E-commerce platform consumption risk assessment method
By analyzing the seller's product reviews and equipment information, and combining IP information, we automatically determine whether there are risks in the products and sellers on the e-commerce platform, solving the problem of inability to effectively evaluate seller risks in the existing technology, and realizing a timely warning of buyer risks and fund protection reduces the risk of fraud.
Patent Information
- Application Number
- CN202410922117.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-10
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-07-10
AI Technical Summary
The existing consumption risk assessment methods of e-commerce platforms cannot automatically obtain the risk situation of the seller corresponding to the product, and cannot judge the authenticity of the product review, making it difficult for buyers to judge whether the seller has committed fraud and cause property losses.
By obtaining the seller's product review information, analyzing the review scores and comment texts, combining device information and IP information, using information analysis strategies, judgment strategies and evaluation strategies, automatically determine whether there are risks in the product and seller, and warn and funds are provided to buyers. Protect.
It effectively reduces the risk of buyers being fraudulent, prevents property losses, improves transaction security, and reduces misjudgment and economic losses.
Smart Images

Figure HDA0004937010960000011 
Figure HDA0004937010960000012
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of fraudulent behavior identification, and in particular to a method for assessing consumer risk on an e-commerce platform. Background Art
[0002] The e-commerce platform consumer risk assessment method can ensure that consumers on the e-commerce platform can shop and trade more safely and with greater confidence during the transaction process, effectively reduce the risk perception of consumers on the e-commerce platform, enhance the reputation and competitiveness of the platform, and promote the healthy development of the e-commerce industry. It includes reputation rating systems, payment security measures, consumer protection policies, data encryption and privacy protection, risk warning mechanisms, consumer education and guidelines, and technical review and certification;
[0003] Establish and maintain a credit rating system for consumers and merchants, and conduct assessments based on historical transaction records, reviews, and complaints, which can help consumers judge the credibility and service quality of merchants. Use secure payment platforms and third-party payment systems to ensure the safety of consumers' funds during transactions. Formulate clear consumer protection policies, including return and exchange policies, complaint handling procedures, etc., to protect consumer rights and interests, reduce consumer risks, strengthen the encrypted storage and transmission of user data, ensure that personal privacy information is not leaked, and prevent information from being used by criminals. Establish a risk early warning mechanism, monitor transaction behaviors and patterns, and promptly detect and respond to potential fraud and risky behaviors. Provide consumer education and shopping guides to help consumers identify and avoid risks in transactions, enhance consumers' self-protection awareness, conduct technical audits and certifications on merchants and products to ensure that they comply with relevant regulations and quality standards, and reduce consumers' selection and purchase risks;
[0004] The existing consumer risk assessment methods on e-commerce platforms cannot automatically obtain the risk status of the sellers corresponding to the goods purchased by the buyers, cannot automatically determine the authenticity of the product reviews based on the comments of the sellers' products, and thus determine whether the sellers and products are risky, cannot automatically determine whether the sellers are selling normally based on the sellers' IP, and remind the buyers that there are risks in this consumption. Due to limited information channels, it is difficult for buyers to determine whether the sellers have committed fraud, thereby causing property losses. Its practicality is definitely limited. Summary of the invention
[0005] The present invention provides a method for assessing consumption risks on an e-commerce platform, which is used to promote solving the problems raised in the background technology.
[0006] The present invention provides the following technical solution: a method for assessing consumption risk on an e-commerce platform, comprising:
[0007] Get target information;
[0008] The target information includes information about the seller and information about the goods purchased by the buyer;
[0009] The seller information includes the seller's products, review information, and device information;
[0010] Based on the target information, through information analysis strategies, analytical data is formed to determine whether the comments of the goods purchased by the buyer and other goods of the seller of the goods are valid comments, so as to determine whether the goods purchased by the buyer and the seller of the goods have risks;
[0011] Based on the analysis data, through the seller judgment strategy, judgment data is formed to further judge the products that are initially judged to be high-risk and the sellers of the products to prevent misjudgment;
[0012] Based on the judgment data, the seller is evaluated through the evaluation strategy to form evaluation data, so as to determine whether there is a problem with the product or the seller based on further judgment results, so as to warn the buyer and protect the funds;
[0013] Based on the evaluation data, when there is a problem with the goods purchased by the buyer or the seller of the goods, the buyer will be warned and his funds will be protected.
