CN113468486A - Big data watermarking method based on artificial intelligence - Google Patents
Big data watermarking method based on artificial intelligence Download PDFInfo
- Publication number
- CN113468486A CN113468486A CN202110495216.XA CN202110495216A CN113468486A CN 113468486 A CN113468486 A CN 113468486A CN 202110495216 A CN202110495216 A CN 202110495216A CN 113468486 A CN113468486 A CN 113468486A
- Authority
- CN
- China
- Prior art keywords
- data
- watermark
- target file
- historical
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 36
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 29
- 238000003062 neural network model Methods 0.000 claims abstract description 22
- 238000000605 extraction Methods 0.000 claims abstract description 4
- 238000001514 detection method Methods 0.000 claims description 58
- 230000003828 downregulation Effects 0.000 claims description 47
- 239000013598 vector Substances 0.000 claims description 33
- 238000012549 training Methods 0.000 claims description 19
- 238000004364 calculation method Methods 0.000 claims description 12
- 238000012360 testing method Methods 0.000 claims description 11
- 238000007781 pre-processing Methods 0.000 claims description 10
- 230000003827 upregulation Effects 0.000 claims description 10
- 238000012216 screening Methods 0.000 claims description 4
- 230000001939 inductive effect Effects 0.000 claims description 3
- 230000001105 regulatory effect Effects 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 description 7
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000012795 verification Methods 0.000 description 3
- 230000007613 environmental effect Effects 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 239000000047 product Substances 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/10—Protecting distributed programs or content, e.g. vending or licensing of copyrighted material ; Digital rights management [DRM]
- G06F21/16—Program or content traceability, e.g. by watermarking
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Technology Law (AREA)
- Computer Hardware Design (AREA)
- Computer Security & Cryptography (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Editing Of Facsimile Originals (AREA)
Abstract
The invention provides a big data watermarking method based on artificial intelligence, which comprises the following steps: when a request of adding a watermark of a user is received, acquiring first target file data and first environment data corresponding to the request; performing feature extraction on the first target file data and the first environment data to obtain a plurality of feature values; inputting the characteristic value into a neural network model trained in advance to obtain the parameter of the watermark; based on the parameters, watermarking is performed on the target file. The big data watermarking method based on artificial intelligence analyzes the historical added data of the user based on artificial intelligence so as to realize intelligent watermarking selection.
Description
Technical Field
The invention relates to the technical field of watermarking, in particular to a big data watermarking method based on artificial intelligence.
Background
At present, Digital Watermarking (Digital Watermarking) is an important method for protecting data assets, and Digital signals, such as images, characters, symbols, numbers and the like, are usually embedded in Digital products and can be used as identification and marking information, so that the purpose of the Digital Watermarking is to perform copyright protection, ownership certification, fingerprint and integrity protection and the like.
When a user adds the watermark, the user often needs to manually set the watermark, and after the setting is finished, the user does not change the setting, so that the added watermark cannot be changed.
Disclosure of Invention
One of the purposes of the invention is to provide a big data watermarking method based on artificial intelligence, which analyzes historical added data of a user based on artificial intelligence so as to realize intelligent watermarking selection.
The embodiment of the invention provides a big data watermarking method based on artificial intelligence, which comprises the following steps:
when a request of adding a watermark of a user is received, acquiring first target file data and first environment data corresponding to the request;
performing feature extraction on the first target file data and the first environment data to obtain a plurality of feature values;
inputting the characteristic value into a neural network model trained in advance to obtain the parameter of the watermark;
based on the parameters, watermarking is performed on the target file.
