CN112700265A - Anti-fraud system and method based on big data processing - Google Patents
Anti-fraud system and method based on big data processing Download PDFInfo
- Publication number
- CN112700265A CN112700265A CN202110304854.9A CN202110304854A CN112700265A CN 112700265 A CN112700265 A CN 112700265A CN 202110304854 A CN202110304854 A CN 202110304854A CN 112700265 A CN112700265 A CN 112700265A
- Authority
- CN
- China
- Prior art keywords
- rate
- comment
- picture
- abnormal
- time period
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/018—Certifying business or products
- G06Q30/0185—Product, service or business identity fraud
Landscapes
- Business, Economics & Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Finance (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses an anti-fraud system and method based on big data processing, which comprises a picture background data storage module, a comment abnormity monitoring module, an abnormal picture detection module, a picture background abnormity detection module, an abnormal volume of transaction detection module, a comment blueprint rate detection module, a comment good evaluation rate detection module, a comment follow-up evaluation rate detection module and a fraud behavior judgment module, and has the beneficial effects that: the method comprises the steps of monitoring comments, comparing an image to be uploaded of the abnormal comments with an uploaded image and a seller display image when the abnormal comments are monitored, detecting the transaction amount of a shop when the image to be uploaded is abnormal, and calculating a fraud evaluation value according to an evaluation value of good rate abnormality, an evaluation value of blueprint rate abnormality and an evaluation value of chasing rate abnormality if the transaction amount is abnormal, so as to judge whether a merchant has fraud, reduce sales of an online shopping platform and guarantee benefits of buyers and other sellers in the same party.
Description
Technical Field
The invention relates to the technical field of computer information, in particular to an anti-fraud system and an anti-fraud method based on big data processing.
Background
The e-commerce refers to the business activities carried out by selling products on the internet, the business activities in real life are transferred to a virtual e-commerce platform for carrying out, the transaction mode is very convenient and fast, the time and space limits are broken, compared with the traditional business form, the e-commerce is undoubtedly a great change, and the e-commerce plays an increasingly important role in the economy of China at present.
At present, in the e-commerce industry, in order to forge good reputation conditions of sellers, including good comments, sales volumes and the like, and in order to obtain more flow entries and activity support on a platform, a bill brushing is performed through various channels, the bill brushing refers to a cheating action taken by a virtual shop in order to obtain a better search ranking of single products or shops, the bill brushing is often matched with a flow brushing and an express delivery of an empty package, the bill brushing can be divided into single product brushing and sales volume preparation and credit brushing to improve the overall reputation of the shop, the mode is generally a sales faking, so that benefits of other goods shops of the same type are damaged, misleading is also caused to buyers, the buyers generally refer to buyer evaluations when buying goods on the virtual shop platform, and therefore, the importance of the evaluation of one shop or one goods is self-evident to one buyer, correspondingly, poor appraisal can influence the overall dynamic scoring and the good appraisal rate of the stores, further influence store ranking, especially for medium and small sellers, poor appraisal can influence buyer judgment to a great extent, and influence conversion rate of commodities, under the background, black product teams consisting of professional good appraisers and professional poor appraisers can be produced, some bad merchants can brush sales volume and good appraisal rate for own stores through the black product teams, and conversely, the bad appraisal is carried out maliciously by competitors, so that the competitiveness of the bad merchants in the same type of commodities is improved, the competitors are pressed, the buyers are misled to consume, the benefits of the buyers and the competitors are damaged in a sales counterfeiting manner, and malignant competition in the industry is formed.
Based on the above problems, it is highly desirable to provide an anti-fraud system and method based on big data processing, wherein comments are monitored, when abnormal comments are monitored, uploaded pictures of all buyers and displayed pictures of corresponding sellers under a comment plate to which the abnormal comments belong are obtained, whether the uploaded pictures are abnormal or not is determined by comparing the uploaded pictures with the uploaded pictures and the displayed pictures of the abnormal comments, the transaction amount of shops within a certain period of time is detected, when the transaction amount is abnormal, a fraud evaluation value is calculated according to an abnormal evaluation value of a good evaluation rate, an abnormal evaluation value of a blueprint rate and an abnormal evaluation value of a chasing evaluation rate, and whether fraud behaviors exist in merchants is determined, so that the sales of online shopping platforms are reduced, and the benefits of the buyers and other sellers in the same party are guaranteed.
Disclosure of Invention
The present invention is directed to a system and method for anti-fraud based on big data processing, so as to solve the above problems in the background art.
In order to solve the technical problems, the invention provides the following technical scheme:
an anti-fraud system based on big data processing comprises a picture background data storage module, a comment abnormity monitoring module, an abnormal picture detection module, a picture background abnormity detection module, an abnormal volume of transaction detection module, a comment blueprint rate detection module, a comment good rate detection module, a comment rate of review detection module and a fraud behavior determination module, wherein the picture background data storage module is pre-stored with commodity picture background data of a virtual shop, the comment abnormity monitoring module is used for monitoring whether user comments are abnormal in real time, the abnormal picture detection module is used for detecting whether pictures uploaded on comment plates by users are abnormal, the picture background abnormity detection module is used for carrying out background detection on uploaded pictures detected to be abnormal, the abnormal volume of transaction detection module is used for acquiring the volume of transaction of commodities of any shop within a certain period of time, and monitoring the abnormal volume of bargaining in a certain period of time according to the volume of bargaining of the commodities, wherein the comment blueprint rate abnormal detection module is used for detecting whether the blueprint rate of a comment plate under any commodity in any shop is abnormal or not and calculating the blueprint rate abnormal evaluation valueThe comment goodness detection module is used for detecting whether the goodness of comment of a comment plate under any commodity in any shop is abnormal or not and calculating the goodness of comment abnormal evaluation valueThe comment follow-up rate detection module is used for detecting whether the follow-up rate of the comment plate under any commodity in any shop is abnormal or not and calculating the abnormal follow-up rate evaluation valueThe fraud determination module estimates the value of the fraud rate anomaly according to the blueprint rateEvaluation value of abnormality of favorable rateEvaluation value for evaluation rate abnormalityA determination is made of fraudulent behavior.
Further, the comment abnormity monitoring module monitors comment abnormity when a user comments on any commodity comment plate through the mobile terminal at the current moment, the comment abnormity monitoring module acquires the current comment state, the comment state is good comment, medium comment or poor comment, and if the current comment is good comment, the comment abnormity monitoring module acquires the number of words of comment of the userAnd number of uploaded picturesWhen the number of words of the comment is greater than or equal to a first preset value and the number of pictures is greater than or equal to a second preset value, calculating a comment abnormity evaluation value according to the number of words of the comment and the number of uploaded picturesWherein, in the step (A),as a function of the number of the coefficients,in order to be at the first preset value,is the second presetThe method includes the steps that when abnormal comments are monitored, whether the current comments are good comments or not needs to be considered, the first step of detection of the malicious good comments is aimed at, the general malicious good comments have the characteristics of large sections of characters and multiple images, whether the current comments are abnormal or not is monitored according to the quantity of the comment characters and the quantity of the images, a preset value is set firstly, when the quantity of the comment characters and the quantity of the uploaded images are larger than the preset value, an abnormal comment evaluation value is calculated, the abnormal comment evaluation value can reflect the abnormal degree of the comments, further, the next step of judgment is carried out according to the abnormal degree, and when the abnormal comment evaluation value is larger than or equal to a comment abnormal evaluation value threshold value, abnormal detection is carried out on the images uploaded by a user under a current comment plate through an abnormal image detection module.