[0014] As an optional solution to the e-commerce platform consumption risk assessment method of the present invention, the information analysis strategy is specifically:
[0015] Obtain the goods that the buyer has confirmed the order for and set them as target goods;
[0016] Get the seller corresponding to the target product and set it as the target seller;
[0017] Obtain all products of the target seller and set them as the target seller's products;
[0018] Obtaining review information of each target seller's product, wherein the review information includes a review score and a review text;
[0019] Get the total score and effective score of the review;
[0020] Among them, the effective score < the total score of the review × 80%;
[0021] The target seller's products with review scores ≥ total review scores × 80% are defined as favorable products;
[0022] The target seller's products with review scores ≥ valid scores and review scores < total review scores × 80% are defined as valid products;
[0023] The number of target seller products, the number of well-received products, and the number of valid products are obtained respectively, and are defined as the total number of sellers, the number of well-received products, and the number of valid products respectively;
[0024] If the number of positive reviews ≥ the total number of sellers × 80%, the review judgment strategy is executed for each positive review product;
[0025] Obtain the quantity of the target seller's products that are determined to have low product risk, and set it as the low-risk quantity;
[0026] If the low-risk quantity ≥ the effective quantity, the target seller is deemed to have low risk;
[0027] If the low-risk quantity is less than the effective quantity, the target seller is deemed to have high risk and the target product determination strategy is implemented;
[0028] If the number of favorable reviews is less than the total number of sellers × 80% and the number of favorable reviews is greater than or equal to the valid number, the target seller is considered to have low risk;
[0029] If the number of positive reviews is less than the effective number, the target seller is deemed to be high risk and the target product determination strategy is executed.
[0030] As an optional solution of the e-commerce platform consumption risk assessment method of the present invention, the target commodity determination strategy is specifically:
[0031] Get the review score of the target product and set it as the target score;
[0032] If the target score is ≥ the total review score × 80%, the review judgment strategy is executed on the target product;
[0033] If the target score is less than the total score of the review × 80% and the target score is greater than or equal to the effective score, the target product is considered to have low risk;
[0034] If the target score is less than the effective score, the target product is deemed to be high risk and the seller's judgment strategy is executed.
[0035] As an optional solution of the e-commerce platform consumption risk assessment method of the present invention, the comment judgment strategy is specifically:
[0036] Get all the comments of the product;
[0037] Set comment keywords;
[0038] Comment text containing comment keywords will be deemed invalid text;
[0039] Get the number of comments on the product, which is defined as the number of comments;
[0040] Get the number of invalid words of the product and set it as invalid quantity;
[0041] If the number of invalid reviews ≥ the number of reviews × 80%, it means that the product has few valid reviews and is considered to have high risk;
[0042] If (number of reviews - number of invalid reviews) ≥ number of reviews × 80%, it means that the product has many valid reviews and is judged to have low risk.
[0043] As an optional solution of the e-commerce platform consumption risk assessment method of the present invention, the seller determination strategy is specifically:
[0044] Obtain the seller's device information, including the seller's device IP and the number of the seller's devices;
[0045] If the number of seller devices = 1, and the number of seller device IPs = 1, then obtain the network IP to which the seller device IP is connected and identify it as the target IP;
[0046] If the number of seller devices = 1, and the number of seller device IPs > 1, then obtain the number of times each seller device IP is used, extract the seller device IP corresponding to the largest number of times, and define it as the commonly used IP;
[0047] Get the network IP to which the commonly used IP is connected and identify it as the target IP;
[0048] If the number of seller devices is greater than 1, and the number of seller device IPs is equal to 1, then obtain the network IP to which the seller device IP is connected and identify it as the target IP;
[0049] If the number of seller devices is greater than 1, and the number of seller device IPs is greater than 1, then obtain the network IPs to which each seller device IP corresponding to each seller device is connected, and identify them as target IPs respectively;
[0050] Execute risk assessment strategy on target IP.