Preferably, the neural network model is generated by training as follows:
acquiring historical watermark adding data of a user from a big data platform based on user parameters;
preprocessing the historical watermark adding data to obtain a historical data set;
dividing a historical data set into a training set and a testing set based on a preset rule;
training the initial second neural network model based on the training set and the test set, and taking the initial second neural network model as the neural network model after the initial second neural network model converges;
preprocessing the historical watermark adding data to obtain a historical data set, wherein the preprocessing the historical watermark adding data to obtain the historical data set comprises the following steps:
analyzing the historical watermark adding data to obtain second target file data and second environment data in the historical watermark adding data and parameter data of the added watermark;
respectively calculating a first similarity of second target file data, a second similarity of second link data and a third similarity of parameter data in each historical watermark adding data;
inductive grouping is carried out on the historical watermark adding data based on the first similarity, the second similarity and the third similarity;
determining the number of the historical watermark adding data in each group, and deleting the groups when the number is smaller than a preset number threshold;
the preset rules include: and equally dividing the historical data set into a training set and a testing set according to a preset proportion.
Preferably, the big data watermarking method based on artificial intelligence further includes:
when a request of adding a watermark of a user is received, detecting a first target file corresponding to the request;
determining whether a watermark has been included; when the watermark exists, returning first prompt information; when the watermark does not exist, acquiring first target file data and first environment data corresponding to the request;
detecting a first target file corresponding to the request; the method comprises the following steps:
extracting key data of the first target file, and determining that the watermark exists in the first target file when extracting the key data corresponding to the added watermark;
and/or the presence of a gas in the gas,
sending the first target file to a plurality of detection ends connected with a big data platform;
and determining whether the first target file has the watermark or not based on the data fed back by the detection end.
Preferably, the determining whether the watermark exists in the first target file based on the data fed back by the detecting end includes:
analyzing the data fed back by the detection end based on a preset analysis template, and assigning a result corresponding to the data to obtain a result value;
acquiring a credit coefficient of a detection end;
determining a feedback value based on the credit coefficient and the result value, wherein the calculation formula of the feedback value is as follows:
wherein F is a feedback value;
the credit coefficient of the ith detection end; a. theiThe result value of the ith detection end is obtained; n is the number of feedback detection ends;
and when the feedback value is larger than a preset threshold value, determining that the first target file has the watermark, otherwise, determining that the first target file does not have the watermark.
Preferably, when the feedback of the detection end and the judgment result of the first target file are positive, the previous N times of historical feedback of the detection end are obtained;
judging whether the previous N times of historical feedbacks are all positive directions, and if so, up-regulating the credit coefficient of the detection end; the calculation formula of the credit coefficient after the adjustment is as follows:
h is a credit coefficient after the up-regulation, H is a credit coefficient before the up-regulation, gamma is a preset up-regulation amplitude value, and W is the credit coefficient and value of all detection ends fed back at this time;
when the feedback of the detection end is opposite to the judgment result of the first target file, acquiring the previous N times of historical feedback of the detection end;
constructing a first vector determined by a down-regulation function based on the previous N times of historical feedback;
acquiring a down-regulation amplitude determination library, wherein the second vectors in the down-regulation amplitude determination library correspond to the down-regulation amplitudes one to one;
querying a down-regulation amplitude determination library based on the first vector to determine a down-regulation amplitude;
based on the down-regulation amplitude value, the credit coefficient of the detection end is down-regulated; the calculation formula of the credit coefficient after the down regulation is as follows:
q is the credit coefficient after the down regulation, Q is the credit coefficient before the down regulation, delta is the down regulation amplitude, and W is the credit coefficient sum value of all detection ends fed back at this time.
Preferably, a down-regulation amplitude determination library is queried based on the first vector to determine a down-regulation amplitude; the method comprises the following steps:
and calculating the matching degree between the first vector and each second vector, wherein the calculation formula is as follows:
wherein, P is the matching degree of the first vector and the second vector; x is the number ofkA data value of a k-th dimension of the first vector; y iskIs a data value of the kth dimension of the second vector; m is the total number of data dimensions; c is a preset constant;
and acquiring a down-regulation amplitude value corresponding to the second vector with the maximum matching degree.
Preferably, the big data watermarking method based on artificial intelligence further includes:
when a request of adding a watermark of a user is received, acquiring a questionnaire answer of a watermark preference questionnaire filled in when the user registers;
a pre-set watermark library is obtained and,
analyzing the answers of the questionnaire to obtain the preference of the user;
parameters of the watermark are retrieved from a watermark repository based on the preferences.