Further, the abnormal picture detection module acquires all buyer uploaded pictures and seller displayed pictures of any one commodity under the comment plate of the commodity, wherein the buyer uploaded pictures are first reference pictures, the seller displayed pictures are second reference pictures, the abnormal picture detection module extracts figures and object outlines in the first reference pictures and the second reference pictures, and further extracts figures and object outlines in the pictures to be detected uploaded by the current user under the comment plate, the abnormal picture detection module compares the figures and the object outlines in the pictures to be detected with the figures and the object outlines in the first reference pictures and the second reference pictures respectively, and the picture abnormal detection module adjusts the size of the pictures to be detected to be consistent with the first reference pictures or the second reference pictures during each comparison, superposing the picture to be detected with the first reference picture or the second reference picture, further obtaining the superposition proportion of the character of the picture to be detected with the first reference picture or the second reference picture and the character of the object outline part occupying the first reference picture or the second reference picture and the character of the object outline part, when the superposition proportion is more than or equal to the superposition proportion threshold value, further carrying out abnormity detection on the picture background in the picture to be detected through a picture background abnormity detection module, wherein the commonly and maliciously brushed and commented comments are generally consistent with the comment pictures of different users, or the comment pictures are consistent with the display pictures of the seller, because the brushing list can involve sending an empty packet, namely the seller can not really post the goods, the brushing list buyer can not receive the goods, so the brushing list pictures are provided for the seller, or the brushing list buyer directly transmits the comment to the seller display pictures, therefore, the picture to be detected, namely the picture to be uploaded, can be compared with the picture uploaded by the buyer and the picture displayed by the seller, whether the same picture exists or not can be judged, and if a plurality of same pictures exist, the probability of malicious and good comment is improved.
Further, the picture background abnormality detection module is connected to the picture background data storage module, and the picture background abnormality detection module further acquires a picture background in the picture to be detected, a picture background of an uploaded picture by a buyer, and commodity picture background data of a corresponding virtual shop pre-stored in the picture background data storage module, wherein the picture background of the uploaded picture by the buyer and the commodity picture background are first picture backgrounds, the picture background in the picture to be detected is a second picture background, the picture background abnormality detection module compares the second picture backgrounds with the first picture backgrounds one by one, detects whether the second picture background is consistent with any one of the first picture backgrounds, if only the outlines of people and articles in the picture are detected, a misjudgment situation may occur, and further aims at the picture backgrounds, and judging whether the background of the picture to be uploaded is consistent with the background of the picture displayed by the merchant and the picture uploaded by the buyer.
Further, if the picture background abnormality detection module detects that the second picture background is consistent with any one of the first picture backgrounds, the abnormality volume of transaction detection module further detects whether the volume of transaction of the order of any commodity in a certain time period of the corresponding store is abnormal, the abnormality volume of transaction detection module obtains the volume of transaction of the order in the certain time period of the corresponding store, divides the certain time period into a plurality of time periods, the time period closest to the current time is a first time period, the time period closest to the first time period is a second time period, further obtains the volume of transaction of the order in other time periods except the first time period, and calculates the volume of transaction of other time periods except the first time periodThe increase rate of the order volume of every two adjacent periods of the period is further calculated according to the increase rate of the order volume of the adjacent periods and the number of the periods except the first periodAnd calculating the increase rate of the order volume from the second time interval to the first time intervalAccording to the average growth rate of the volume of trafficRate of increase of volume of transactions with ordersCalculating traffic anomaly evaluation valueAs a traffic anomaly evaluation valueWhen the volume of the commodity is larger than or equal to the volume anomaly evaluation value threshold value, the volume of the commodity corresponding to the commodity in the certain time period is anomalous, the blueprint rate, the goodness of evaluation and the evaluation rate of the evaluation plate under the corresponding commodity are respectively detected by the comment blueprint rate detection module, the comment goodness of evaluation detection module and the comment evaluation rate detection module, when the picture is anomalous, the volume of the commodity data in the certain time period of the corresponding merchant is further obtained, and whether the volume of the commodity in the two recent time periods is anomalous is judged by combining the previous volume of the commodity increase rate average value, when the merchant conducts recent bill brushing, the volume of the commodity is most intuitively reflected, but whether the single-line brushing of the merchant is unreasonable or not is judged only according to the volume anomaly, the fraud behavior is possibly unreasonable, the misjudgment is likely to occur, the increase speed of the volume of the commodity is also likely to be the propaganda of the merchant in place, therefore, it is necessary to detect the traffic abnormality instead of the traffic abnormality as a judgment basisWhen the volume of the transaction is abnormally increased, the favorable rating, the blueprint rate and the chasing rate of the order comments increased in the period are further detected, so that whether the merchant has fraud or not can be better judged.
Further, the comment goodness-of-comment detection module acquires comment plates of commodities with abnormal transaction volume corresponding to stores, further acquires the total number of comments and the goodness-of-comment quantity in a certain time period, divides the certain time period into a plurality of time periods, calculates the goodness-of-comment rate of the corresponding commodities in the plurality of time periods according to the proportion of the goodness-of-comment quantity in each time period to the total number of the comments, calculates the goodness-of-comment rate increase rate of every two adjacent time periods except the first time period, and further calculates the average goodness-of-comment rate increase rate according to the goodness-of-comment rate increase rate of the adjacent time periods and the number of the time periods except the first time periodAnd calculating the rate of increase of the goodness of the second time interval to the first time intervalAverage growth rate according to good scoreAnd rate of increase of goodnessCalculating a goodness-of-evaluation anomaly evaluation valueThe average increase rate of the future favorable evaluation rate is calculated by combining the data of the future favorable evaluation rate, the past sales condition of the commodity can be visually known according to the average increase rate, the product quality is known, on the basis of the average increase rate of the future favorable evaluation rate, if the recent sudden increase rate is sharply increased compared with the average increase rate of the future favorable evaluation rate, the recent sudden increase rate does not accord with the increase rule of the future evaluation rate, the action of refreshing the bill is generally carried out in a short time, and the division of the time period is carried outIn the observation period, when the comment of the current user is monitored to be abnormal, the high rating rate increase rate of the time period nearest to the current time is most valuable to reference, because the current time when the comment is detected to be abnormal and the nearest time period are most probably in the same billing period, the high rating rate increase rate is combined to provide reference for whether the recent high rating rate increase is abnormal or not, the high rating rate increase rate is increased sharply, the fact that the deal amount may be abnormal is reflected, and the high rating rate increase abnormality and the deal amount increase abnormality are related.
Further, the comment blueprint rate detection module acquires comment plates of commodities with abnormal volume of bargaining corresponding stores, further acquires the total number of comments and the number of blueprint comments in a certain time period, divides the certain time period into a plurality of time periods, calculates the blueprint rate of the corresponding commodities in the plurality of time periods according to the proportion of the number of blueprint comments in each time period to the total number of comments, calculates the blueprint rate increase rate of every two adjacent time periods except the first time period, and further calculates the average blueprint rate increase rate according to the blueprint rate increase rate of the adjacent time periods and the number of time periods except the first time periodAnd calculating the blueprint rate increase rate from the second time interval to the first time intervalAverage growth rate according to blueprint rateAnd the increase rate of blueprint rateCalculating an abnormal evaluation value of blueprint rate。
Further, the comment rate detection module acquires the occurrence rate of the corresponding shopThe method comprises the steps of obtaining the total number of comments and the number of review comments in a certain time period, dividing the certain time period into a plurality of time periods, calculating the review rate of corresponding commodities in the plurality of time periods according to the proportion of the number of review comments in each time period to the total number of the reviews, calculating the review rate increase rate of every two adjacent time periods except for the first time period, and further calculating the average review rate increase rate according to the review rate increase rate of the adjacent time periods and the number of the time periods except for the first time periodAnd calculating the rate of increase of the rate of the follow-up evaluation from the second time period to the first time periodAverage growth rate according to the evaluation rateAnd rate of increase of rate of pursuitCalculating abnormal evaluation value of evaluation rateGenerally, people generally rarely comment on commodities after purchasing the commodities, so that blueprint rate and evaluation rate are not high, but blueprint rate and evaluation rate are high and are just one embodiment of a bill brushing, according to related data, blueprint rate and evaluation rate of a comment plate under a general merchant are about one value, if the blueprint rate and evaluation rate are far higher than the value, malicious sales of the merchant are likely to occur, but good evaluation is not excluded when a buyer receives goods at the moment, but in the using process, problems occur to the commodities or the quality of the commodities is good, the buyer carries out evaluation again, and similarly, blueprint is also satisfactory to the commodities when receiving the commodities, so that blueprint is carried out while being good evaluation is good, but the situation that the merchant brushes the bill by the merchant cannot be excludedThe reputation of the shops is improved to obtain more flow entries, so that reference is provided for whether the recent increase rate is abnormal or not by combining the average increase rate of the blueprint rate and the average increase rate of the evaluation rate, the abnormal evaluation value of the blueprint rate and the abnormal evaluation value of the evaluation rate are obtained by calculation, the blueprint rate and the evaluation rate are used as influence parameters, whether the merchants have fraud behaviors or not is judged by calculation, the merchants with the fraud behaviors are marked, and the good pictures in the comments under the marked merchants are hidden, so that the inducement factors of the buying behaviors of the buyers are reduced, the evaluation of the swiped bills does not influence the judgment of the buyers, the benefits of other merchants of the same type are guaranteed, and the selling false behaviors of the online shopping platform are reduced.