[0051] As an optional solution to the e-commerce platform consumption risk assessment method of the present invention, the risk determination strategy is specifically:
[0052] Get all device IPs that the target IP has connected to and set them as the judgment IP;
[0053] Obtaining a risk level for each determined IP, wherein the risk level includes high risk and low risk;
[0054] If there is a high-risk judgment IP, the judgment IP and the target IP are considered abnormal;
[0055] If all the determined IPs corresponding to the target IP are low risk, the target IP is considered normal;
[0056] If the target IP is abnormal, the seller's device IP is determined to be abnormal, and the misjudgment analysis strategy is executed;
[0057] If all target IPs are normal, the seller's device IP is considered normal.
[0058] As an optional solution of the e-commerce platform consumption risk assessment method of the present invention, the misjudgment analysis strategy is specifically:
[0059] Get all the determined IPs corresponding to the target IP;
[0060] Get the number of times each determination IP uses the target IP;
[0061] Sort the IP addresses from large to small based on the number of times used, and set the IP address corresponding to the largest number of times used as the comparison IP address.
[0062] Get the abnormal judgment IP and define it as abnormal IP;
[0063] Get the IP address of the seller's device;
[0064] If the compared IP is consistent with the abnormal IP, the time periods when the seller's device IP and the abnormal IP used the target IP are obtained respectively, and are defined as the seller's use period and the abnormal use period respectively;
[0065] If the seller's usage period has a high overlap with the abnormal usage period, the seller's device IP is determined to be abnormal and the target seller is considered to be high risk;
[0066] If the seller's usage period has a low overlap with the abnormal usage period, the seller's device IP is considered normal and the target seller is considered to have low risk;
[0067] If the compared IP is inconsistent with the abnormal IP, the time period determination strategy is executed.
[0068] As an optional solution of the e-commerce platform consumption risk assessment method of the present invention, the time period determination strategy is specifically as follows:
[0069] Get the seller's usage period and abnormal usage period respectively;
[0070] Obtain the time period when the comparison IP uses the target IP, and define it as the comparison usage period;
[0071] If the comparison usage period has a high overlap with the abnormal usage period, the comparison IP is considered abnormal;
[0072] If the seller's usage period has a high overlap with the abnormal usage period, or the seller's usage period has a high overlap with the comparison usage period, then the seller's device IP is determined to be abnormal and the target seller is considered to have a high risk;
[0073] If the seller's usage period has a low overlap with the abnormal usage period, and the seller's usage period has a low overlap with the comparison usage period, then the seller's device IP is judged to be normal and the target seller is deemed to have a low risk.
[0074] As an optional solution of the e-commerce platform consumption risk assessment method of the present invention, the assessment strategy is specifically:
[0075] Obtain the determination results of the target seller and the target product respectively;
[0076] If the target product is determined to have low risk, the payment interface will be entered;
[0077] If the target seller is determined to have low risk, the payment interface will be entered;
[0078] If the target product is identified as high risk, the buyer will be notified that the product he / she purchased has risks and will be advised to cancel the order;
[0079] If the target seller is identified as high risk, the buyer will be notified that the merchant corresponding to the product he / she purchased has risks and the buyer is advised to cancel the order;
[0080] If the buyer chooses to continue paying, the funds in the target seller's account will be locked after the buyer makes the payment until the buyer confirms that the purchase is correct and unlocks the funds;
[0081] If the buyer reports to the platform that there is a problem with the consumption and provides proof within the time period of locking the funds, the funds will be directly transferred back to the buyer's payment account.
[0082] The present invention has the following beneficial effects:
[0083] 1. The consumption risk assessment method of the e-commerce platform obtains the review information of the target product of the buyer's confirmed order, and obtains the review information of the merchant corresponding to the target product and all the merchant's products, compares the review information with the set keywords, and automatically determines whether the product review is valid, that is, whether the product review is an automatic positive review by the system or a professional review hired by the merchant, or a valid review published by the buyer based on his own usage and the product itself after normal consumption. It automatically determines whether the target product is a false sale that defrauds the buyer, such as concert tickets, and reminds the buyer that there is a risk in this consumption, thereby reducing the risk of the buyer being defrauded and preventing the buyer from causing property losses.