Preferably, the parameters include: the character of the watermark, the color of the font, the size of the font, the angle of the font and the definition of the font are combined.
Preferably, the first environment data includes: one or more of account ID, equipment ID, login position and software running environment.
Preferably, the big data watermarking method based on artificial intelligence further includes:
when a watermark adding request of a user is received, screening out first data related to user preference from a big data platform;
a pre-set watermark library is obtained and,
analyzing the first data to obtain the preference of the user;
parameters of the watermark are retrieved from a watermark repository based on the preferences.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
The technical solution of the present invention is further described in detail by the accompanying drawings and embodiments.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention. In the drawings:
fig. 1 is a schematic diagram of a big data watermarking method based on artificial intelligence in an embodiment of the present invention.
Detailed Description
The preferred embodiments of the present invention will be described in conjunction with the accompanying drawings, and it will be understood that they are described herein for the purpose of illustration and explanation and not limitation.
The embodiment of the invention provides a big data watermarking method based on artificial intelligence, as shown in figure 1, comprising the following steps:
step S1: when a request of adding a watermark of a user is received, acquiring first target file data and first environment data corresponding to the request;
step S2: performing feature extraction on the first target file data and the first environment data to obtain a plurality of feature values;
step S3: inputting the characteristic value into a neural network model trained in advance to obtain the parameter of the watermark;
step S4: based on the parameters, watermarking is performed on the target file.
The working principle and the beneficial effects of the technical scheme are as follows:
the neural network model is applied to analyzing the target file added with the watermark and the added environmental data, the requirements of the user are grasped, so that watermark parameters meeting the requirements of the user are selected, the watermark adding is set based on the watermark parameters, and the watermark adding of the target file is further realized.
The big data watermarking method based on artificial intelligence analyzes the historical added data of the user based on artificial intelligence so as to realize intelligent watermarking selection and improve the watermarking adding efficiency of the user. And batch different watermark adding operation can be realized through the method.
In one embodiment, the neural network model is generated by training as follows:
acquiring historical watermark adding data of a user from a big data platform based on user parameters;
preprocessing the historical watermark adding data to obtain a historical data set;
dividing a historical data set into a training set and a testing set based on a preset rule;
training the initial second neural network model based on the training set and the test set, and taking the initial second neural network model as the neural network model after the initial second neural network model converges;
preprocessing the historical watermark adding data to obtain a historical data set, wherein the preprocessing the historical watermark adding data to obtain the historical data set comprises the following steps:
analyzing the historical watermark adding data to obtain second target file data and second environment data in the historical watermark adding data and parameter data of the added watermark;
respectively calculating a first similarity of second target file data, a second similarity of second link data and a third similarity of parameter data in each historical watermark adding data;
inductive grouping is carried out on the historical watermark adding data based on the first similarity, the second similarity and the third similarity;
determining the number of the historical watermark adding data in each group, and deleting the groups when the number is smaller than a preset number threshold;
the preset rules include: and equally dividing the historical data set into a training set and a testing set according to a preset proportion.
The working principle and the beneficial effects of the technical scheme are as follows:
the neural network model is trained in advance, the embodiment provides a method for acquiring training data trained in advance, and historical watermark adding data of a user is acquired from a big data platform through user parameters; the user parameters include user name, ID and other information characterizing the user identity. After the data are obtained, preprocessing operations such as abnormal data elimination, duplicate removal, data supplement and the like are required to be removed, so that the data meet the requirements of training data; in addition, the preprocessing further includes: and (4) screening. Adding second target file data, second environment data and added watermark parameter data in data according to historical watermarks, and adding a first similarity corresponding to the second target file data, a second similarity corresponding to the second link data and a third similarity corresponding to the parameter data according to each historical data; when the first similarity, the second similarity and the third similarity are respectively larger than the preset corresponding threshold values, the data are subjected to induction grouping, and then the groups smaller than the preset number threshold value in the groups are deleted to remove the data which do not have representativeness, so that the accuracy of the trained neural network model is improved. Wherein, the preset rule comprises: equally dividing a historical data set into a training set and a testing set according to a preset proportion; for example, 5:1 classification into a training set test set is adopted.