Further, the fraud determination module obtains an abnormal blueprint rate evaluation valueEvaluation value of abnormality of favorable rateEvaluation value for evaluation rate abnormalityAnd further performing fraud evaluation value calculation, the fraud evaluation valueWherein, in the step (A),、、are all the coefficients of the light-emitting diode,、、are all less than 1, when the fraud assessment value isWhen the fraud behavior evaluation value is larger than or equal to the fraud behavior evaluation value threshold value, the fraud behavior judgment module judges that the merchant to which the corresponding commodity belongs has fraud behavior, the fraud behavior judgment module marks the merchant, hides the favorable evaluation of the picture attached to the commodity comment plate under the marked merchant, and calculates the fraud behavior evaluation value of the merchant by combining the blueprint rate abnormal evaluation value, the favorable evaluation rate abnormal evaluation value and the evaluation rate abnormal evaluation value, the blueprint rate abnormal evaluation value, the favorable evaluation value and the evaluation rate abnormal evaluation value can reflect the brushing single-line evaluation value of the merchant, but the detection of whether the merchant has brushing single-line evaluation value is unreasonable only according to one of the blueprint rate abnormal evaluation value, so that the fraud behavior abnormity of the merchant is evaluated in a weighted average mode to reduce the inducement factors of the buying behavior of the buyer, the brushing comments have no influence on the judgment of the buyer, and the benefits of other similar merchants are guaranteed, and the sales faking behavior of the online shopping platform is reduced.
Further, an anti-fraud method based on big data processing comprises the following steps:
s1: the comment abnormity monitoring module monitors comment abnormity conditions of a user commenting in any commodity comment plate through the mobile terminal at the current moment, the comment abnormity monitoring module acquires the current comment state, the comment state is good comment, medium comment or poor comment, and if the current comment is good comment, the comment abnormity monitoring module acquires the word number of comments of the userAnd number of uploaded picturesWhen the number of words of the comment is greater than or equal to a first preset value and the number of pictures is greater than or equal to a second preset value, calculating a comment abnormity evaluation value according to the number of words of the comment and the number of uploaded picturesWherein, in the step (A),as a function of the number of the coefficients,in order to be at the first preset value,when the comment abnormal evaluation value is greater than or equal to the comment abnormal evaluation value threshold value, abnormal detection is carried out on the picture uploaded by the user under the current comment plate through an abnormal picture detection module;
s2: the abnormal picture detection module acquires all buyer uploaded pictures and any seller displayed pictures of the commodity under a comment plate of the buyer, wherein the buyer uploaded pictures are first reference pictures, the seller displayed pictures are second reference pictures, the abnormal picture detection module extracts figures and object outlines in the first reference pictures and the second reference pictures and further extracts figures and object outlines in the pictures to be detected uploaded under the comment plate of a current user, the abnormal picture detection module compares the figures and the object outlines in the pictures to be detected with the figures and the object outlines in the first reference pictures and the second reference pictures respectively, the picture abnormal detection module adjusts the size of the pictures to be detected to be consistent with the first reference pictures or the second reference pictures during each comparison and enables the pictures to be detected to be coincident with the first reference pictures or the second reference pictures, further acquiring the coincidence proportion of the figure and the article outline of the picture to be detected and the first reference picture or the second reference picture in the figure and article outline part of the first reference picture or the second reference picture, and when the coincidence proportion is more than or equal to the coincidence proportion threshold value, further performing anomaly detection on the picture background in the picture to be detected through a picture background anomaly detection module;
s3: the image background anomaly detection module is connected with the image background data storage module, and further acquires an image background in the image to be detected and commodity image background data of a corresponding virtual shop, which is pre-stored in the image background data storage module, wherein the commodity image background is a first image background, the image background in the image to be detected is a second image background, and the image background anomaly detection module compares the second image background with the first image background one by one and detects whether the second image background is consistent with any one of the first image backgrounds;
s4: if the picture background abnormality detection module detects that the second picture background is consistent with any first picture background, the abnormality volume detection module further detects whether the order volume of any commodity in a certain time period of the corresponding store is abnormal or not, the abnormality volume detection module acquires the order volume in the certain time period of the corresponding store, divides the certain time period into a plurality of time periods, the time period closest to the current time is a first time period, the time period closest to the first time period is a second time period, further acquires the order volume of other time periods except the first time period, calculates the order volume increase rate of every two adjacent time periods except the first time period, and further calculates the average volume increase rate according to the order volume increase rate of the adjacent time periods and the number of the time periods except the first time periodAnd calculating the increase rate of the order volume from the second time interval to the first time intervalAccording to the average growth rate of the volume of trafficRate of increase of volume of transactions with ordersCalculating traffic anomaly evaluation valueAs a traffic anomaly evaluation valueWhen the volume of the corresponding commodity is larger than or equal to the volume of the transaction abnormal evaluation value threshold value, the volume of the corresponding commodity is abnormal within a certain time period, and the blueprint rate, the good evaluation rate and the evaluation rate of the comment plate under the corresponding commodity are respectively detected by the comment blueprint rate detection module, the comment good evaluation rate detection module and the comment evaluation rate detection module;
s5: the comment goodness-of-comment detection module acquires comment plates of commodities with abnormal transaction volumes corresponding to shops, further acquires the total number of comments and the goodness-of-comment quantity in a certain time period, divides the certain time period into a plurality of time periods, calculates the goodness-of-comment rate of the corresponding commodities in the plurality of time periods according to the proportion of the goodness-of-comment quantity in each time period to the total number of the comments, calculates the goodness-of-comment rate increase rate of every two adjacent time periods except the first time period, and further calculates the average goodness-of-comment rate increase rate according to the goodness-of-comment rate increase rate of the adjacent time periods and the number of the time periods except theAnd calculating the rate of increase of the goodness of the second period to the first periodAverage growth rate according to good scoreAnd rate of increase of goodnessCalculating a goodness-of-evaluation anomaly evaluation value;
S6: the comment blueprint rate detection module acquires comment plates of commodities with abnormal volume of bargaining corresponding stores, further acquires the total number of comments and the number of blueprint comments in a certain time period, divides the certain time period into a plurality of time periods, calculates the blueprint rate of the corresponding commodities in the plurality of time periods according to the proportion of the number of the blueprint comments in each time period to the total number of the comments, calculates the blueprint rate increase rate of every two adjacent time periods except the first time period in other time periods, and further calculates the average blueprint rate increase rate according to the blueprint rate increase rate of the adjacent time periods and the number of the time periods except the first time periodAnd calculating the blueprint rate increase rate from the second time interval to the first time intervalAverage growth rate according to blueprint rateAnd the increase rate of blueprint rateCalculating an abnormal evaluation value of blueprint rate;
S7: the comment evaluation rate detection module acquires comment plates of commodities with abnormal transaction volume corresponding to shops, further acquires the total number of comments and the number of evaluation comments in a certain time period, divides the certain time period into a plurality of time periods, calculates the evaluation rate of the corresponding commodities in the plurality of time periods according to the proportion of the number of evaluation comments in each time period to the total number of the comments, calculates the evaluation rate increase rate of every two adjacent time periods except the first time period, and further calculates the evaluation rate average increase rate according to the evaluation rate increase rate of the adjacent time periods and the number of the time periods except the first time periodAnd calculating the rate of increase of the rate of the follow-up evaluation from the second time period to the first time periodAverage growth rate according to the evaluation rateAnd rate of increase of rate of pursuitCalculating abnormal evaluation value of evaluation rate;
S8: fraud behavior determination module acquires abnormal blueprint rate evaluation valueEvaluation value of abnormality of favorable rateEvaluation value for evaluation rate abnormalityAnd further performing fraud evaluation value calculation and fraud evaluation value calculationWherein, in the step (A),、、are all the coefficients of the light-emitting diode,、、are all less than 1, when the fraud assessment value isAnd when the fraud behavior evaluation value is larger than or equal to the threshold value of the fraud behavior evaluation value, the fraud behavior judging module judges that the merchant to which the corresponding commodity belongs has fraud behaviors.
Compared with the prior art, the invention has the following beneficial effects: the method comprises the steps of monitoring comments, when abnormal comments are monitored, obtaining uploaded pictures of all buyers and displayed pictures of corresponding sellers under a comment plate to which the abnormal comments belong, comparing the pictures to be uploaded and the uploaded pictures and the displayed pictures of the abnormal comments, determining whether the pictures to be uploaded are abnormal, further detecting the transaction amount of shops within a certain time period, when the transaction amount is abnormal, calculating a fraud behavior evaluation value according to an evaluation value of the good rate abnormality, an evaluation value of the blueprint rate abnormality and an evaluation value of the chasing rate abnormality, and further judging whether fraud behaviors exist in merchants, so that sales of online shopping platforms are reduced, and benefits of the buyers and other sellers in the same party are guaranteed.
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 block diagram of an anti-fraud system based on big data processing according to the present invention;
fig. 2 is a schematic diagram of the steps of an anti-fraud method based on big data processing.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1-2, the present invention provides a technical solution:
an anti-fraud system based on big data processing comprises a picture background data storage module, a comment abnormity monitoring module, an abnormal picture detection module, a picture background abnormity detection module, an abnormal volume of transaction detection module, a comment blueprint rate detection module, a comment good rate detection module, a comment rate detection module and a fraud behavior judgment module, wherein the picture background data storage module is stored with commodity picture background data of a virtual shop in advance, the comment abnormity monitoring module is used for monitoring whether user comments are abnormal in real time, the abnormal picture detection module is used for detecting whether pictures uploaded on a comment plate by a user are abnormal or not, the picture background abnormity detection module is used for carrying out background detection on the uploaded pictures with the abnormality detected, the abnormal volume of transaction detection module is used for acquiring the volume of transaction of commodities in any shop within a certain period of time and monitoring the volume of transaction abnormity within the certain period of time according to the volume of transaction of commodities, the comment blueprint rate abnormity detection module is used for detecting whether the blueprint rate of a comment plate under any commodity in any shop is abnormal or not and calculating the blueprint rate abnormity evaluation valueThe comment goodness detection module is used for detecting whether the goodness of comment of a comment plate under any commodity in any shop is abnormal or not and calculating the goodness of comment abnormal evaluation valueThe comment evaluation rate detection module is used for detecting whether the evaluation rate of the comment plate under any commodity in any shop is abnormal or not and calculating the evaluation value of the evaluation rate abnormalityThe fraud determination module is used for determining the fraud rate according to the blueprint rateAnomaly evaluation valueEvaluation value of abnormality of favorable rateEvaluation value for evaluation rate abnormalityA determination is made of fraudulent behavior.
The comment abnormity monitoring module monitors comment abnormity conditions of a user commenting in any commodity comment plate through the mobile terminal at the current moment, the comment abnormity monitoring module acquires the current comment state, the comment state is good comment, medium comment or poor comment, and if the current comment is good comment, the comment abnormity monitoring module acquires the word number of comments of the userAnd number of uploaded picturesWhen the number of words of the comment is greater than or equal to a first preset value and the number of pictures is greater than or equal to a second preset value, calculating a comment abnormity evaluation value according to the number of words of the comment and the number of uploaded picturesWherein, in the step (A),as a function of the number of the coefficients,in order to be at the first preset value,when the comment abnormal evaluation value is greater than or equal to the comment abnormal evaluation value threshold value, the second preset value is used for displaying the picture uploaded by the user under the current comment plate through the abnormal picture detection moduleThe sheet is subjected to abnormality detection.
The abnormal picture detection module acquires all buyer uploaded pictures and any seller displayed pictures of the commodity under a comment plate of the buyer, wherein the buyer uploaded pictures are first reference pictures, the seller displayed pictures are second reference pictures, the abnormal picture detection module extracts figures and object outlines in the first reference pictures and the second reference pictures and further extracts figures and object outlines in the pictures to be detected uploaded under the comment plate of a current user, the abnormal picture detection module compares the figures and the object outlines in the pictures to be detected with the figures and the object outlines in the first reference pictures and the second reference pictures respectively, the picture abnormal detection module adjusts the size of the pictures to be detected to be consistent with the first reference pictures or the second reference pictures during each comparison and enables the pictures to be detected to be coincident with the first reference pictures or the second reference pictures, and further acquiring the coincidence proportion of the figure and the article outline of the picture to be detected and the first reference picture or the second reference picture in the figure and article outline part of the first reference picture or the second reference picture, and when the coincidence proportion is more than or equal to the coincidence proportion threshold, further performing anomaly detection on the picture background in the picture to be detected through a picture background anomaly detection module.
The picture background abnormity detection module is connected with the picture background data storage module, and further acquires a picture background in the picture to be detected, a picture background of the picture uploaded by the buyer, and commodity picture background data of a corresponding virtual shop, which are stored in the picture background data storage module in advance, wherein the picture background and the commodity picture background of the picture uploaded by the buyer are first picture backgrounds, the picture background in the picture to be detected is second picture backgrounds, and the picture background abnormity detection module compares the second picture backgrounds with the first picture backgrounds one by one and detects whether the second picture backgrounds are consistent with any one of the first picture backgrounds.
If the picture background abnormity detection module detects that the second picture background is consistent with any first picture background, the abnormal volume detection module is used for detecting any commodity in a certain time period of the corresponding shopDetecting whether the order volume is abnormal or not, acquiring the order volume in a certain time period of the corresponding shop by an abnormal volume detection module, dividing the certain time period into a plurality of time periods, wherein the time period closest to the current moment is a first time period, the time period closest to the first time period is a second time period, further acquiring the order volume in other time periods except the first time period, calculating the order volume growth rate of every two adjacent time periods except the first time period, and further calculating the average volume growth rate according to the order volume growth rate of the adjacent time periods and the number of the time periods except the first time periodAnd calculating the increase rate of the order volume from the second time interval to the first time intervalAccording to the average growth rate of the volume of trafficRate of increase of volume of transactions with ordersCalculating traffic anomaly evaluation valueAs a traffic anomaly evaluation valueWhen the volume of the corresponding commodity is larger than or equal to the volume of the transaction abnormal evaluation value threshold value, the volume of the corresponding commodity is abnormal within a certain time period, and the blueprint rate, the good evaluation rate and the evaluation rate of the comment plate under the corresponding commodity are respectively detected through the comment blueprint rate detection module, the comment good evaluation rate detection module and the comment evaluation rate detection module.
