other things.
It is to be understood that the described embodiments are for purposes of illustration only and that the scope of the appended claims is not limited to such structures.
The deep learning based similarity information merging program 12 stored in the memory 11 of the electronic device 1 is a combination of a plurality of instructions, and when running in the processor 10, can realize:
acquiring a first original image set and a second original image set, and preprocessing the first original image set and the second original image set to obtain a first image set to be combined and a second image set to be combined;
performing text recognition on the first image set to be combined and the second image set to be combined by using a pre-trained text recognition model to obtain a first word vector and a second word vector, encoding the first word vector and the second word vector to generate a first key value, a first result value and a second key value corresponding to the first key value and a second result value corresponding to the second key value, and converting the first image set to be combined and the second image set to be combined into a first output text and a second output text according to the first key value, the first result value, the second key value and the second result value;
calculating the editing distance between the first output text and the second output text by using the key value;
comparing the editing distance with a preset threshold value;
if the editing distance is smaller than or equal to a preset threshold value, merging the first output text and the second output text to obtain and output a merged data set;
and if the editing distance is larger than a preset threshold value, directly outputting the first output text and the second output text.
The method comprises the steps of obtaining two original image sets, preprocessing the two original image sets respectively to obtain two image sets to be combined, performing text recognition on the two image sets to be combined by utilizing a text recognition model trained in advance to obtain two output texts, calculating the editing distance of the two output texts, judging the editing distance of the two output texts and a preset threshold value, and determining whether the two output texts are combined or directly outputting the texts. Therefore, the method, the device and the computer-readable storage medium for merging the similar information based on deep learning provided by the invention can merge the similar information and reduce manual intervention.
Further, the integrated modules/units of the electronic device 1, if implemented in the form of software functional units and sold or used as separate products, may be stored in a computer readable storage medium. The computer-readable medium may include: any entity or device capable of carrying said computer program code, recording medium, U-disk, removable hard disk, magnetic disk, optical disk, computer Memory, Read-Only Memory (ROM).
Further, the computer usable storage medium may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function, and the like; the storage data area may store data created according to the use of the blockchain node, and the like.
In the embodiments provided in the present invention, it should be understood that the disclosed apparatus, device and method can be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the modules is only one logical functional division, and other divisions may be realized in practice.
The modules described as separate parts may or may not be physically separate, and parts displayed as modules may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment.
In addition, functional modules in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, or in a form of hardware plus a software functional module.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof.
The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any accompanying claims should not be construed as limiting the claim concerned.
The block chain is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.
Furthermore, it is obvious that the word "comprising" does not exclude other elements or steps, and the singular does not exclude the plural. A plurality of units or means recited in the system claims may also be implemented by one unit or means in software or hardware. The terms second, etc. are used to denote names, but not any particular order.
Finally, it should be noted that the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting, and although the present invention is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions may be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.
Claims (10)
1. A similar information merging method based on deep learning is characterized by comprising the following steps:
acquiring a first original image set and a second original image set, and preprocessing the first original image set and the second original image set to obtain a first image set to be combined and a second image set to be combined;
performing text recognition on the first image set to be combined and the second image set to be combined by using a pre-trained text recognition model to obtain a first word vector and a second word vector, encoding the first word vector and the second word vector to generate a first key value, a first result value and a second key value corresponding to the first key value, and a second result value corresponding to the second key value, and converting the first image set to be combined and the second image set to be combined into a first output text and a second output text according to the first key value, the first result value, the second key value and the second result value;
calculating the editing distance between the first output text and the second output text by using the key value;
comparing the editing distance with a preset threshold value;
if the editing distance is smaller than or equal to a preset threshold value, merging the first output text and the second output text to obtain and output a merged data set;
and if the editing distance is larger than a preset threshold value, directly outputting the first output text and the second output text.
2. The deep learning based similarity information merging method according to claim 1, wherein the encoding the first word vector and the second word vector comprises:
obtaining a unique coding function of the word vector;
compiling the unique coding function of the word vector into a code generation statement by using a compiler;
and encoding the first word vector and the second word vector by using the encoding generation statement.
3. The deep learning based similar information merging method according to claim 1, wherein the text recognition model comprises:
the word vector layer is used for converting words and characters in the text contained in the first image set to be combined into a first word vector and converting words and characters in the text contained in the second image set to be combined into a second word vector;
the Bi-LSTM layer is used for dividing the first word vector and the second word vector, coding the divided first word vector and the divided second word vector to obtain a first coding representation of the first word vector and a second coding representation of the second word vector, and labeling the divided first word vector and the divided second word vector by using the first coding representation and the second coding representation to obtain a first key value and a corresponding first result value and a corresponding second key value and a corresponding second result value;
and the CRF layer is used for splicing the first key value and the corresponding first result value and the second key value and the result value of the same type in the corresponding second result value to generate an output text.
4. The deep learning based similarity information merging method according to claim 3, wherein labeling the segmented first word vector and the segmented second word vector using the first coded representation and the second coded representation comprises:
setting a labeling queue task;
and labeling the first word vector and the second word vector according to the sequence of the labeling queue task.