In one embodiment, the artificial intelligence based big data watermarking method further comprises:
when a request of adding a watermark of a user is received, detecting a first target file corresponding to the request;
determining whether a watermark has been included; when the watermark exists, returning first prompt information; when the watermark does not exist, acquiring first target file data and first environment data corresponding to the request;
detecting a first target file corresponding to the request; the method comprises the following steps:
extracting key data of the first target file, and determining that the watermark exists in the first target file when extracting the key data corresponding to the added watermark;
and/or the presence of a gas in the gas,
sending the first target file to a plurality of detection ends connected with a big data platform;
and determining whether the first target file has the watermark or not based on the data fed back by the detection end.
The working principle and the beneficial effects of the technical scheme are as follows:
before adding the watermark, whether the target file is added with the watermark or not needs to be determined; the watermarked file does not need to be added twice. When adding the watermark, a key datum is embedded as an added mark of the watermark; when the key data exists in the data of the file, the watermark can be determined to be added. After the critical data is searched, the first target file is sent to a big data platform, whether the watermark exists is determined through multi-party verification, and the multi-party verification is mainly applied to the scene of adding the watermark of the file in batches, and a user cannot determine the watermark one by one, so that the multi-party verification is needed, and the efficiency and the accuracy of determining the watermark are improved.
In one embodiment, the determining whether the watermark exists in the first target file based on the data fed back by the detection end comprises:
analyzing the data fed back by the detection end based on a preset analysis template, and assigning a result corresponding to the data to obtain a result value;
acquiring a credit coefficient of a detection end;
determining a feedback value based on the credit coefficient and the result value, wherein the calculation formula of the feedback value is as follows:
wherein F is a feedback value;
the credit coefficient of the ith detection end; a. theiAs a result of the ith detection terminalA value; n is the number of feedback detection ends;
and when the feedback value is larger than a preset threshold value, determining that the first target file has the watermark, otherwise, determining that the first target file does not have the watermark.
The working principle and the beneficial effects of the technical scheme are as follows:
and judging whether the first target file has the watermark or not based on the credit coefficient of the detection end and the result value of the identification according to the feedback data. The credit coefficient is initially configured for the platform to the detection end, and fine tuning is performed according to the detection process of the detection end. When the detection end feeds back, and the result value is 1, the first target has a watermark; when the result value is-1, the first target has no watermark; the preset threshold value can be determined according to the number of the sent detection ends and the comparison table of the number of the detection ends and the threshold value; the number of the detection ends and the threshold value comparison table are set according to experience.
In one embodiment, when the feedback of the detection end and the judgment result of the first target file are positive, the previous N times of historical feedback of the detection end are obtained;
judging whether the previous N times of historical feedbacks are all positive directions, and if so, up-regulating the credit coefficient of the detection end; the calculation formula of the credit coefficient after the adjustment is as follows:
h is a credit coefficient after the up-regulation, H is a credit coefficient before the up-regulation, gamma is a preset up-regulation amplitude value, and W is the credit coefficient and value of all detection ends fed back at this time;
when the feedback of the detection end is opposite to the judgment result of the first target file, acquiring the previous N times of historical feedback of the detection end;
constructing a first vector determined by a down-regulation function based on the previous N times of historical feedback;
acquiring a down-regulation amplitude determination library, wherein the second vectors in the down-regulation amplitude determination library correspond to the down-regulation amplitudes one to one;
querying a down-regulation amplitude determination library based on the first vector to determine a down-regulation amplitude;
based on the down-regulation amplitude value, the credit coefficient of the detection end is down-regulated; the calculation formula of the credit coefficient after the down regulation is as follows:
q is the credit coefficient after the down regulation, Q is the credit coefficient before the down regulation, delta is the down regulation amplitude, and W is the credit coefficient sum value of all detection ends fed back at this time.