[0084] 2. The e-commerce platform's consumer risk assessment method obtains the seller's device IP, the seller's commonly used network IP, and all device IPs that have been connected to the network IP for buyers who are suspected of having many professional reviews. By obtaining the risk level of each device IP, that is, whether the device has committed fraud, the seller's risk level is automatically determined. If there is a high-risk device IP, it is identified as an abnormal device and the risk level of the seller's device IP is increased, thereby automatically determining whether the seller is selling normally and reminding the buyer that there is a risk in this consumption, reducing the risk of the buyer being defrauded, and preventing the buyer from causing property losses.
[0085] 3. The consumer risk assessment method of the e-commerce platform obtains the devices that use the most network IPs and identifies them as comparison devices. For the device IPs of sellers determined to be at high risk, the time periods when the seller's device, abnormal device, and comparison device use the network are obtained respectively. If the overlap between the network used by the abnormal device and the comparison device is high, the network risk is determined to be high. If the overlap between the network used by the abnormal device and the comparison device is low, the network risk is determined to be low, indicating that the abnormal device may be using the network free. If the overlap between the network used by the seller's device and the abnormal device is high, the seller's risk is determined to be high. If the overlap between the network used by the seller's device and the abnormal device is low, the seller's risk is determined to be low, indicating that the seller's device may be using the network free. Based on the seller's risk assessment results, it is automatically determined whether the seller is selling normally, and the buyer is reminded that there is a risk in this consumption, thereby reducing the risk of the buyer being defrauded and preventing the buyer from causing property losses. At the same time, it also prevents misjudgment of the seller, which may cause certain economic losses. BRIEF DESCRIPTION OF THE DRAWINGS
[0086] Figure 1 This is a flow chart of the e-commerce platform consumption risk assessment method of the present invention;
[0087] Figure 2 It is a schematic diagram of the association map of the present invention. DETAILED DESCRIPTION
[0088] 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.
[0089] Example 1: A method for assessing consumption risk on an e-commerce platform, see Figure 1 ,include:
[0090] Get target information;
[0091] The target information includes information about the seller and information about the goods purchased by the buyer;
[0092] The seller information includes the seller's products, review information, and device information;
[0093] Based on the target information, through information analysis strategies, analytical data is formed to determine whether the comments of the goods purchased by the buyer and other goods of the seller of the goods are valid comments, so as to determine whether the goods purchased by the buyer and the seller of the goods have risks;
[0094] Based on the analysis data, through the seller judgment strategy, judgment data is formed to further judge the products that are initially judged to be high-risk and the sellers of the products to prevent misjudgment;
[0095] Based on the judgment data, the seller is evaluated through the evaluation strategy to form evaluation data, so as to determine whether there is a problem with the product or the seller based on further judgment results, so as to warn the buyer and protect the funds;
[0096] Based on the evaluation data, when there is a problem with the goods purchased by the buyer or the seller of the goods, the buyer will be warned and his funds will be protected.
[0097] Through the above method, the risk situation of the seller corresponding to the product is automatically obtained according to the product purchased by the buyer, and the authenticity of the product review is automatically judged according to the review text of the seller's product, so as to judge whether the seller and the product have risks. According to the seller's IP, it is automatically judged whether the seller is selling normally, and the buyer is reminded that there is a risk in this consumption, thereby reducing the risk of the buyer being defrauded and preventing the buyer from causing property losses.