The comment goodness detection module acquires comment plates of commodities with abnormal transaction volumes corresponding to shops, further acquires the total number of comments and the goodness number in a certain period of time, and sends the comment plates to the comment quality detection moduleA certain time period is divided into a plurality of time periods, the good evaluation rate of the corresponding commodities in the time periods is calculated according to the proportion of the good evaluation number in each time period to the total number of the comments, the good evaluation rate increase rate of every two adjacent time periods except the first time period is calculated, and the good evaluation rate average increase rate is further calculated according to the good evaluation rate increase rate of the adjacent time periods and the number of the time periods except the first time periodAnd calculating the rate of increase of the goodness of the second period to the first periodAverage growth rate according to good scoreAnd rate of increase of goodnessCalculating a goodness-of-evaluation anomaly evaluation value。
The comment blueprint rate detection module acquires comment plates of commodities with abnormal volume of bargaining corresponding stores, further acquires the total number of comments and the number of blueprint comments in a certain time period, divides the certain time period into a plurality of time periods, calculates the blueprint rate of the corresponding commodities in the plurality of time periods according to the proportion of the number of the blueprint comments in each time period to the total number of the comments, calculates the blueprint rate increase rate of every two adjacent time periods except the first time period in other time periods, and further calculates the average blueprint rate increase rate according to the blueprint rate increase rate of the adjacent time periods and the number of the time periods except the first time periodAnd calculating the blueprint rate increase rate from the second time interval to the first time intervalAverage growth rate according to blueprint rateAnd the increase rate of blueprint rateCalculating an abnormal evaluation value of blueprint rate。
The comment evaluation rate detection module acquires comment plates of commodities with abnormal transaction volume corresponding to shops, further acquires the total number of comments and the number of evaluation comments in a certain time period, divides the certain time period into a plurality of time periods, calculates the evaluation rate of the corresponding commodities in the plurality of time periods according to the proportion of the number of evaluation comments in each time period to the total number of the comments, calculates the evaluation rate increase rate of every two adjacent time periods except the first time period, and further calculates the evaluation rate average increase rate according to the evaluation rate increase rate of the adjacent time periods and the number of the time periods except the first time periodAnd calculating the rate of increase of the rate of the follow-up evaluation from the second time period to the first time periodAverage growth rate according to the evaluation rateAnd rate of increase of rate of pursuitCalculating abnormal evaluation value of evaluation rate。
Fraud behavior determination module acquires abnormal blueprint rate evaluation valueEvaluation value of abnormality of favorable rateEvaluation value for evaluation rate abnormalityAnd further performing fraud evaluation value calculation and fraud evaluation value calculationWherein, in the step (A),、、are all the coefficients of the light-emitting diode,、、are all less than 1, when the fraud assessment value isWhen the fraud behavior evaluation value is larger than or equal to the fraud behavior evaluation value threshold value, the fraud behavior judging module judges that the merchant to which the corresponding commodity belongs has fraud behaviors, the fraud behavior judging module marks the merchant and hides the favorable comment of the picture attached to the comment plate of the corresponding commodity under the marked merchant.
An anti-fraud method based on big data processing comprises the following steps:
s1: comment exception monitoring module passes through current userMonitoring the comment abnormal condition of the mobile terminal when commenting under any commodity comment plate, acquiring the current comment state by a comment abnormal monitoring module, wherein the comment state is good comment, medium comment or poor comment, and if the current comment is good comment, acquiring the number of words of comment of a user by the comment abnormal monitoring moduleAnd number of uploaded picturesWhen the number of words of the comment is greater than or equal to a first preset value and the number of pictures is greater than or equal to a second preset value, calculating a comment abnormity evaluation value according to the number of words of the comment and the number of uploaded picturesWherein, in the step (A),as a function of the number of the coefficients,in order to be at the first preset value,when the comment abnormal evaluation value is greater than or equal to the comment abnormal evaluation value threshold value, abnormal detection is carried out on the picture uploaded by the user under the current comment plate through an abnormal picture detection module;
s2: the abnormal picture detection module acquires all buyer uploaded pictures and any seller displayed pictures of the commodity under a comment plate of the buyer, wherein the buyer uploaded pictures are first reference pictures, the seller displayed pictures are second reference pictures, the abnormal picture detection module extracts figures and object outlines in the first reference pictures and the second reference pictures and further extracts figures and object outlines in the pictures to be detected uploaded under the comment plate of a current user, the abnormal picture detection module compares the figures and the object outlines in the pictures to be detected with the figures and the object outlines in the first reference pictures and the second reference pictures respectively, the picture abnormal detection module adjusts the size of the pictures to be detected to be consistent with the first reference pictures or the second reference pictures during each comparison and enables the pictures to be detected to be coincident with the first reference pictures or the second reference pictures, further acquiring the coincidence proportion of the figure and the article outline of the picture to be detected and the first reference picture or the second reference picture in the figure and article outline part of the first reference picture or the second reference picture, and when the coincidence proportion is more than or equal to the coincidence proportion threshold value, further performing anomaly detection on the picture background in the picture to be detected through a picture background anomaly detection module;
s3: the image background anomaly detection module is connected with the image background data storage module, and further acquires an image background in the image to be detected and commodity image background data of a corresponding virtual shop, which is pre-stored in the image background data storage module, wherein the commodity image background is a first image background, the image background in the image to be detected is a second image background, and the image background anomaly detection module compares the second image background with the first image background one by one and detects whether the second image background is consistent with any one of the first image backgrounds;
s4: if the picture background abnormality detection module detects that the second picture background is consistent with any first picture background, the abnormality volume detection module further detects whether the order volume of any commodity in a certain time period of the corresponding store is abnormal or not, the abnormality volume detection module acquires the order volume in the certain time period of the corresponding store, divides the certain time period into a plurality of time periods, the time period closest to the current time is a first time period, the time period closest to the first time period is a second time period, further acquires the order volume of other time periods except the first time period, calculates the order volume increase rate of every two adjacent time periods except the first time period, and further calculates the average volume increase rate according to the order volume increase rate of the adjacent time periods and the number of the time periods except the first time periodAnd calculating the increase rate of the order volume from the second time interval to the first time intervalAccording to the average growth rate of the volume of trafficRate of increase of volume of transactions with ordersCalculating traffic anomaly evaluation valueAs a traffic anomaly evaluation valueWhen the volume of the corresponding commodity is larger than or equal to the volume of the transaction abnormal evaluation value threshold value, the volume of the corresponding commodity is abnormal within a certain time period, and the blueprint rate, the good evaluation rate and the evaluation rate of the comment plate under the corresponding commodity are respectively detected by the comment blueprint rate detection module, the comment good evaluation rate detection module and the comment evaluation rate detection module;
s5: the comment goodness-of-comment detection module acquires comment plates of commodities with abnormal transaction volumes corresponding to shops, further acquires the total number of comments and the goodness-of-comment quantity in a certain time period, divides the certain time period into a plurality of time periods, calculates the goodness-of-comment rate of the corresponding commodities in the plurality of time periods according to the proportion of the goodness-of-comment quantity in each time period to the total number of the comments, calculates the goodness-of-comment rate increase rate of every two adjacent time periods except the first time period, and further calculates the average goodness-of-comment rate increase rate according to the goodness-of-comment rate increase rate of the adjacent time periods and the number of the time periods except theAnd calculating the rate of increase of the goodness of the second period to the first periodAverage growth rate according to good scoreAnd rate of increase of goodnessCalculating a goodness-of-evaluation anomaly evaluation value;
S6: the comment blueprint rate detection module acquires comment plates of commodities with abnormal volume of bargaining corresponding stores, further acquires the total number of comments and the number of blueprint comments in a certain time period, divides the certain time period into a plurality of time periods, calculates the blueprint rate of the corresponding commodities in the plurality of time periods according to the proportion of the number of the blueprint comments in each time period to the total number of the comments, calculates the blueprint rate increase rate of every two adjacent time periods except the first time period in other time periods, and further calculates the average blueprint rate increase rate according to the blueprint rate increase rate of the adjacent time periods and the number of the time periods except the first time periodAnd calculating the blueprint rate increase rate from the second time interval to the first time intervalAverage growth rate according to blueprint rateAnd the increase rate of blueprint rateCalculating an abnormal evaluation value of blueprint rate;
S7: the comment follow-up rate detection module acquires a comment plate under a commodity with abnormal transaction amount corresponding to the shop, and further acquires a certain amount of commentThe total number of the comments and the number of the additional comments in the time period are divided into a plurality of time periods, the additional rating rate of the corresponding commodity in the plurality of time periods is calculated according to the proportion of the additional rating comment number in each time period to the total number of the comments, the additional rating rate increase rate of every two adjacent time periods except the first time period is calculated, and the average additional rating rate increase rate is further calculated according to the additional rating rate increase rate of the adjacent time periods and the number of the time periods except the first time periodAnd calculating the rate of increase of the rate of the follow-up evaluation from the second time period to the first time periodAverage growth rate according to the evaluation rateAnd rate of increase of rate of pursuitCalculating abnormal evaluation value of evaluation rate;
S8: fraud behavior determination module acquires abnormal blueprint rate evaluation valueEvaluation value of abnormality of favorable rateEvaluation value for evaluation rate abnormalityAnd further performing fraud evaluation value calculation and fraud evaluation value calculationWherein, in the step (A),、、are all the coefficients of the light-emitting diode,、、are all less than 1, when the fraud assessment value isAnd when the fraud behavior evaluation value is larger than or equal to the threshold value of the fraud behavior evaluation value, the fraud behavior judging module judges that the merchant to which the corresponding commodity belongs has fraud behaviors.
It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.
Finally, it should be noted that: although the present invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that changes may be made in the embodiments and/or equivalents thereof without departing from the spirit and scope of the invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. An anti-fraud system based on big data processing, characterized in that: the anti-fraud system comprises a picture background data storage module, a comment abnormity monitoring module, an abnormal picture detection module, a picture background abnormity detection module, an abnormal volume of transaction detection module, a comment blueprint rate detection module, a comment good rate detection module, a comment chasing rate detection module and a fraud behavior judgment module, wherein the picture background data storage module is stored with commodity picture background data of a virtual shop in advance, the comment abnormity monitoring module is used for monitoring whether user comments are abnormal in real time, the abnormal picture detection module is used for detecting whether pictures uploaded on a comment plate by a user are abnormal or not, the picture background abnormity detection module is used for carrying out background detection on uploaded pictures with the detected abnormalities, the abnormal volume of transaction detection module is used for acquiring the volume of transaction of commodities of any shop in a certain period of time and monitoring the volume of transaction abnormity in a certain period of time according to the volume of transaction of commodities, the comment blueprint rate abnormity detection module is used for detecting whether the blueprint rate of a comment plate under any commodity in any shop is abnormal or not and calculating a blueprint rate abnormity evaluation valueThe comment goodness detection module is used for detecting whether the goodness of comment of a comment plate under any commodity in any shop is abnormal or not and calculating the goodness of comment abnormal evaluation valueThe comment follow-up rate detection module is used for detecting whether the follow-up rate of the comment plate under any commodity in any shop is abnormal or not and calculating the abnormal follow-up rate evaluation valueThe fraud determination module estimates the value of the fraud rate anomaly according to the blueprint rateEvaluation value of abnormality of favorable rateEvaluation value for evaluation rate abnormalityA determination is made of fraudulent behavior.
2. An anti-fraud system based on big data processing according to claim 1, characterized in that: the comment abnormity monitoring module monitors comment abnormity conditions when a user comments on any commodity comment plate through the mobile terminal at the current moment, the comment abnormity monitoring module acquires the current comment state, the comment state is good comment, medium comment or poor comment, and if the current comment is good comment, the comment abnormity monitoring module acquires the word number of comment words of the user commentAnd number of uploaded picturesWhen the number of words of the comment is greater than or equal to a first preset value and the number of pictures is greater than or equal to a second preset value, calculating a comment abnormity evaluation value according to the number of words of the comment and the number of uploaded picturesWherein, in the step (A),as a function of the number of the coefficients,in order to be at the first preset value,and when the comment abnormal evaluation value is greater than or equal to the comment abnormal evaluation value threshold value, the abnormal picture detection module is used for carrying out abnormal detection on the picture uploaded by the user in the current comment plate.
3. An anti-fraud system based on big data processing according to claim 1 or 2, characterized in that: the abnormal picture detection module acquires all buyer uploaded pictures and seller displayed pictures of any commodity under the comment plate of the commodity, wherein the buyer uploaded pictures are first reference pictures, the seller displayed pictures are second reference pictures, the abnormal picture detection module extracts figures and object outlines in the first reference pictures and the second reference pictures and further extracts figures and object outlines in the pictures to be detected uploaded under the comment plate of the current user, the abnormal picture detection module compares the figures and the object outlines in the pictures to be detected with the figures and the object outlines in the first reference pictures and the second reference pictures respectively, and the picture abnormal detection module adjusts the size of the pictures to be detected to be consistent with the first reference pictures or the second reference pictures during each comparison, the method comprises the steps of enabling a picture to be detected to be overlapped with a first reference picture or a second reference picture, further obtaining the overlapping proportion of a figure and an article outline overlapping part of the picture to be detected and the first reference picture or the second reference picture in the figure and article outline part of the first reference picture or the second reference picture, and further conducting abnormity detection on a picture background in the picture to be detected through a picture background abnormity detection module when the overlapping proportion is larger than or equal to an overlapping proportion threshold value.
4. An anti-fraud system based on big data processing according to claim 3, characterized in that: the picture background abnormity detection module is connected with the picture background data storage module, and further acquires a picture background in the picture to be detected, a picture background of a picture uploaded by a buyer, and commodity picture background data of a corresponding virtual shop pre-stored in the picture background data storage module, wherein the picture background and the commodity picture background of the picture uploaded by the buyer are first picture backgrounds, the picture background in the picture to be detected is second picture backgrounds, and the picture background abnormity detection module compares the second picture backgrounds with the first picture backgrounds one by one to detect whether the second picture backgrounds are consistent with any first picture background.
5. An anti-fraud system based on big data processing according to claim 1 or 4, characterized in that: if the picture background abnormality detection module detects that the second picture background is consistent with any first picture background, whether the order volume of any commodity in a certain time period of the corresponding store is abnormal or not is further detected through the abnormal volume detection module, the abnormal volume detection module obtains the order volume in the certain time period of the corresponding store and divides the certain time period into a plurality of time periods, the time period closest to the current moment is a first time period, the time period closest to the first time period is a second time period, the order volume in other time periods except the first time period is further obtained, the order volume growth rate of every two adjacent time periods except the first time period is calculated, and the average volume growth rate is further calculated according to the order volume growth rate of the adjacent time periods and the number of the time periods except the first time periodAnd calculating the increase rate of the order volume from the second time interval to the first time intervalAccording to the average growth rate of the volume of trafficRate of increase of volume of transactions with ordersCalculating traffic anomaly evaluation valueAs a traffic anomaly evaluation valueWhen the volume of the corresponding commodity is larger than or equal to the volume of the transaction abnormal evaluation value threshold value, the volume of the corresponding commodity in the certain time period is abnormal, and the blueprint rate, the good evaluation rate and the evaluation rate of the comment plate under the corresponding commodity are respectively detected through the comment blueprint rate detection module, the comment good evaluation rate detection module and the comment evaluation rate detection module.
6. An anti-fraud system based on big data processing according to claim 5, characterized in that: the comment goodness-of-comment detection module acquires comment plates of commodities with abnormal transaction volume corresponding to shops, further acquires the total number of comments and the goodness-of-comment quantity in a certain time period, divides the certain time period into a plurality of time periods, calculates the goodness-of-comment rate of the corresponding commodities in the plurality of time periods according to the proportion of the goodness-of-comment quantity in each time period to the total number of the comments, calculates the goodness-of-comment rate increase rate of every two adjacent time periods except the first time period, and further calculates the average goodness-of-comment rate increase rate according to the goodness-of-comment rate increase rate of the adjacent time periods and the number of the time periods except the first time periodAnd calculating the rate of increase of the goodness of the second time interval to the first time intervalAverage growth rate according to good scoreAnd rate of increase of goodnessCalculating a goodness-of-evaluation anomaly evaluation value。
7. An anti-fraud system based on big data processing according to claim 5, characterized in that: the comment blueprint rate detection module acquires comment plates of commodities with abnormal volume of bargaining corresponding stores, further acquires the total number of comments and the number of blueprint comments in a certain time period, divides the certain time period into a plurality of time periods, calculates the blueprint rate of the corresponding commodities in the plurality of time periods according to the proportion of the number of blueprint comments in each time period to the total number of comments, calculates the blueprint rate increase rate of every two adjacent time periods except the first time period, and further calculates the average blueprint rate increase rate according to the blueprint rate increase rate of the adjacent time periods and the number of time periods except the first time periodAnd calculating the blueprint rate increase rate from the second time interval to the first time intervalAverage growth rate according to blueprint rateAnd the increase rate of blueprint rateCalculating an abnormal evaluation value of blueprint rate。
8. An anti-fraud system based on big data processing according to claim 5, characterized in that: the comment evaluation rate detection module acquires comment plates of commodities with abnormal transaction volume corresponding to shops, further acquires the total number of comments and the number of evaluation comments in a certain time period, divides the certain time period into a plurality of time periods, calculates the evaluation rate of the corresponding commodities in the plurality of time periods according to the proportion of the number of evaluation comments in each time period to the total number of the comments, calculates the evaluation rate increase rate of every two adjacent time periods except the first time period, and further calculates the evaluation rate average increase rate according to the evaluation rate increase rate of the adjacent time periods and the number of the time periods except the first time periodAnd calculating the rate of increase of the rate of the follow-up evaluation from the second time period to the first time periodAverage growth rate according to the evaluation rateAnd rate of increase of rate of pursuitCalculating abnormal evaluation value of evaluation rate。
9. An anti-fraud system based on big data processing according to claim 1, characterized in that: the fraud behavior determination module acquires an abnormal blueprint rate evaluation valueEvaluation value of abnormality of favorable rateEvaluation value for evaluation rate abnormalityAnd further performing fraud evaluation value calculation, the fraud evaluation valueWherein, in the step (A),、、are all the coefficients of the light-emitting diode,、、are all less than 1, when the fraud assessment value isAnd when the fraud behavior evaluation value is larger than or equal to the threshold value of the fraud behavior evaluation value, the fraud behavior judging module judges that the merchant to which the corresponding commodity belongs has fraud behaviors.
10. An anti-fraud method based on big data processing is characterized in that: the anti-fraud method comprises the following steps:
s1: the comment abnormity monitoring module is used for commenting the abnormal conditions of the comment of the user at the current moment under any commodity comment plate through the mobile terminalMonitoring, wherein the comment abnormity monitoring module acquires the current comment state, the comment state is good comment, medium comment or poor comment, and if the current comment is good comment, the comment abnormity monitoring module acquires the word number of the comment of the userAnd number of uploaded picturesWhen the number of words of the comment is greater than or equal to a first preset value and the number of pictures is greater than or equal to a second preset value, calculating a comment abnormity evaluation value according to the number of words of the comment and the number of uploaded picturesWherein, in the step (A),as a function of the number of the coefficients,in order to be at the first preset value,when the comment abnormal evaluation value is greater than or equal to the comment abnormal evaluation value threshold value, abnormal detection is carried out on the picture uploaded by the user under the current comment plate through an abnormal picture detection module;
s2: the abnormal picture detection module acquires all buyer uploaded pictures and any seller displayed pictures of the commodity under a comment plate of the buyer, wherein the buyer uploaded pictures are first reference pictures, the seller displayed pictures are second reference pictures, the abnormal picture detection module extracts figures and object outlines in the first reference pictures and the second reference pictures and further extracts figures and object outlines in the pictures to be detected uploaded under the comment plate of a current user, the abnormal picture detection module compares the figures and the object outlines in the pictures to be detected with the figures and the object outlines in the first reference pictures and the second reference pictures respectively, the picture abnormal detection module adjusts the size of the pictures to be detected to be consistent with the first reference pictures or the second reference pictures during each comparison and enables the pictures to be detected to be coincident with the first reference pictures or the second reference pictures, further acquiring the coincidence proportion of the figure and the article outline of the picture to be detected and the first reference picture or the second reference picture in the figure and article outline part of the first reference picture or the second reference picture, and when the coincidence proportion is more than or equal to the coincidence proportion threshold value, further performing anomaly detection on the picture background in the picture to be detected through a picture background anomaly detection module;
s3: the image background anomaly detection module is connected with the image background data storage module, and further acquires an image background in the image to be detected and commodity image background data of a corresponding virtual shop, which is pre-stored in the image background data storage module, wherein the commodity image background is a first image background, the image background in the image to be detected is a second image background, and the image background anomaly detection module compares the second image background with the first image background one by one and detects whether the second image background is consistent with any one of the first image backgrounds;
s4: if the picture background abnormality detection module detects that the second picture background is consistent with any first picture background, the abnormality volume detection module further detects whether the order volume of any commodity in a certain time period of the corresponding store is abnormal or not, the abnormality volume detection module acquires the order volume in the certain time period of the corresponding store, divides the certain time period into a plurality of time periods, the time period closest to the current time is a first time period, the time period closest to the first time period is a second time period, further acquires the order volume of other time periods except the first time period, calculates the order volume increase rate of every two adjacent time periods except the first time period, and further calculates the average volume increase rate according to the order volume increase rate of the adjacent time periods and the number of the time periods except the first time periodAnd calculating the increase rate of the order volume from the second time interval to the first time intervalAccording to the average growth rate of the volume of trafficRate of increase of volume of transactions with ordersCalculating traffic anomaly evaluation valueAs a traffic anomaly evaluation valueWhen the volume of the corresponding commodity is larger than or equal to the volume of the transaction abnormal evaluation value threshold value, the volume of the corresponding commodity is abnormal within a certain time period, and the blueprint rate, the good evaluation rate and the evaluation rate of the comment plate under the corresponding commodity are respectively detected by the comment blueprint rate detection module, the comment good evaluation rate detection module and the comment evaluation rate detection module;
s5: the comment goodness-of-comment detection module acquires comment plates of commodities with abnormal transaction volumes corresponding to shops, further acquires the total number of comments and the goodness-of-comment quantity in a certain time period, divides the certain time period into a plurality of time periods, calculates the goodness-of-comment rate of the corresponding commodities in the plurality of time periods according to the proportion of the goodness-of-comment quantity in each time period to the total number of the comments, calculates the goodness-of-comment rate increase rate of every two adjacent time periods except the first time period, and further calculates the average goodness-of-comment rate increase rate according to the goodness-of-comment rate increase rate of the adjacent time periods and the number of the time periods except theAnd calculating the rate of increase of the goodness of the second period to the first periodAverage growth rate according to good scoreAnd rate of increase of goodnessCalculating a goodness-of-evaluation anomaly evaluation value;
S6: the comment blueprint rate detection module acquires comment plates of commodities with abnormal volume of bargaining corresponding stores, further acquires the total number of comments and the number of blueprint comments in a certain time period, divides the certain time period into a plurality of time periods, calculates the blueprint rate of the corresponding commodities in the plurality of time periods according to the proportion of the number of the blueprint comments in each time period to the total number of the comments, calculates the blueprint rate increase rate of every two adjacent time periods except the first time period in other time periods, and further calculates the average blueprint rate increase rate according to the blueprint rate increase rate of the adjacent time periods and the number of the time periods except the first time periodAnd calculating the blueprint rate increase rate from the second time interval to the first time intervalAverage growth rate according to blueprint rateAnd the increase rate of blueprint rateCalculating an abnormal evaluation value of blueprint rate;
S7: the comment evaluation rate detection module acquires comment plates of commodities with abnormal transaction volume corresponding to shops, further acquires the total number of comments and the number of evaluation comments in a certain time period, divides the certain time period into a plurality of time periods, calculates the evaluation rate of the corresponding commodities in the plurality of time periods according to the proportion of the number of evaluation comments in each time period to the total number of the comments, calculates the evaluation rate increase rate of every two adjacent time periods except the first time period, and further calculates the evaluation rate average increase rate according to the evaluation rate increase rate of the adjacent time periods and the number of the time periods except the first time periodAnd calculating the rate of increase of the rate of the follow-up evaluation from the second time period to the first time periodAverage growth rate according to the evaluation rateAnd rate of increase of rate of pursuitCalculating abnormal evaluation value of evaluation rate;
S8: fraud behavior determination module acquires abnormal blueprint rate evaluation valueEvaluation value of abnormality of favorable rateEvaluation value for evaluation rate abnormalityAnd further performing fraud evaluation value calculation and fraud evaluation value calculationWherein, in the step (A),、、are all the coefficients of the light-emitting diode,、、are all less than 1, when the fraud assessment value isAnd when the fraud behavior evaluation value is larger than or equal to the fraud behavior evaluation value threshold value, the fraud behavior judging module judges that the merchant to which the corresponding commodity belongs has fraud behavior, marks the merchant and hides the favorable comment of the picture attached to the comment plate of the corresponding commodity under the marked merchant.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110304854.9A CN112700265A (en) | 2021-03-23 | 2021-03-23 | Anti-fraud system and method based on big data processing |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110304854.9A CN112700265A (en) | 2021-03-23 | 2021-03-23 | Anti-fraud system and method based on big data processing |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112700265A true CN112700265A (en) | 2021-04-23 |
Family
ID=75515405
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110304854.9A Pending CN112700265A (en) | 2021-03-23 | 2021-03-23 | Anti-fraud system and method based on big data processing |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112700265A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113436712A (en) * | 2021-06-08 | 2021-09-24 | 四川数字链享科技有限公司 | Evaluation management system for smart medical cloud service platform |
CN115511584A (en) * | 2022-11-08 | 2022-12-23 | 深圳市必凡娱乐科技有限公司 | E-commerce platform false transaction order monitoring method and system |
CN115983873A (en) * | 2022-12-29 | 2023-04-18 | 珠海蓝度科技有限公司 | Big data based user data analysis management system and method |
CN117474686A (en) * | 2023-12-11 | 2024-01-30 | 万链指数(青岛)信息科技有限公司 | Financial data prediction system based on blockchain and big data |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108573432A (en) * | 2018-04-23 | 2018-09-25 | 重庆南米电子商务有限公司 | Transaction supervisory systems and method for e-commerce |
CN110766708A (en) * | 2019-10-24 | 2020-02-07 | 西安工业大学 | Image comparison method based on contour similarity |
CN112269920A (en) * | 2020-10-14 | 2021-01-26 | 广州五子科技有限公司 | Method for identifying brush bills |
CN112330833A (en) * | 2020-10-30 | 2021-02-05 | 中国平安人寿保险股份有限公司 | Face recognition attendance data verification method and device and computer equipment |
CN112396433A (en) * | 2020-11-30 | 2021-02-23 | 翼果(深圳)科技有限公司 | Method and system for identifying false commodity comments based on behavior of person to be evaluated |
-
2021
- 2021-03-23 CN CN202110304854.9A patent/CN112700265A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108573432A (en) * | 2018-04-23 | 2018-09-25 | 重庆南米电子商务有限公司 | Transaction supervisory systems and method for e-commerce |
CN110766708A (en) * | 2019-10-24 | 2020-02-07 | 西安工业大学 | Image comparison method based on contour similarity |
CN112269920A (en) * | 2020-10-14 | 2021-01-26 | 广州五子科技有限公司 | Method for identifying brush bills |
CN112330833A (en) * | 2020-10-30 | 2021-02-05 | 中国平安人寿保险股份有限公司 | Face recognition attendance data verification method and device and computer equipment |
CN112396433A (en) * | 2020-11-30 | 2021-02-23 | 翼果(深圳)科技有限公司 | Method and system for identifying false commodity comments based on behavior of person to be evaluated |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113436712A (en) * | 2021-06-08 | 2021-09-24 | 四川数字链享科技有限公司 | Evaluation management system for smart medical cloud service platform |
CN113436712B (en) * | 2021-06-08 | 2024-01-09 | 四川临丰医疗科技有限公司 | Evaluation management system for intelligent medical cloud service platform |
CN115511584A (en) * | 2022-11-08 | 2022-12-23 | 深圳市必凡娱乐科技有限公司 | E-commerce platform false transaction order monitoring method and system |
CN115983873A (en) * | 2022-12-29 | 2023-04-18 | 珠海蓝度科技有限公司 | Big data based user data analysis management system and method |
CN117474686A (en) * | 2023-12-11 | 2024-01-30 | 万链指数(青岛)信息科技有限公司 | Financial data prediction system based on blockchain and big data |
CN117474686B (en) * | 2023-12-11 | 2024-03-29 | 万链指数(青岛)信息科技有限公司 | Financial data prediction system based on blockchain and big data |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112700265A (en) | Anti-fraud system and method based on big data processing | |
CN108038696B (en) | Method and system for detecting bill swiping based on equipment identification code and social group information | |
CN108573432B (en) | Transaction supervision system and method for electronic commerce | |
US11030622B2 (en) | Card systems and methods | |
US10438228B2 (en) | Systems and methods for price matching and comparison | |
US20210233103A1 (en) | Sales promotion system and sales promotion method | |
CN112561543A (en) | E-commerce platform false transaction order monitoring method and system based on full-period logistics data analysis and cloud server | |
Qureshi et al. | BARRIERS TO ADOPTION OF ONLINE SHOPPING IN PAKISTAN. | |
CN113362126A (en) | Order data detection method and system applied to online business | |
CN110866807A (en) | Electronic commerce transaction system | |
CN115187065A (en) | Garment commodity quality analysis system based on electronic commerce platform | |
CN110866220A (en) | Selection recommendation system in e-commerce activity process | |
Olotewo | Examining the antecedents of in-store and online purchasing behavior: a case of Nigeria | |
CN115904936A (en) | Block chain evaluation system for node performance test | |
CN111143575B (en) | Optimization method and system for store commodity structure | |
CN113256372A (en) | Commodity sale system and method based on electronic commerce | |
CN114266629A (en) | E-commerce data processing system based on fair transaction | |
CN112734508A (en) | E-commerce transaction data analysis system based on cloud platform | |
CN114493415A (en) | Commodity platform control method, system and readable storage medium | |
US10127575B1 (en) | Systems and methods for determining impact of high-affinity baskets | |
CN116611796B (en) | Exception detection method and device for store transaction data | |
CN117078294B (en) | Cloud computing-based big data information intelligent analysis method and system | |
CN117236996B (en) | User behavior prediction method and system based on big data analysis | |
Zhang | Analysis of influencing factors of purchase decision for online consumption of clothes | |
CN114626888A (en) | Malicious behavior prediction method for online shopping platform |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210423 |