Querying a down-regulation amplitude determination library based on the first vector to determine a down-regulation amplitude; the method comprises the following steps:
and calculating the matching degree between the first vector and each second vector, wherein the calculation formula is as follows:
wherein, P is the matching degree of the first vector and the second vector; x is the number ofkA data value of a k-th dimension of the first vector; y iskIs a data value of the kth dimension of the second vector; m is the total number of data dimensions; c is a preset constant;
and acquiring a down-regulation amplitude value corresponding to the second vector with the maximum matching degree.
The working principle and the beneficial effects of the technical scheme are as follows:
adjusting the credit coefficient of the detection end according to the accuracy of the feedback of the detection end; so as to ensure the accuracy of the feedback of the next detection end. In order to optimize the adjustment scheme, the up-regulation amplitude value is larger than the down-regulation amplitude value; when the down-regulation amplitude is determined according to the down-regulation amplitude determination library, in N times of historical feedback of a user, if the feedback error ratio is larger, the down-regulation amplitude is larger; the more the continuous times of continuous feedback errors are, the larger the down-regulation amplitude is; the more recent historical feedback has a greater effect on the current turndown magnitude.
In one embodiment, the artificial intelligence based big data watermarking method further comprises:
when a request of adding a watermark of a user is received, acquiring a questionnaire answer of a watermark preference questionnaire filled in when the user registers;
a pre-set watermark library is obtained and,
analyzing the answers of the questionnaire to obtain the preference of the user;
parameters of the watermark are retrieved from a watermark repository based on the preferences.
The working principle and the beneficial effects of the technical scheme are as follows:
the user preference is mastered through the questionnaire, so that the watermark corresponding to the preference is selected for the user, and the individual requirements of the user are met.
In one embodiment, the parameters include: the character of the watermark, the color of the font, the size of the font, the angle of the font and the definition of the font are combined.
The location where the embedding of the watermark is achieved by the parameters and the determined location of the watermark for the respective parameters, in one embodiment the first environment data comprises: one or more of account ID, equipment ID, login position and software running environment.
The precise watermark selection is achieved by the environmental data.
In one embodiment, the artificial intelligence based big data watermarking method further comprises:
when a watermark adding request of a user is received, screening out first data related to user preference from a big data platform;
a pre-set watermark library is obtained and,
analyzing the first data to obtain the preference of the user;
parameters of the watermark are retrieved from a watermark repository based on the preferences.
The working principle and the beneficial effects of the technical scheme are as follows:
the user preference is mastered through the questionnaire, so that the watermark corresponding to the preference is selected for the user, and the individual requirements of the user are met.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.
Claims (10)
1. A big data watermarking method based on artificial intelligence is characterized by comprising the following steps:
when a request of adding a watermark of a user is received, acquiring first target file data and first environment data corresponding to the request;
performing feature extraction on the first target file data and the first environment data to obtain a plurality of feature values;
inputting the characteristic value into a neural network model trained in advance to obtain parameters of the watermark;
and based on the parameters, carrying out watermarking on the target file.
2. The artificial intelligence based big data watermarking method of claim 1, wherein the neural network model is generated by training through the following steps:
acquiring historical watermarking data of the user from a big data platform based on user parameters;
preprocessing the historical watermark adding data to obtain a historical data set;
dividing the historical data set into a training set and a testing set based on a preset rule;
training an initial second neural network model based on the training set and the test set, and taking the initial second neural network model as the neural network model after the initial second neural network model converges;
wherein, the preprocessing the historical watermark adding data to obtain a historical data set comprises:
analyzing the historical watermark adding data to obtain second target file data and second environment data in the historical watermark adding data and parameter data of the added watermark;
respectively calculating a first similarity of second target file data, a second similarity of second link data and a third similarity of the parameter data in the historical watermark adding data;
inductive grouping of the historical watermarking data based on the first similarity, the second similarity and the third similarity;
determining the quantity of the historical watermark adding data in each group, and deleting the groups when the quantity is smaller than a preset quantity threshold value;
the preset rules include: and equally dividing the historical data set into the training set and the test set according to a preset proportion.