[0098] Embodiment 2: This embodiment is an improvement made on the basis of embodiment 1. The e-commerce platform consumption risk assessment method and the information analysis strategy are specifically as follows:
[0099] Obtain the goods that the buyer has confirmed the order for and set them as target goods;
[0100] Get the seller corresponding to the target product and set it as the target seller;
[0101] Obtain all products of the target seller and set them as the target seller's products;
[0102] Obtaining review information of each target seller's product, the review information including review scores and review text, the review score being the overall review score of the product, that is, review score = score evaluated by each buyer who purchased the product / total number of buyers who purchased the product;
[0103] Obtain the total score and effective score of the review, where the total score of the review is 100 and the effective score is 60;
[0104] Among them, the effective score < the total score of the review × 80%;
[0105] The target seller's products with review scores ≥ total review scores × 80% are defined as favorable products;
[0106] The target seller's products with review scores ≥ valid scores and review scores < total review scores × 80% are defined as valid products;
[0107] The number of target seller products, the number of well-received products, and the number of valid products are obtained respectively, and are defined as the total number of sellers, the number of well-received products, and the number of valid products respectively;
[0108] If the number of positive reviews ≥ the total number of sellers × 80%, the review judgment strategy is executed for each positive review product;
[0109] Obtain the quantity of the target seller's products that are determined to have low product risk, and set it as the low-risk quantity;
[0110] If the low-risk quantity ≥ the effective quantity, the target seller is deemed to have low risk;
[0111] If the low-risk quantity is less than the effective quantity, the target seller is deemed to have high risk and the target product determination strategy is implemented;
[0112] If the number of favorable reviews is less than the total number of sellers × 80% and the number of favorable reviews is greater than or equal to the valid number, the target seller is considered to have low risk;
[0113] If the number of positive reviews is less than the effective number, the target seller is deemed to be high risk and the target product determination strategy is executed.
[0114] The target product determination strategy is specifically as follows:
[0115] Get the review score of the target product and set it as the target score;
[0116] If the target score is ≥ the total review score × 80%, the review judgment strategy is executed on the target product;
[0117] If the target score is less than the total score of the review × 80% and the target score is greater than or equal to the effective score, the target product is considered to have low risk;
[0118] If the target score is less than the effective score, the target product is deemed to be high risk and the seller's judgment strategy is executed.
[0119] The comment determination strategy is specifically as follows:
[0120] Obtain all the comments on the product, which are comments posted by all buyers of each product after purchasing the product based on their own usage and the product itself. The comments are of certain guidance to buyers who want to buy the product. Based on the comment keywords, it is determined whether the comments are default comments set by the system or professional comments hired by the seller, or valid comments posted by buyers based on their own usage and the product itself after normal consumption, where normal consumption refers to a complete set of consumption behaviors in which the buyer receives the product after spending the product amount.
[0121] Setting review keywords, where the review keywords are common words used in the default reviews set by the system or the professional reviews hired by the seller, such as system default good reviews, fast logistics speed good reviews, etc.;
[0122] Comment text containing comment keywords will be deemed invalid text;
[0123] Get the number of comments on the product, which is defined as the number of comments;
[0124] Get the number of invalid words of the product and set it as invalid quantity;
[0125] If the number of invalid reviews ≥ the number of reviews × 80%, it means that the product has few valid reviews and is considered to have high risk;
[0126] If (number of reviews - number of invalid reviews) ≥ number of reviews × 80%, it means that the product has many valid reviews and is judged to have low risk.
[0127] Embodiment 3: This embodiment is an improvement made on the basis of embodiment 2. Figure 2 In this embodiment, the seller determination strategy is specifically as follows:
[0128] Obtain the seller's device information, including the seller's device IP and the number of the seller's devices;
[0129] If the number of seller devices = 1, and the number of seller device IPs = 1, then obtain the network IP to which the seller device IP is connected and identify it as the target IP;
[0130] If the number of seller devices = 1, and the number of seller device IPs > 1, then obtain the number of times each seller device IP is used, extract the seller device IP corresponding to the largest number of times, and define it as the commonly used IP;
[0131] Get the network IP to which the commonly used IP is connected and identify it as the target IP;
[0132] If the number of seller devices is greater than 1, and the number of seller device IPs is equal to 1, then obtain the network IP to which the seller device IP is connected and identify it as the target IP;
[0133] If the number of seller devices is greater than 1, and the number of seller device IPs is greater than 1, then obtain the network IPs to which each seller device IP corresponding to each seller device is connected, and identify them as target IPs respectively;
[0134] Execute risk assessment strategy on target IP.
