A portal website audit method based on image processing

Abstract

The present invention discloses a portal website audit method based on image processing, which specifically relates to the field of image processing, and includes an audit area division step, an audit content collection step, an audit content analysis step, an abnormal index calculation step, a comprehensive analysis step, and a manual audit step. The present invention divides a target portal website into sub-data areas and numbers them through the audit area division step, collects parameters of each sub-data area through the audit content collection step, calculates the flow change coefficient, page content change coefficient, and product sales change coefficient of each sub-data area through the audit content analysis step, calculates the abnormal index of each sub-data area through the abnormal index calculation step, calculates the comprehensive abnormal index of the target portal website through the comprehensive analysis step and audits it, and manually audits the portal website that needs manual audit through the manual audit step.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and more specifically, to a portal website auditing method based on image processing. Background Art

[0002] With the global development of the Internet, it has had a profound impact on people's lives and production methods. Portal websites initially provided search engines and directory services. Later, due to increasingly fierce market competition, portal websites had to quickly expand into various new business types, hoping to attract and retain Internet users through a wide range of business categories. As a result, portal websites' business is all-encompassing, becoming the "department store" or "online supermarket" of the online world.

[0003] The current portal website review methods mainly include manual review, which is to employ professional content reviewers or establish review teams to review the published content one by one. They will remove some complex or gray content in the review process based on the website's regulations and the requirements of laws and regulations. At the same time, in order to ensure the accuracy and legality of the review, reviewers will receive regular training and update their judgment standards to cope with the ever-changing content formats and review requirements.

[0004] However, there are still some shortcomings in its actual use. For example, manual review requires a lot of human resources and time costs, especially when the amount of content is huge, the review efficiency may be affected. Although technical review can improve efficiency, due to the limitations of algorithms and models, misjudgments and omissions may occur. Different reviewers may have different judgment criteria for the same content, which may lead to inconsistency and unfairness in the review results. With the development of network technology, new content forms are constantly emerging, and existing review methods may not be able to cope with these new forms of content and need to be constantly updated and improved. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a portal website audit method based on image processing, which solves the problems raised in the above-mentioned background technology through the following scheme.

[0006] To achieve the above purpose, the present invention provides the following technical solution: a portal website audit method based on image processing, comprising:

[0007] Step S1: Audit area division step: used to determine the audit content of the target portal website as the target data area, divide the target data area into various sub-data areas by equal time division, and mark them as 1, 2...n in sequence.

[0008] Step S2: Audit content collection step: used to collect pictures of the target portal website and collect website traffic parameters, page content parameters and product sales parameters of each sub-data area through image processing technology, and transmit them to the audit content analysis step.

[0009] Step S3: Audit content analysis step: including network traffic analysis unit, page content analysis unit and product sales analysis unit. Each analysis unit is used to establish an analysis model. The parameters transmitted in the audit content collection step are imported into the analysis model to calculate the traffic change coefficient, page content change coefficient and product sales change coefficient of each sub-data area, and transmit them to the abnormal index calculation step.

[0010] Step S4: Abnormal index calculation step: used to establish an abnormal index calculation model, import the analysis results transmitted by the audit content analysis step into the abnormal index calculation model, calculate the abnormal index of each sub-data area, and transmit it to the comprehensive analysis step.

[0011] Step S5: Comprehensive analysis step: used to establish a comprehensive analysis model, import the abnormal index of each sub-data area transmitted by each abnormal index calculation step into the comprehensive analysis model, calculate the comprehensive abnormal index of the target portal website, review the comprehensive abnormal index of the target portal website through the comprehensive abnormal index security threshold, and transmit the review result to the manual review step.

[0012] Step S6: Manual review step: used to receive the portal websites that need manual review status transmitted in the comprehensive analysis step, and perform manual review on the portal websites that need manual review status.

[0013] Preferably, the audit content collection step sets an image tag on each page of the target portal website through image processing technology. When a visitor visits the page, the image tag will collect and record the visitor's behavior data, and collect website traffic data of each sub-data area by analyzing the visitor's mouse movement trajectory or keyboard strokes on the page. By uploading the page content picture of the target portal website to an optical character recognition tool, the optical character recognition tool will automatically recognize the text in the picture and convert it into digital information to collect page content parameters. By setting an image tag on the product picture of the target portal website, when a visitor visits the page, the image tag will automatically record the product information, and collect product sales parameters by recognizing the picture numbers through the optical character recognition tool.

[0014] Preferably, the website traffic parameters include the number of visits, the number of unique visitors, the bounce rate and the conversion rate, which are marked as Lf i , Ld i , Lt i and Lh i,Page content parameters include content length, content update frequency, number of internal links, and keyword density, which are marked as Yc i , Yp i 、Ys i and Ym i , product sales parameters include sales, order volume, return volume, and repeat purchase rate, which are marked as Xe i ,Xb i ,Xq i and Xu i , where i=1, 2,...n, i represents the i-th sub-data area.

[0015] Preferably, the network traffic analysis unit is used to establish a network traffic analysis model, and the network traffic parameters are imported into the network traffic analysis model to obtain: L i represents the flow change coefficient of the ith sub-data area, Lf i Indicates the number of accesses to the ith sub-data area, Lf max Indicates the maximum number of visits to the target portal website, Ld i Indicates the number of unique visitors in the ith sub-data area, Lt i Indicates the bounce rate of the ith sub-data area, Lh i represents the conversion rate of the ith sub-data area, and μ1 represents other influencing factors of the flow variation coefficient.

[0016] Preferably, the page content analysis unit is used to establish a page content analysis model, and the page content parameters are imported into the page content analysis model to obtain: Y i represents the page content change coefficient of the ith sub-data area, Yc i Indicates the content length of the i-th sub-data area, Yp i Indicates the content update frequency of the ith sub-data area, Yp max Indicates the maximum content update frequency of the target portal website, Ys i Indicates the number of internal links in the ith sub-data area, Ym i represents the keyword density of the ith sub-data area, and μ2 represents other influencing factors of the page content variation coefficient.

[0017] Preferably, the product sales analysis unit is used to establish a product sales analysis model, and import product sales parameters into the product sales analysis model to obtain: X i represents the product sales change coefficient of the ith sub-data area, Xe i represents the sales volume of the ith sub-data area, Xb i represents the order volume of the ith sub-data area, Xqi represents the return quantity of the ith sub-data area, Xb max represents the maximum order quantity of the target portal website, Xq min represents the minimum return quantity of the target portal website, Xu i represents the repeat purchase rate of the ith sub-data area, and μ3 represents other influencing factors of the product sales variation coefficient.

[0018] Preferably, the abnormal index calculation model is: η i represents the abnormal index of the ith sub-data region, L i represents the flow change coefficient of the ith sub-data area, Y i represents the page content change coefficient of the i-th sub-data area, X i It represents the product sales change coefficient of the i-th sub-data area, L0 represents the preset safety value of the traffic change coefficient, Y0 represents the preset safety value of the page content change coefficient, X0 represents the preset safety value of the product sales change, and other influencing factors of the λ anomaly index.

[0019] Preferably, the comprehensive analysis model is: η represents the comprehensive anomaly index of the target portal website, η i It represents the anomaly index of the ith sub-data area. The comprehensive anomaly index safety threshold is: Indicates the maximum safety threshold of the comprehensive anomaly index, Represents the minimum safety threshold of the comprehensive abnormality index. When η ν+ >η>η ν- When η ν+ ≤η or when η≤η ν- When an abnormal situation occurs on the target portal website, the target portal website is marked as requiring manual review.

[0020] Technical effects and advantages of the present invention:

[0021] The present invention uses image processing technology to quickly identify the content of images, reduce the burden of manual review, and improve review efficiency. The present invention can reduce misjudgments and missed judgments caused by human factors and improve the accuracy of review. The present invention uses image processing methods to reduce the need for manual review of a large number of images, reduce labor costs, enable portal websites to be managed more efficiently, and save human resources. The present invention can promote technological innovation and promote the continuous development and improvement of portal website review technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a schematic diagram of the overall structure of the present invention. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0024] As attached Figure 1 The portal website audit method based on image processing shown includes an audit area division step, an audit content collection step, an audit content analysis step, an abnormal index calculation step, a comprehensive analysis step and a manual audit step.

[0025] Step S1: Audit area division step: used to determine the audit content of the target portal website as the target data area, divide the target data area into various sub-data areas by equal time division, and mark them as 1, 2...n in sequence.

[0026] Step S2: Audit content collection step: used to collect pictures of the target portal website and collect website traffic parameters, page content parameters and product sales parameters of each sub-data area through image processing technology, and transmit them to the audit content analysis step.

[0027] The audit content collection step sets an image tag on each page of the target portal website through image processing technology. When a visitor visits the page, the image tag will collect and record the visitor's behavior data, and collect website traffic data for each sub-data area by analyzing the visitor's mouse movement trajectory or keyboard strokes on the page. By uploading the page content image of the target portal website to the optical character recognition tool, the optical character recognition tool will automatically recognize the text in the image and convert it into digital information to collect page content parameters. By setting an image tag on the product image of the target portal website, when a visitor visits the page, the image tag will automatically record the product information, and collect product sales parameters by recognizing the image numbers through the optical character recognition tool.

[0028] The website traffic parameters include the number of visits, the number of unique visitors, the bounce rate and the conversion rate, which are marked as Lf i , Ld i , Lt i and Lh i ,Page content parameters include content length, content update frequency, number of internal links, and keyword density, which are marked as Yc i , Yp i 、Ys i and Ym i, product sales parameters include sales, order volume, return volume, and repeat purchase rate, which are marked as Xe i ,Xb i ,Xq i and Xu i , where i=1, 2,...n, i represents the i-th sub-data area.

[0029] Step S3: Audit content analysis step: including network traffic analysis unit, page content analysis unit and product sales analysis unit. Each analysis unit is used to establish an analysis model. The parameters transmitted in the audit content collection step are imported into the analysis model to calculate the traffic change coefficient, page content change coefficient and product sales change coefficient of each sub-data area, and transmit them to the abnormal index calculation step.

[0030] The network traffic analysis unit is used to establish a network traffic analysis model, and import network traffic parameters into the network traffic analysis model to obtain: L i represents the flow change coefficient of the ith sub-data area, Lf i Indicates the number of accesses to the ith sub-data area, Lf max Indicates the maximum number of visits to the target portal website, Ld i Indicates the number of unique visitors in the ith sub-data area, Lt i Indicates the bounce rate of the ith sub-data area, Lh i represents the conversion rate of the ith sub-data area, and μ1 represents other influencing factors of the flow variation coefficient.

[0031] The page content analysis unit is used to establish a page content analysis model, and import the page content parameters into the page content analysis model to obtain: Y i represents the page content change coefficient of the ith sub-data area, Yc i Indicates the content length of the i-th sub-data area, Yp i Indicates the content update frequency of the ith sub-data area, Yp max Indicates the maximum content update frequency of the target portal website, Ys i Indicates the number of internal links in the ith sub-data area, Ym i represents the keyword density of the ith sub-data area, and μ2 represents other influencing factors of the page content variation coefficient.

[0032] The product sales analysis unit is used to establish a product sales analysis model, and import product sales parameters into the product sales analysis model to obtain: X i represents the product sales change coefficient of the ith sub-data area, Xe irepresents the sales volume of the ith sub-data area, Xb i represents the order volume of the ith sub-data area, Xq i represents the return quantity of the ith sub-data area, Xb max represents the maximum order quantity of the target portal website, Xq min represents the minimum return quantity of the target portal website, Xu i represents the repeat purchase rate of the ith sub-data area, and μ3 represents other influencing factors of the product sales variation coefficient.

[0033] Step S4: Abnormal index calculation step: used to establish an abnormal index calculation model, import the analysis results transmitted by the audit content analysis step into the abnormal index calculation model, calculate the abnormal index of each sub-data area, and transmit it to the comprehensive analysis step.

[0034] The abnormal index calculation model is: η i represents the abnormal index of the ith sub-data region, L i represents the flow change coefficient of the ith sub-data area, Y i represents the page content change coefficient of the i-th sub-data area, X i It represents the product sales change coefficient of the i-th sub-data area, L0 represents the preset safety value of the traffic change coefficient, Y0 represents the preset safety value of the page content change coefficient, X0 represents the preset safety value of the product sales change, and other influencing factors of the λ anomaly index.

[0035] Step S5: Comprehensive analysis step: used to establish a comprehensive analysis model, import the abnormal index of each sub-data area transmitted by each abnormal index calculation step into the comprehensive analysis model, calculate the comprehensive abnormal index of the target portal website, review the comprehensive abnormal index of the target portal website through the comprehensive abnormal index security threshold, and transmit the review result to the manual review step.

[0036] The comprehensive analysis model is: η represents the comprehensive anomaly index of the target portal website, η i represents the anomaly index of the ith sub-data area, and the comprehensive anomaly index safety threshold is: Indicates the maximum safety threshold of the comprehensive anomaly index, Represents the minimum safety threshold of the comprehensive abnormality index. When or when When an abnormal situation occurs on the target portal website, the target portal website is marked as requiring manual review.

[0037] Step S6: Manual review step: used to receive the portal websites that need manual review status transmitted in the comprehensive analysis step, and perform manual review on the portal websites that need manual review status.

[0038] The manual review step is used to receive portal website information to be reviewed, including text, pictures and digital information, and review the portal website information to be reviewed with reference to the content policy, guidelines and laws and regulations of the portal website, and record the decision and reasons of the manual review.

[0039] The present invention divides the target portal website into various sub-data areas and numbers them through an audit area division step, collects website traffic parameters, page content parameters and product sales parameters of each sub-data area through an audit content collection step, calculates the traffic change coefficient, page content change coefficient and product sales change coefficient of each sub-data area through an audit content analysis step, calculates the abnormal index of each sub-data area through an abnormal index calculation step, calculates the comprehensive abnormal index of the target portal website through a comprehensive analysis step and reviews it, and manually reviews the portal website that needs manual review through a manual review step.

[0040] The present invention uses image processing technology to quickly identify the content of images, reduce the burden of manual review, and improve review efficiency. The present invention can reduce misjudgments and missed judgments caused by human factors and improve the accuracy of review. The present invention uses image processing methods to reduce the need for manual review of a large number of images, reduce labor costs, enable portal websites to be managed more efficiently, and save human resources. The present invention can promote technological innovation and promote the continuous development and improvement of portal website review technology.

[0041] Secondly: In the drawings of the embodiments disclosed in the present invention, only the structures related to the embodiments disclosed in the present invention are involved, and other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of the present invention can be combined with each other;

[0042] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A portal website audit method based on image processing, comprising: Step S1: Audit area division step: used to determine the audit content of the target portal website as the target data area, divide the target data area into various sub-data areas by equal time division, and mark them as 1, 2...n in sequence; Step S2: Audit content collection step: used to collect images of the target portal website and collect website traffic parameters, page content parameters and product sales parameters of each sub-data area through image processing technology, and transmit them to the audit content analysis step; The website traffic parameters include the number of visits, the number of unique visitors, the bounce rate and the conversion rate, which are marked as Lf i , Ld i , Lt i and Lh i ,Page content parameters include content length, content update frequency, number of internal links, and keyword density, which are marked as Yc i , Yp i 、Ys i and Ym i , product sales parameters include sales, order volume, return volume, and repeat purchase rate, which are marked as Xe i , Xb i ,Xq i and Xu i , where i=1, 2, ..., n, i represents the i-th sub-data area; Step S3: Audit content analysis step: including a network traffic analysis unit, a page content analysis unit and a product sales analysis unit, each analysis unit is used to establish an analysis model, import the parameters transmitted in the audit content collection step into the analysis model to calculate the traffic change coefficient, page content change coefficient and product sales change coefficient of each sub-data area, and transmit them to the abnormal index calculation step; The network traffic analysis unit is used to establish a network traffic analysis model, and import network traffic parameters into the network traffic analysis model to obtain: L i represents the flow change coefficient of the ith sub-data area, Lf i Indicates the number of accesses to the ith sub-data area, Lf max Indicates the maximum number of visits to the target portal website, Ld i Indicates the number of unique visitors in the ith sub-data area, Lt i Indicates the bounce rate of the ith sub-data area, Lh i represents the conversion rate of the ith sub-data area, μ1 represents other influencing factors of the flow rate variation coefficient; The page content analysis unit is used to establish a page content analysis model, and import the page content parameters into the page content analysis model to obtain: Y i represents the page content change coefficient of the ith sub-data area, Yc i Indicates the content length of the i-th sub-data area, Yp i Indicates the content update frequency of the ith sub-data area, Yp max Indicates the maximum content update frequency of the target portal website, Ys i Indicates the number of internal links in the ith sub-data area, Ym i represents the keyword density of the ith sub-data area, μ2 represents other influencing factors of the page content variation coefficient; The product sales analysis unit is used to establish a product sales analysis model, and import product sales parameters into the product sales analysis model to obtain: X i represents the product sales change coefficient of the ith sub-data area, Xe i represents the sales volume of the ith sub-data area, Xb i represents the order volume of the ith sub-data area, Xq i represents the return quantity of the ith sub-data area, Xb max represents the maximum order quantity of the target portal website, Xq min represents the minimum return quantity of the target portal website, Xu i represents the repeat purchase rate of the ith sub-data area, μ3 represents other influencing factors of the product sales variation coefficient; Step S4: Abnormal index calculation step: used to establish an abnormal index calculation model, import the analysis results transmitted in the audit content analysis step into the abnormal index calculation model, calculate the abnormal index of each sub-data area, and transmit it to the comprehensive analysis step; Step S5: Comprehensive analysis step: used to establish a comprehensive analysis model, import the abnormal index of each sub-data area transmitted in each abnormal index calculation step into the comprehensive analysis model, calculate the comprehensive abnormal index of the target portal website, review the comprehensive abnormal index of the target portal website through the comprehensive abnormal index security threshold, and transmit the review result to the manual review step; Step S6: Manual review step: used to receive the portal websites that need manual review status transmitted in the comprehensive analysis step, and perform manual review on the portal websites that need manual review status.

2. The portal website audit method based on image processing according to claim 1, characterized in that: The audit content collection step sets an image tag on each page of the target portal website through image processing technology. When a visitor visits the page, the image tag will collect and record the visitor's behavior data, and collect website traffic data for each sub-data area by analyzing the visitor's mouse movement trajectory or keyboard strokes on the page. By uploading the page content image of the target portal website to the optical character recognition tool, the optical character recognition tool will automatically recognize the text in the image and convert it into digital information to collect page content parameters. By setting an image tag on the product image of the target portal website, when a visitor visits the page, the image tag will automatically record the product information, and collect product sales parameters by recognizing the image numbers through the optical character recognition tool.

3. The portal website audit method based on image processing according to claim 1, characterized in that: The abnormal index calculation model is: η i represents the abnormal index of the ith sub-data region, L i represents the flow change coefficient of the ith sub-data area, Y i represents the page content change coefficient of the i-th sub-data area, X i It represents the product sales change coefficient of the i-th sub-data area, L0 represents the preset safety value of the traffic change coefficient, Y0 represents the preset safety value of the page content change coefficient, X0 represents the preset safety value of the product sales change, and other influencing factors of the λ anomaly index.

4. The portal website audit method based on image processing according to claim 1, characterized in that: The comprehensive analysis model is: η represents the comprehensive anomaly index of the target portal website, η i represents the anomaly index of the ith sub-data area, and the comprehensive anomaly index safety threshold is: Indicates the maximum safety threshold of the comprehensive anomaly index, Represents the minimum safety threshold of the comprehensive abnormality index. When or when When an abnormal situation occurs on the target portal website, the target portal website is marked as requiring manual review.

Images