3. The artificial intelligence based big data watermarking method of claim 1, further comprising:
when a request of adding a watermark of a user is received, detecting the first target file corresponding to the request;
determining whether a watermark has been included; when the watermark exists, returning first prompt information; when the watermark does not exist, acquiring the first target file data and the first environment data corresponding to the request;
detecting the first target file corresponding to the request; the method comprises the following steps:
extracting key data of the first target file, and determining that the watermark exists in the first target file when extracting the key data corresponding to the added watermark;
and/or the presence of a gas in the gas,
sending the first target file to a plurality of detection ends connected with a big data platform;
and determining whether the first target file has the watermark or not based on the data fed back by the detection end.
4. The big data watermarking method based on artificial intelligence, wherein the determining whether the watermark exists in the first target file based on the data fed back by the detection end comprises:
analyzing the data fed back by the detection end based on a preset analysis template, and assigning a result corresponding to the data to obtain a result value;
acquiring a credit coefficient of the detection end;
determining a feedback value based on the credit coefficient and the result value, wherein the calculation formula of the feedback value is as follows:
wherein F is the feedback value;
the credit coefficient of the ith detection end; a. theiThe result value of the ith detection end is obtained; n is the number of the feedback detection ends;
and when the feedback value is larger than a preset threshold value, determining that the first target file has the watermark, otherwise, determining that the first target file does not have the watermark.
5. The big data watermarking method based on artificial intelligence as claimed in claim 4, wherein when the feedback of the detection end and the judgment result of the first target file are positive, the previous N times of historical feedback of the detection end are obtained;
judging whether the previous N times of historical feedbacks are all positive directions, and if so, up-regulating the credit coefficient of the detection end; the calculation formula of the credit coefficient after the adjustment is as follows:
h is a credit coefficient after the up-regulation, H is a credit coefficient before the up-regulation, gamma is a preset up-regulation amplitude value, and W is the credit coefficient and value of all detection ends fed back at this time;
when the feedback of the detection end is opposite to the judgment result of the first target file, obtaining the previous N times of historical feedback of the detection end;
constructing a first vector determined by a down-regulation function based on the previous N times of historical feedback;
acquiring a down-regulation amplitude determination library, wherein second vectors in the down-regulation amplitude determination library correspond to down-regulation amplitudes one to one;
querying the down regulation amplitude determination library based on the first vector to determine the down regulation amplitude;
based on the down-regulation amplitude value, the credit coefficient of the detection end is down-regulated; the calculation formula of the credit coefficient after the down regulation is as follows:
q is the credit coefficient after the down regulation, Q is the credit coefficient before the down regulation, delta is the down regulation amplitude, and W is the credit coefficient sum value of all detection ends fed back at this time.
6. The artificial intelligence based big data watermarking method of claim 5, wherein the down-regulation magnitude is determined by querying the down-regulation magnitude determination library based on the first vector; the method comprises the following steps:
calculating the matching degree between the first vector and each second vector, wherein the calculation formula is as follows:
wherein P is the degree of match of the first vector and the second vector; x is the number ofkA data value of a k-th dimension of the first vector; y iskA data value of a k-th dimension of the second vector; m is the total number of data dimensions; c is a preset constant;
and acquiring the down-regulation amplitude value corresponding to the second vector with the maximum matching degree.
7. The artificial intelligence based big data watermarking method of claim 1, further comprising:
when a request of adding a watermark of a user is received, acquiring a questionnaire answer of a watermark preference questionnaire filled in when the user registers;
a pre-set watermark library is obtained and,
analyzing the questionnaire answers to obtain the preferences of the user;
retrieving parameters of a watermark from the watermark library based on the preferences.
8. The artificial intelligence based big data watermarking method of claim 1, wherein the parameters comprise: the character of the watermark, the color of the font, the size of the font, the angle of the font and the definition of the font are combined.