[0135] The risk determination strategy is specifically as follows:
[0136] Get all device IPs that the target IP has connected to and set them as the judgment IP;
[0137] Obtaining the risk level of each determined IP, wherein the risk level includes high risk and low risk, wherein high risk means that the device IP has been involved in fraudulent behavior in the past, and low risk means that the device IP has not been involved in fraudulent behavior for the time being;
[0138] If there is a high-risk judgment IP, the judgment IP and the target IP are considered abnormal;
[0139] If all the determined IPs corresponding to the target IP are low risk, the target IP is considered normal;
[0140] If the target IP is abnormal, the seller's device IP is determined to be abnormal, and the misjudgment analysis strategy is executed;
[0141] If all target IPs are normal, the seller's device IP is considered normal.
[0142] The misjudgment analysis strategy is specifically as follows:
[0143] Get all the determined IPs corresponding to the target IP;
[0144] Get the number of times each determination IP uses the target IP;
[0145] Sort the IP addresses from large to small based on the number of times used, and set the IP address corresponding to the largest number of times used as the comparison IP address.
[0146] Get the abnormal judgment IP and define it as abnormal IP;
[0147] Get the IP address of the seller's device;
[0148] If the compared IP is consistent with the abnormal IP, the time periods when the seller's device IP and the abnormal IP used the target IP are obtained respectively, and are defined as the seller's use period and the abnormal use period respectively;
[0149] If the seller's usage period has a high overlap with the abnormal usage period, the seller's device IP is determined to be abnormal and the target seller is considered to be high risk;
[0150] If the seller's usage period has a low overlap with the abnormal usage period, the seller's device IP is considered normal and the target seller is considered to have low risk;
[0151] If the compared IP is inconsistent with the abnormal IP, the time period determination strategy is executed;
[0152] The time period determination strategy is specifically as follows:
[0153] Get the seller's usage period and abnormal usage period respectively;
[0154] Obtain the time period when the comparison IP uses the target IP, and define it as the comparison usage period;
[0155] If the comparison usage period has a high overlap with the abnormal usage period, the comparison IP is considered abnormal;
[0156] If the seller's usage period has a high overlap with the abnormal usage period, or the seller's usage period has a high overlap with the comparison usage period, then the seller's device IP is determined to be abnormal and the target seller is considered to have a high risk;
[0157] If the seller's usage period has a low overlap with the abnormal usage period, and the seller's usage period has a low overlap with the comparison usage period, then the seller's device IP is judged to be normal and the target seller is deemed to have a low risk.
[0158] This embodiment also provides that the evaluation strategy is specifically:
[0159] Obtain the determination results of the target seller and the target product respectively;
[0160] If the target product is determined to have low risk, the payment interface will be entered;
[0161] If the target seller is determined to have low risk, the payment interface will be entered;
[0162] If the target product is identified as high risk, the buyer will be notified that the product he / she purchased is risky, i.e., there may be fraud, and the buyer is advised to cancel the order;
[0163] If the target seller is identified as high risk, the buyer will be notified that the merchant corresponding to the product he / she purchased is risky, that is, there may be fraud, and the buyer is advised to cancel the order;
[0164] If the buyer chooses to continue paying, the funds in the target seller's account will be locked after the buyer makes the payment until the buyer confirms that the purchase is correct and unlocks the funds;
[0165] If the buyer reports to the platform that there is a problem with the consumption and provides proof within the time period of locking the funds, the funds will be directly transferred back to the buyer's payment account.
[0166] In this embodiment, a correlation map is formed for data such as the seller's device IP, target IP, judgment IP, abnormal IP and comparison IP. According to the goods purchased by the buyer, the risk situation of the seller corresponding to the goods is automatically obtained. According to the comments of the seller's goods, the authenticity of the goods comments is automatically judged to judge whether the seller and the goods are risky. According to the seller's IP, it is automatically judged whether the seller's sales are normal, and the buyer is reminded that there are risks in this consumption, thereby reducing the risk of buyers being defrauded and preventing buyers from causing property losses.
[0167] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.
[0168] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.
Claims
1. A method for assessing consumption risks on e-commerce platforms. Features: include: Get target information; The target information includes information about the seller and information about the goods purchased by the buyer; The seller information includes the seller's products, review information, and device information; Based on the target information, through information analysis strategies, analytical data is formed to determine whether the comments of the goods purchased by the buyer and other goods of the seller of the goods are valid comments, so as to determine whether the goods purchased by the buyer and the seller of the goods have risks; Based on the analysis data, through the seller judgment strategy, judgment data is formed to further judge the products that are initially judged to be high-risk and the sellers of the products to prevent misjudgment; Based on the judgment data, the seller is evaluated through the evaluation strategy to form evaluation data, so as to determine whether there is a problem with the product or the seller based on further judgment results, so as to warn the buyer and protect the funds; Based on the evaluation data, when there is a problem with the goods purchased by the buyer or the seller of the goods, the buyer will be warned and his funds will be protected.
2. According to the e-commerce platform consumption risk assessment method according to claim 1, Features: The information analysis strategy is specifically as follows: Obtain the goods that the buyer has confirmed the order for and set them as target goods; Get the seller corresponding to the target product and set it as the target seller; Obtain all products of the target seller and set them as the target seller's products; Obtaining review information of each target seller's product, wherein the review information includes a review score and a review text; Get the total score and effective score of the review; Among them, the effective score < the total score of the review × 80%; The target seller's products with review scores ≥ total review scores × 80% are defined as favorable products; The target seller's products with review scores ≥ valid scores and review scores < total review scores × 80% are defined as valid products; The number of target seller products, the number of well-received products, and the number of valid products are obtained respectively, and are defined as the total number of sellers, the number of well-received products, and the number of valid products respectively; If the number of positive reviews ≥ the total number of sellers × 80%, the review judgment strategy is executed for each positive review product; Obtain the quantity of the target seller's products that are determined to have low product risk, and set it as the low-risk quantity; If the low-risk quantity ≥ the effective quantity, the target seller is deemed to have low risk; If the low-risk quantity is less than the effective quantity, the target seller is deemed to have high risk and the target product determination strategy is implemented; If the number of favorable reviews is less than the total number of sellers × 80% and the number of favorable reviews is greater than or equal to the valid number, the target seller is considered to have low risk; If the number of positive reviews is less than the effective number, the target seller is deemed to be high risk and the target product determination strategy is executed.
3. According to the e-commerce platform consumption risk assessment method according to claim 2, Features: The target commodity determination strategy is specifically as follows: Get the review score of the target product and set it as the target score; If the target score is ≥ the total review score × 80%, the review judgment strategy is executed on the target product; If the target score is less than the total score of the review × 80% and the target score is greater than or equal to the effective score, the target product is considered to have low risk; If the target score is less than the effective score, the target product is deemed to be high risk and the seller's judgment strategy is executed.
4. According to the e-commerce platform consumption risk assessment method of claim 3, Features: The comment determination strategy is specifically as follows: Get all the comments of the product; Set comment keywords; Comment text containing comment keywords will be deemed invalid text; Get the number of comments on the product, which is defined as the number of comments; Get the number of invalid words of the product and set it as invalid quantity; If the number of invalid reviews ≥ the number of reviews × 80%, it means that the product has few valid reviews and is considered to have high risk; If (number of reviews - number of invalid reviews) ≥ number of reviews × 80%, it means that the product has many valid reviews and is judged to have low risk.
5. The e-commerce platform consumption risk assessment method according to claim 1, Features: The seller determination strategy is specifically: Obtain the seller's device information, including the seller's device IP and the number of the seller's devices; If the number of seller devices = 1, and the number of seller device IPs = 1, then obtain the network IP to which the seller device IP is connected and identify it as the target IP; If the number of seller devices = 1, and the number of seller device IPs > 1, then obtain the number of times each seller device IP is used, extract the seller device IP corresponding to the largest number of times, and define it as the commonly used IP; Get the network IP to which the commonly used IP is connected and identify it as the target IP; If the number of seller devices is greater than 1, and the number of seller device IPs is equal to 1, then obtain the network IP to which the seller device IP is connected and identify it as the target IP; If the number of seller devices is greater than 1, and the number of seller device IPs is greater than 1, then obtain the network IPs to which each seller device IP corresponding to each seller device is connected, and identify them as target IPs respectively; Execute risk assessment strategy on target IP.
6. According to the e-commerce platform consumption risk assessment method of claim 5, Features: The risk determination strategy is specifically as follows: Get all device IPs that the target IP has connected to and set them as the judgment IP; Obtaining a risk level for each determined IP, wherein the risk level includes high risk and low risk; If there is a high-risk judgment IP, the judgment IP and the target IP are considered abnormal; If all the determined IPs corresponding to the target IP are low risk, the target IP is considered normal; If the target IP is abnormal, the seller's device IP is determined to be abnormal, and the misjudgment analysis strategy is executed; If all target IPs are normal, the seller's device IP is considered normal.
7. The e-commerce platform consumption risk assessment method according to claim 6, Features: The misjudgment analysis strategy is specifically as follows: Get all the determined IPs corresponding to the target IP; Get the number of times each determination IP uses the target IP; Sort the IP addresses from large to small based on the number of times used, and set the IP address corresponding to the largest number of times used as the comparison IP address. Get the abnormal judgment IP and define it as abnormal IP; Get the IP address of the seller's device; If the compared IP is consistent with the abnormal IP, the time periods when the seller's device IP and the abnormal IP used the target IP are obtained respectively, and are defined as the seller's use period and the abnormal use period respectively; If the seller's usage period has a high overlap with the abnormal usage period, the seller's device IP is determined to be abnormal and the target seller is considered to be high risk; If the seller's usage period has a low overlap with the abnormal usage period, the seller's device IP is considered normal and the target seller is considered to have low risk; If the compared IP is inconsistent with the abnormal IP, the time period determination strategy is executed.
8. According to the e-commerce platform consumption risk assessment method according to claim 7, Features: The time period determination strategy is specifically as follows: Get the seller's usage period and abnormal usage period respectively; Obtain the time period when the comparison IP uses the target IP, and define it as the comparison usage period; If the comparison usage period has a high overlap with the abnormal usage period, the comparison IP is considered abnormal; If the seller's usage period has a high overlap with the abnormal usage period, or the seller's usage period has a high overlap with the comparison usage period, then the seller's device IP is determined to be abnormal and the target seller is considered to have a high risk; If the seller's usage period has a low overlap with the abnormal usage period, and the seller's usage period has a low overlap with the comparison usage period, then the seller's device IP is judged to be normal and the target seller is deemed to have a low risk.
9. The e-commerce platform consumption risk assessment method according to claim 1, Features: The evaluation strategy is specifically as follows: Obtain the determination results of the target seller and the target product respectively; If the target product is determined to have low risk, the payment interface will be entered; If the target seller is determined to have low risk, the payment interface will be entered; If the target product is identified as high risk, the buyer will be notified that the product he / she purchased has risks and will be advised to cancel the order; If the target seller is identified as high risk, the buyer will be notified that the merchant corresponding to the product he purchased has risks and the buyer is advised to cancel the order; If the buyer chooses to continue paying, the funds in the target seller's account will be locked after the buyer makes the payment until the buyer confirms that the purchase is correct. If the buyer reports to the platform that there is a problem with the consumption and provides proof within the time period of locking the funds, the funds will be directly transferred back to the buyer's payment account.
Citation Information
Patent Citations
E-commerce false comment judging and recognizing method
CN104881795A
E-commerce credit spreading method based on two-dimensional dynamic display
CN105069627A
Risk assessing method for e-commerce product quality
CN107977798A
Subjective emotion measure-based e-commerce product reliability index and realization method thereof
CN108665339A
Financial anti-fraud analysis method based on big data
CN110111109A