5. The method for merging similar information based on deep learning of claim 3, wherein the step of splicing key values and result values of the same type to generate an output text comprises:
splicing key values and result values of the same type;
and decoding the spliced text according to the inverse process of the encoding to generate the first output text and the second output text.
6. The method for merging similar information based on deep learning of claim 1, wherein the calculating the edit distance of the first output text and the second output text by using the key value comprises:
calculating the edit distance Sim using an edit distance algorithmtopic:
Simtopic=Pearson(R,S)
Wherein, R is the key value of the first output text, S is the key value of the second output text, and Pearson is the edit distance operation.
7. The deep learning based similarity information merging method according to any one of claims 1 to 6, wherein the preprocessing comprises:
amplifying the image signals in the first original image set and the image signals in the second original image set to obtain a first amplified image signal and a second amplified image signal;
and filtering the first amplified image signal and the second amplified image signal to obtain the first image set to be combined and the second image set to be combined.
8. An apparatus for merging similar information based on deep learning, the apparatus comprising:
the image merging module is used for merging the first original image set and the second original image set into a second image set to be merged;
a text recognition module, configured to perform text recognition on the first image set to be merged and the second image set to be merged by using a pre-trained text recognition model to obtain a first word vector and a second word vector, encode the first word vector and the second word vector to generate a first key value, a corresponding first result value, a corresponding second key value, and a corresponding second result value, and convert the first image set to be merged and the second image set to be merged into a first output text and a second output text according to the first key value, the corresponding first result value, the second key value, and the corresponding second result value;
the editing distance calculation module is used for calculating the editing distance between the first output text and the second output text by using the key value;
and the merging processing module is used for comparing the size between the editing distance and a preset threshold value, merging the first output text and the second output text to obtain and output a merged data set if the editing distance is smaller than or equal to the preset threshold value, and directly outputting the first output text and the second output text if the editing distance is larger than the preset threshold value.
9. An electronic device, characterized in that the electronic device comprises:
at least one processor; and the number of the first and second groups,
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the deep learning based similarity information merging method of any one of claims 1 to 7.
10. A computer-readable storage medium comprising a storage data area and a storage program area, wherein the storage data area stores created data, and the storage program area stores a computer program; wherein the computer program when executed by a processor implements the deep learning based similarity information merging method according to any one of claims 1 to 7.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010917729.0A CN111985491A (en) | 2020-09-03 | 2020-09-03 | Similar information merging method, device, equipment and medium based on deep learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010917729.0A CN111985491A (en) | 2020-09-03 | 2020-09-03 | Similar information merging method, device, equipment and medium based on deep learning |
Publications (1)
Publication Number | Publication Date |
---|---|
CN111985491A true CN111985491A (en) | 2020-11-24 |
Family
ID=73448709
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010917729.0A Pending CN111985491A (en) | 2020-09-03 | 2020-09-03 | Similar information merging method, device, equipment and medium based on deep learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111985491A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116704079A (en) * | 2023-08-03 | 2023-09-05 | 腾讯科技(深圳)有限公司 | Image generation method, device, equipment and storage medium |
CN117093717A (en) * | 2023-10-20 | 2023-11-21 | 湖南财信数字科技有限公司 | Similar text aggregation method, device, equipment and storage medium thereof |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH07200745A (en) * | 1993-12-17 | 1995-08-04 | Xerox Corp | Comparison method of at least two image sections |
WO2011041672A1 (en) * | 2009-10-02 | 2011-04-07 | Massachusetts Institute Of Technology | Translating text to, merging, and optimizing graphical user interface tasks |
CN103959282A (en) * | 2011-09-28 | 2014-07-30 | 谷歌公司 | Selective feedback for text recognition systems |
CN106997379A (en) * | 2017-03-20 | 2017-08-01 | 杭州电子科技大学 | A kind of merging method of the close text based on picture text click volume |
CN107491730A (en) * | 2017-07-14 | 2017-12-19 | 浙江大学 | A kind of laboratory test report recognition methods based on image procossing |
CN109885842A (en) * | 2018-02-22 | 2019-06-14 | 谷歌有限责任公司 | Handle text neural network |
CN110619119A (en) * | 2019-07-23 | 2019-12-27 | 平安科技(深圳)有限公司 | Intelligent text editing method and device and computer readable storage medium |
CN110782885A (en) * | 2019-09-29 | 2020-02-11 | 深圳和而泰家居在线网络科技有限公司 | Voice text correction method and device, computer equipment and computer storage medium |
CN111414916A (en) * | 2020-02-29 | 2020-07-14 | 中国平安财产保险股份有限公司 | Method and device for extracting and generating text content in image and readable storage medium |
CN111539211A (en) * | 2020-04-17 | 2020-08-14 | 中移(杭州)信息技术有限公司 | Entity and semantic relation recognition method and device, electronic equipment and storage medium |
-
2020
- 2020-09-03 CN CN202010917729.0A patent/CN111985491A/en active Pending
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH07200745A (en) * | 1993-12-17 | 1995-08-04 | Xerox Corp | Comparison method of at least two image sections |
WO2011041672A1 (en) * | 2009-10-02 | 2011-04-07 | Massachusetts Institute Of Technology | Translating text to, merging, and optimizing graphical user interface tasks |
CN103959282A (en) * | 2011-09-28 | 2014-07-30 | 谷歌公司 | Selective feedback for text recognition systems |
CN106997379A (en) * | 2017-03-20 | 2017-08-01 | 杭州电子科技大学 | A kind of merging method of the close text based on picture text click volume |
CN107491730A (en) * | 2017-07-14 | 2017-12-19 | 浙江大学 | A kind of laboratory test report recognition methods based on image procossing |
CN109885842A (en) * | 2018-02-22 | 2019-06-14 | 谷歌有限责任公司 | Handle text neural network |
CN110619119A (en) * | 2019-07-23 | 2019-12-27 | 平安科技(深圳)有限公司 | Intelligent text editing method and device and computer readable storage medium |
CN110782885A (en) * | 2019-09-29 | 2020-02-11 | 深圳和而泰家居在线网络科技有限公司 | Voice text correction method and device, computer equipment and computer storage medium |
CN111414916A (en) * | 2020-02-29 | 2020-07-14 | 中国平安财产保险股份有限公司 | Method and device for extracting and generating text content in image and readable storage medium |
CN111539211A (en) * | 2020-04-17 | 2020-08-14 | 中移(杭州)信息技术有限公司 | Entity and semantic relation recognition method and device, electronic equipment and storage medium |
Non-Patent Citations (2)
Title |
---|
褚力;张世永;: "基于集成合并的文本特征提取方法", 计算机应用与软件, no. 10, 15 October 2008 (2008-10-15), pages 218 - 219 * |
谢腾;杨俊安;刘辉;: "基于BERT-BiLSTM-CRF模型的中文实体识别", 计算机系统应用, no. 07, 15 July 2020 (2020-07-15), pages 52 - 59 * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116704079A (en) * | 2023-08-03 | 2023-09-05 | 腾讯科技(深圳)有限公司 | Image generation method, device, equipment and storage medium |
CN116704079B (en) * | 2023-08-03 | 2023-09-29 | 腾讯科技(深圳)有限公司 | Image generation method, device, equipment and storage medium |
CN117093717A (en) * | 2023-10-20 | 2023-11-21 | 湖南财信数字科技有限公司 | Similar text aggregation method, device, equipment and storage medium thereof |
CN117093717B (en) * | 2023-10-20 | 2024-01-30 | 湖南财信数字科技有限公司 | Similar text aggregation method, device, equipment and storage medium thereof |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112597312A (en) | Text classification method and device, electronic equipment and readable storage medium | |
CN111681681A (en) | Voice emotion recognition method and device, electronic equipment and storage medium | |
CN112446207A (en) | Title generation method and device, electronic equipment and storage medium | |
CN111985491A (en) | Similar information merging method, device, equipment and medium based on deep learning | |
CN113064994A (en) | Conference quality evaluation method, device, equipment and storage medium | |
CN112733551A (en) | Text analysis method and device, electronic equipment and readable storage medium | |
CN114898373A (en) | File desensitization method and device, electronic equipment and storage medium | |
CN111626118A (en) | Text error correction method and device, electronic equipment and computer readable storage medium | |
CN113205814A (en) | Voice data labeling method and device, electronic equipment and storage medium | |
CN112883346A (en) | Safety identity authentication method, device, equipment and medium based on composite data | |
CN115409041B (en) | Unstructured data extraction method, device, equipment and storage medium | |
CN116127925B (en) | Text data enhancement method and device based on destruction processing of text | |
CN113626605B (en) | Information classification method, device, electronic equipment and readable storage medium | |
WO2022156088A1 (en) | Fingerprint signature generation method and apparatus, and electronic device and computer storage medium | |
CN114943306A (en) | Intention classification method, device, equipment and storage medium | |
CN112580505B (en) | Method and device for identifying network point switch door state, electronic equipment and storage medium | |
CN113822215A (en) | Equipment operation guide file generation method and device, electronic equipment and storage medium | |
CN114186540A (en) | Mail content intelligent filling method and device, electronic equipment and storage medium | |
CN112528984A (en) | Image information extraction method, device, electronic equipment and storage medium | |
CN113486646A (en) | Product report issuing method and device, electronic equipment and readable storage medium | |
CN112712797A (en) | Voice recognition method and device, electronic equipment and readable storage medium | |
CN113536782A (en) | Sensitive word recognition method and device, electronic equipment and storage medium | |
CN112183520A (en) | Intelligent data information processing method and device, electronic equipment and storage medium | |
CN113704478B (en) | Text element extraction method, device, electronic equipment and medium | |
CN116110056B (en) | Information extraction method and device, electronic equipment and storage medium |
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 |