9. The artificial intelligence based big data watermarking method of claim 1, wherein the first environment data comprises: one or more of account ID, equipment ID, login position and software running environment.
10. The artificial intelligence based big data watermarking method of claim 1, further comprising:
when a watermark adding request of a user is received, screening out first data related to user preference from a big data platform;
a pre-set watermark library is obtained and,
analyzing the first data to obtain the preference of the user;
retrieving parameters of a watermark from the watermark library based on the preferences.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110495216.XA CN113468486B (en) | 2021-05-07 | 2021-05-07 | Big data watermarking method based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110495216.XA CN113468486B (en) | 2021-05-07 | 2021-05-07 | Big data watermarking method based on artificial intelligence |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113468486A true CN113468486A (en) | 2021-10-01 |
| CN113468486B CN113468486B (en) | 2023-12-08 |
Family
ID=77870684
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110495216.XA Active CN113468486B (en) | 2021-05-07 | 2021-05-07 | Big data watermarking method based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113468486B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109993300A (en) * | 2017-12-29 | 2019-07-09 | 华为技术有限公司 | A kind of training method and device of neural network model |
| CN110781463A (en) * | 2019-10-29 | 2020-02-11 | 北京中电普华信息技术有限公司 | Method and device for adding watermark information |
| CN112199944A (en) * | 2020-10-10 | 2021-01-08 | 深圳壹账通智能科技有限公司 | Method and device for adding watermark in text, electronic equipment and storage medium |
-
2021
- 2021-05-07 CN CN202110495216.XA patent/CN113468486B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109993300A (en) * | 2017-12-29 | 2019-07-09 | 华为技术有限公司 | A kind of training method and device of neural network model |
| CN110781463A (en) * | 2019-10-29 | 2020-02-11 | 北京中电普华信息技术有限公司 | Method and device for adding watermark information |
| CN112199944A (en) * | 2020-10-10 | 2021-01-08 | 深圳壹账通智能科技有限公司 | Method and device for adding watermark in text, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113468486B (en) | 2023-12-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wu et al. | A novel convolutional neural network for image steganalysis with shared normalization | |
| CN107515895A (en) | A kind of sensation target search method and system based on target detection | |
| CN105844283A (en) | Method for identifying category of image, image search method and image search device | |
| CN109740606B (en) | Image identification method and device | |
| CN104615676B (en) | One kind being based on the matched picture retrieval method of maximum similarity | |
| CN108197668A (en) | The method for building up and cloud system of model data collection | |
| CN113761259A (en) | Image processing method and device and computer equipment | |
| CN112966685B (en) | Attack network training method and device for scene text recognition and related equipment | |
| CN113139536A (en) | Text verification code identification method and equipment based on cross-domain meta learning and storage medium | |
| CN112948823A (en) | Data leakage risk assessment method | |
| CN108875727A (en) | The detection method and device of graph-text identification, storage medium, processor | |
| CN112580108A (en) | Signature and seal integrity verification method and computer equipment | |
| CN114359553A (en) | Signature positioning method and system based on Internet of things and storage medium | |
| CN110059705B (en) | OCR recognition result judgment method and device based on modeling | |
| CN112364974A (en) | Improved YOLOv3 algorithm based on activation function | |
| CN114998080A (en) | Face tamper-proof watermark generation method, tamper detection method and attribute detection method | |
| CN109617864B (en) | Website identification method and website identification system | |
| CN113987429A (en) | Copyright verification method of neural network model based on watermark embedding | |
| CN107103327B (en) | Dyeing counterfeit image detection method based on color statistical difference | |
| CN111858275B (en) | Method, device, equipment and storage medium for identifying abnormal media behaviors | |
| CN113468486B (en) | Big data watermarking method based on artificial intelligence | |
| CN113128504A (en) | OCR recognition result error correction method and device based on verification rule | |
| CN104484869B (en) | Image matching method and system for ordinal measure features | |
| CN111737688A (en) | Attack defense system based on user portrait | |
| CN111581439A (en) | Big data information processing method and system for machine learning analysis |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |