Risk control method, device, computing device and computer storage medium

Abstract

The present invention discloses a risk control method, apparatus, computing device, and computer storage medium. The method performs text decomposition processing on the item description information of the item to be detected to obtain a plurality of decomposed texts; randomly combines the plurality of decomposed texts to obtain combined texts; performs filtering processing on the plurality of decomposed texts by single character or single word to obtain the filtered decomposed texts; performs text parsing, word retrieval matching, and verification matching on the combined texts and the filtered decomposed texts to obtain a parsing dimension risk score value, a retrieval dimension risk score value, and a verification dimension risk score value; determines a risk control scheme for the item to be detected according to the parsing dimension risk score value, the retrieval dimension risk score value, and the verification dimension risk score value, realizing the simplification of complex risk control, automatically discovering subjective violations, reducing the misjudgment rate of machine review, and reducing resource consumption, computing power consumption, and power consumption.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and particularly to a risk control method, apparatus, computing device, and computer storage medium. Background Art

[0002] With the development of Internet technology, various e-commerce operation platforms have emerged. The e-commerce operation platforms provide platform channels for merchants to sell items. However, various types of item violation incidents emerge in an endless stream, causing annoyance to each e-commerce operation platform. They can only review the items sold by merchants through various risk control algorithms and means.

[0003] Traditional risk control identification mainly involves: logo identification, similar image identification, misclassified category identification, image infringement identification, etc. for pictures, and for text, the review only stays at category review or brand-related review of a single text description. However, text recognition currently only stays at the recognition of the text itself and completely does not introduce the understanding of text content and comprehensive analysis of the text. There are relatively large problems in the timeliness of traditional risk control identification. On the one hand, it consumes a large amount of server resources, and on the other hand, due to a large amount of computation and cross-server calls, it runs slowly, consumes a large amount of power, and the adverse impact on energy conservation and emission reduction is gradually increasing.

[0004] Manual means are relatively accurate in systematically understanding text content, but the efficiency is extremely low. Therefore, there is an urgent need for a risk control solution that can be efficient and accurate. Summary of the Invention

[0005] In view of the above problems, the present invention is proposed to provide a risk control method, apparatus, computing device, and computer storage medium that can overcome the above problems or at least partially solve the above problems.

[0006] According to one aspect of the present invention, there is provided a risk control method, including:

[0007] Obtaining item description information corresponding to an item to be detected;

[0008] Performing text disassembling processing on the item description information to obtain a plurality of disassembled texts;

[0009] Randomly combining the plurality of disassembled texts to obtain combined texts;

[0010] Performing filtering processing on the plurality of disassembled texts by single characters or single words to obtain filtered disassembled texts;

[0011] Performing text parsing on the combined texts and the filtered disassembled texts, and calculating a parsing dimension risk score value according to the text parsing result;

[0012] Perform word retrieval and matching based on the combined text and the filtered disassembled text, and calculate the risk score value of the retrieval dimension according to the matching result;

[0013] Compare and match the combined text and the filtered disassembled text with the thesaurus of items of the same type, and calculate the risk score value of the verification dimension according to the matching result;

[0014] Determine the risk control plan for the item to be detected according to the risk score value of the analysis dimension, the risk score value of the retrieval dimension, and the risk score value of the verification dimension.

[0015] According to another aspect of the present invention, there is provided a risk control device, including:

[0016] An acquisition module, adapted to acquire the item description information corresponding to the item to be detected;

[0017] A disassembling module, adapted to perform text disassembling processing on the item description information to obtain a plurality of disassembled texts;

[0018] A processing module, adapted to randomly combine the plurality of disassembled texts to obtain a combined text; and perform single-character or single-word filtering processing on the plurality of disassembled texts to obtain the filtered disassembled texts;

[0019] A text analysis module, adapted to perform text analysis on the combined text and the filtered disassembled texts, and calculate the risk score value of the analysis dimension according to the text analysis result;

[0020] A word retrieval and matching module, adapted to perform word retrieval and matching based on the combined text and the filtered disassembled texts, and calculate the risk score value of the retrieval dimension according to the matching result;

[0021] A verification and matching module, adapted to compare and match the combined text and the filtered disassembled texts with the thesaurus of items of the same type, and calculate the risk score value of the verification dimension according to the matching result;

[0022] A determination module, adapted to determine the risk control plan for the item to be detected according to the risk score value of the analysis dimension, the risk score value of the retrieval dimension, and the risk score value of the verification dimension.

[0023] According to still another aspect of the present invention, there is provided a computing device, including: a processor, a memory, a communication interface, and a communication bus, and the processor, the memory, and the communication interface complete communication with each other through the communication bus;

[0024] The memory is used to store at least one executable instruction, and the executable instruction causes the processor to perform the operations corresponding to the above risk control method.

[0025] According to another aspect of the present invention, there is provided a computer storage medium storing at least one executable instruction, and the executable instruction causes a processor to perform operations corresponding to the risk control method as described above.

[0026] The solution provided by the present invention conducts an intelligent analysis and research on the text description of items, gives risk control handling opinions or sends them for review. Ultimately, it simplifies complex risk control, realizes the automatic discovery of subjective violations, reduces the misjudgment rate of machine review, and reduces resource consumption, computing power consumption, and power consumption. It solves the problem that in an e-commerce operation platform, due to the inability of traditional risk control methods to understand and control the ultimate selling purpose of items, some abnormal items appear on the e-commerce operation platform, and the actual target sold has no relation to the hanging pictures, titles, and first-level partial text descriptions, or even no relation at all. The reason for doing this is to escape machine review and traditional risk control screening through such deceptive actions, and ultimately achieve the purpose of selling or trading the target item. Eventually, due to the different actual targets, the e-commerce platform has to bear unnecessary legal joint liability.

[0027] The above description is only an overview of the technical solution of the present invention. In order to be able to understand the technical means of the present invention more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features, and advantages of the present invention more obvious and understandable, the specific embodiments of the present invention are specifically exemplified below. Brief Description of the Drawings

[0028] By reading the detailed description of the preferred embodiments below, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present invention. Moreover, throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0029] Figure 1 A flowchart showing the risk control method according to an embodiment of the present invention is shown;

[0030] Figure 2 A structural diagram showing the risk control device according to an embodiment of the present invention is shown;

[0031] Figure 3 A structural diagram showing the computing device according to an embodiment of the present invention is shown. Detailed Embodiments

[0032] Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present invention can be more thoroughly understood and the scope of the present invention can be fully conveyed to those skilled in the art.

[0033] Figure 1 The flowchart of a risk control method according to an embodiment of the present invention is shown. As Figure 1 shown, the method includes the following steps:

[0034] Step S101, obtaining item description information corresponding to the item to be detected.

[0035] The risk control method provided in this embodiment can be used to perform risk control on the items sold on the e-commerce operation platform, and reduce the risk rate of selling problematic items on the e-commerce operation platform.

[0036] The item description information is description information set for the item, including the title of the item, the text in the picture, the item introduction information, etc. The item description information is the information that the merchant needs to provide when selling the item on the e-commerce operation platform, and is mainly used to describe what kind of item is being sold.

[0037] Step S102, performing text disassembling processing on the item description information to obtain multiple disassembled texts.

[0038] After obtaining the item description information, perform text disassembling processing on the item description information. For example, it can be disassembled based on spaces, English characters, punctuation marks, etc. For example, disassembling "ping is very good, pang haha qiu" to obtain the following disassembled texts "ping", "pang", "qiu", "is very good", "haha". This is only an example and has no limiting effect.

[0039] Step S103, randomly combining the multiple disassembled texts to obtain combined texts.

[0040] After obtaining multiple disassembled texts, the multiple disassembled texts can be randomly combined, for example, two disassembled texts can be combined or three disassembled texts can be combined. More specifically, it can be a combination of three character-type disassembled texts, or a combination of two text-type disassembled texts. Continuing with the above example, the following combined texts "pingpangqiu", "pingqiupang", "pangpingqiu", "pangqiuping", "qiupingpang", "qiupangping", "very good haha", "haha very good", etc. can be obtained by random combination. These are just examples and do not have any limiting effect.

[0041] Step S104, performing single character or single word filtering processing on the multiple disassembled texts to obtain filtered disassembled texts.

[0042] The main purpose here is to filter out single characters or single words. By filtering out single characters and single words, the accuracy of risk control can be improved, and it is avoided that good items are mistakenly classified as bad items and the wrong risk control plan is given. For example, "p" is a single character and "pink" is a single word. This step is to filter out the above characters or words.

[0043] Step S105, performing text analysis on the combined text and the filtered disassembled text, and calculating the analysis dimension risk score value according to the text analysis result.

[0044] This step is mainly to interpret the essential meaning of the text and accurately assess the risks by understanding the text content.

[0045] Specifically, the combined text and the filtered disassembled text are analyzed in the following three dimensions:

[0046] This step mainly performs speech recognition processing, and uses speech as a medium to achieve conversion between Chinese and English, as well as between Chinese, English and the language of the country. In specific implementation, the developed external interface for speech pronunciation and language conversion can be called to perform speech conversion on the combined text and the filtered disassembled text. For example, "pingpangqiu" can be transformed into more than 10 speech variants such as "ping pong", "ping ball fat", etc. And "hahahaha" can be voice-varied into "hey heyHenham". For the words after speech conversion, the meaning of the words also needs to be parsed. Specifically, the converted text is matched with the first vocabulary, and the risk score value of the third parsing dimension is calculated according to the matching results. When it is found that the converted text matches a banned word or a banned word in the first vocabulary, the deduction logic is executed. For example, the total score of the third dimension is 4 points. When the number of matches is less than 3, 2 points of the total 4 points of the current dimension are deducted. When 3 (including) or more different banned words or banned words appear, 4 points are deducted.

[0047] For example, through voice conversion, we get "ping pong". If we recognize words like "ping pong" and find that the merchant does not have the qualifications to sell table tennis products, we will consider that the merchant has violated the regulations.

[0048] Step S106, perform word search and matching based on the combined text and the filtered decomposed text, and calculate the search dimension risk score value based on the matching results.

[0049] Specifically, a semantic analysis is performed on the combined text and the filtered disassembled text to screen out semantically ambiguous texts, wherein semantic ambiguity specifically refers to unclear meaning of the text and inability to determine the meaning that the text is intended to express; the semantically ambiguous text is searched and matched with a second vocabulary to determine a first matching number of the semantically ambiguous text and the second vocabulary; alternatively, a network word search and matching is performed based on the semantically ambiguous text to determine a second matching number of the semantically ambiguous text being a preset type of word; the sum of the first matching number and the second matching number is matched with a quantity interval corresponding to the retrieval dimension risk score value to obtain the retrieval dimension risk score value.

[0050] The second word library records some rare and Internet-famous prohibited words, which are incorporated into the second word library through semantic analysis. For text with unclear semantics, the text with unclear semantics is matched with the second word library. If no match is found, an Internet extension search is performed. For words that cannot be retrieved from both the second word library and the Internet, they are fed back to the business personnel for judgment of rare words and Internet-famous words. For example, words such as "fun egg" and "fat top" can be manually determined and then incorporated into the second word library. Here, the corresponding relationship between the quantity intervals corresponding to the risk score values of the retrieval dimension is preset. For example, 10 points - quantity 0, 4 points - quantity interval [1, 3], 2 points - quantity interval [4, 6], 0 points - quantity interval [7, ∞). This is only an example and has no restrictive effect.

[0051] This can be understood as the re-identification and judgment logic for new Internet-famous words to prevent conversion problems between "good words" and "bad words" due to time changes.

[0052] Step S107: Check and match the combined text and the filtered disassembled text with the same-type item word library, and calculate the risk score value of the check dimension according to the matching result.

[0053] The same-type item word library is a word library corresponding to items of the same category or the same brand as the item to be detected. This word library summarizes some characteristic identification words common to items of the same category or the same brand. Taking table tennis balls as an example, the same-type item word library includes dozens of characteristic identification words such as diameter, elasticity, wind resistance, texture, size, standard, and international standard.

[0054] Check and match the combined text and the filtered disassembled text with the words in the same-type item word library, count the third matching quantity of the combined text and the filtered disassembled text that match the words in the same-type item word library, and match the third matching quantity with the quantity interval corresponding to the risk score value of the check dimension to obtain the risk score value of the check dimension. For example, the following regulations are made: when the matching quantity is greater than or equal to 3, it is considered that there is no difference from the same-type item, and the risk score value of the check dimension is set to 10; when the matching quantity is greater than or equal to 1 and less than 3, it is considered that there is a small difference from the same-type item, and the risk score value of the check dimension is set to 5; when the matching quantity is equal to 0, it is considered that there is a large difference from the same-type item, and the risk score value of the check dimension is set to 0.

[0055] If the combined text and the filtered disassembled text corresponding to the item to be detected are checked and matched with the words in the same-type item word library and the number of matches is 2, then the risk score value of the check dimension can be determined to be 5.

[0056] This embodiment does not limit the execution order of steps S105 - S107, which can be executed simultaneously or successively.

[0057] Step S108: Determine the risk control plan for the item to be detected based on the parsing dimension risk score value, the retrieval dimension risk score value, and the verification dimension risk score value.

[0058] Specifically, the parsing dimension risk score value, the retrieval dimension risk score value, and the verification dimension risk score value are used to perform risk scoring on the item to be detected from three different dimensions. Therefore, the risk control plan for the item to be detected can be determined based on these values. For example, perform a weighted sum operation on the parsing dimension risk score value, the retrieval dimension risk score value, and the verification dimension risk score value to obtain the risk control score value corresponding to the item to be detected. Each risk control plan corresponds to a risk control matching score range. Therefore, after obtaining the risk control score value corresponding to the item to be detected, the risk control score value can be matched with the risk control matching score range corresponding to the risk control plan to determine the risk control matching score range into which the risk control score value corresponding to the item to be detected falls, and determine the risk control plan corresponding to the item to be detected according to the mapping relationship.

[0059] For example, risk scores of 10 points are set for the parsing dimension, the retrieval dimension, and the verification dimension respectively. When the total risk control score value corresponding to the item to be detected is lower than 10 points, it is directly determined as a violation, a warning is issued, and the buyer is restricted from querying and browsing the product. When the total score is higher than 10 points and lower than 20 points, it needs to be sent to the manual review link for further review by an artificial person. When the total score is higher than 20 points, it directly passes the review.

[0060] The present invention is a mechanism that can achieve comprehensive intelligent review of item texts by simulating human intelligence's understanding of things. Through the present invention, scientific and dynamic management and disposal can be carried out on the actual meaning and purpose of item texts. For example, for a text of 30mm, through analysis, this kind of data with specific marks can be understood as length units such as 3cm, 0.3m, 3 centimeters, and 0.3 meters. Then, by comparing this information with the item element information, a determination result of whether the item is in violation or a scoring result can be obtained. On the contrary, traditional risk control methods cannot have more understanding of length units and can even only understand them as a string of ordinary characters.

[0061] The solution provided by the present invention is achieved by conducting intelligent analysis and research on the text description of items. For example, through multi-dimensional vocabulary combination, transfer, comparison, and exploration of online public opinion of unknown vocabulary, etc., it gives risk control handling opinions or sends them for review. Ultimately, it simplifies complex risk control, realizes the automatic identification and disposal of subjective violation problems that evade conventional risk control links, reduces the misjudgment rate of machine review, and reduces resource consumption, computing power consumption, and power consumption. It solves the problem that the e-commerce operation platform cannot understand and control the ultimate selling purpose of items due to traditional risk control methods, resulting in some abnormal items on the e-commerce operation platform. The actual target sold has no relation to the hanging pictures, titles, and first-level part of the text description, or even has no relation at all. The reason for doing this is to escape machine review and traditional risk control screening through such deceptive actions, and ultimately achieve the purpose of selling or trading the target item. Finally, due to the different actual targets, the e-commerce platform has to bear unnecessary legal joint liability.

[0062] Figure 2 FIG. shows a schematic structural diagram of a risk control device according to an embodiment of the present invention. As Figure 2 shown, the device includes:

[0063] An acquisition module 201, adapted to acquire item description information corresponding to an item to be detected;

[0064] A disassembling module 202, adapted to perform text disassembling processing on the item description information to obtain a plurality of disassembled texts;

[0065] A processing module 203, adapted to randomly combine the plurality of disassembled texts to obtain combined texts; and perform filtering processing on single characters or single words of the plurality of disassembled texts to obtain filtered disassembled texts;

[0066] A text parsing module 204, adapted to perform text parsing on the combined texts and the filtered disassembled texts, and calculate an analysis dimension risk score value according to the text parsing result;

[0067] A word retrieval and matching module 205, adapted to perform word retrieval and matching according to the combined texts and the filtered disassembled texts, and calculate a retrieval dimension risk score value according to the matching result;

[0068] A verification and matching module 206, adapted to verify and match the combined texts and the filtered disassembled texts with a thesaurus of items of the same type, and calculate a verification dimension risk score value according to the matching result;

[0069] A determination module 207, adapted to determine a risk control solution for the item to be detected according to the analysis dimension risk score value, the retrieval dimension risk score value, and the verification dimension risk score value.

[0070] Optionally, the text parsing module is further adapted to: compare the combined text and the filtered disassembled text with the item element text information, and calculate the first parsing dimension risk score value according to the comparison result.

[0071] Optionally, the text parsing module is further adapted to: match the combined text and the filtered disassembled text with the first thesaurus, and calculate the second parsing dimension risk score value according to the matching result.

[0072] Optionally, the text parsing module is further adapted to: perform speech recognition processing on the combined text and the filtered disassembled text to obtain a converted text;

[0073] Match the converted text with the first thesaurus, and calculate the third parsing dimension risk score value according to the matching result.

[0074] Optionally, the word retrieval and matching module is further adapted to: perform semantic analysis on the combined text and the filtered disassembled text, and screen out the text with unclear semantics;

[0075] Perform word retrieval and matching on the text with unclear semantics with the second thesaurus to determine the first matching quantity of the text with unclear semantics that matches the second thesaurus; or, perform network word retrieval and matching according to the text with unclear semantics to determine the second matching quantity of the text with unclear semantics that is a preset type of word;

[0076] Match the sum of the first matching quantity and the second matching quantity with the quantity interval corresponding to the retrieval dimension risk score value to obtain the retrieval dimension risk score value.

[0077] Optionally, the verification and matching module is further adapted to: count the third matching quantity of the combined text and the filtered disassembled text that match the words in the same type of item thesaurus;

[0078] Match the third matching quantity with the quantity interval corresponding to the verification dimension risk score value to obtain the verification dimension risk score value.

[0079] Optionally, the determination module is further adapted to: perform weighted summation processing on the parsing dimension risk score value, the retrieval dimension risk score value, and the verification dimension risk score value to obtain the risk control score value corresponding to the item to be detected;

[0080] Match the risk control score value with the risk control matching score interval corresponding to the risk control plan to obtain the risk control plan corresponding to the item to be detected.

[0081] The solution provided by the present invention conducts intelligent analysis and research on the text description of items, gives risk control handling opinions or sends them for review, ultimately simplifies complex risk control, realizes the automatic discovery of subjective violations, reduces the misjudgment rate of machine review, and reduces resource consumption, computing power consumption, and power consumption. It solves the problem that the e-commerce operation platform cannot understand and control the ultimate selling purpose of items due to traditional risk control methods, resulting in some abnormal items on the e-commerce operation platform. The actual target sold has no relation or even no relation at all with the hanging pictures, titles, and first-level part of the text description. The reason for doing this is to escape machine review and traditional risk control screening through such deceptive actions, ultimately achieving the purpose of selling or trading the target. Finally, due to the different actual targets, the e-commerce platform has to bear unnecessary legal joint liability.

[0082] The embodiment of the present application also provides a non-volatile computer storage medium, and the computer storage medium stores at least one executable instruction, and this computer executable instruction can execute the risk control method in any of the above method embodiments.

[0083] Figure 3 The structure diagram of the computing device provided by the embodiment of the present invention is shown. The specific implementation of the computing device in the specific embodiment of the present invention is not limited.

[0084] As Figure 3 shown, the computing device may include: a processor, a communications interface, a memory, and a communication bus.

[0085] Among them: the processor, the communications interface, and the memory complete mutual communication through the communication bus. The communications interface is used to communicate with network elements of other devices such as clients or other servers. The processor is used to execute programs, specifically, it can execute the relevant steps in the risk control method embodiment for the computing device described above.

[0086] Specifically, the program may include program code, and the program code includes computer operation instructions.

[0087] The processor may be a central processing unit CPU, or a specific integrated circuit ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of the present invention. One or more processors included in the computing device may be of the same type of processor, such as one or more CPUs; or they may be of different types of processors, such as one or more CPUs and one or more ASICs.

[0088] A memory for storing programs. The memory may include high-speed RAM memory and may also include non-volatile memory, such as multiple disk memories.

[0089] Specifically, the program can be used to enable the processor to execute the risk control method in any of the above method embodiments. For the specific implementation of each step in the program, reference can be made to the corresponding steps and descriptions in the corresponding units in the above risk control embodiments, which will not be elaborated here. Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described devices and modules can refer to the corresponding process descriptions in the foregoing method embodiments, which will not be repeated here.

[0090] The algorithms or displays provided herein are not inherently related to any particular computer, virtual system, or other device. Various general-purpose systems can also be used in conjunction with the teachings provided herein. Based on the above description, the structures required to construct such systems are obvious. In addition, the embodiments of the present invention are not directed to any specific programming language. It should be understood that the content of the present invention described herein can be implemented using various programming languages, and the descriptions of specific languages above are for disclosing the best mode of the present invention.

[0091] In the specification provided herein, a large number of specific details are set forth. However, it can be understood that the embodiments of the present invention can be practiced without these specific details. In some instances, well-known methods, structures, and technologies have not been shown in detail so as not to obscure the understanding of this specification.

[0092] Similarly, it should be understood that in order to streamline the present invention and assist in understanding one or more of the various inventive aspects, in the foregoing description of the exemplary embodiments of the present invention, the various features of the embodiments of the present invention are sometimes grouped together into a single embodiment, figure, or description thereof. However, the disclosed method should not be construed as reflecting the intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as reflected in the following claims, the inventive aspects lie in less than all the features of the single foregoing disclosed embodiment. Thus, the claims following the detailed description are hereby expressly incorporated into the detailed description, with each claim standing on its own as a separate embodiment of the present invention.

[0093] Those skilled in the art can understand that the modules in the devices in the embodiments can be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments can be combined into one module or unit or component, and in addition, they can be divided into multiple sub-modules or sub-units or sub-components. Except that at least some of such features and / or processes or units are mutually exclusive, any combination can be adopted to combine all the features disclosed in this specification (including the accompanying claims, abstract and drawings) and all the processes or units of any method or device so disclosed. Unless otherwise explicitly stated, each feature disclosed in this specification (including the accompanying claims, abstract and drawings) can be replaced by an alternative feature that provides the same, equivalent or similar purpose.

[0094] In addition, those skilled in the art can understand that although some of the embodiments herein include certain features included in other embodiments rather than other features, the combination of the features of different embodiments means that it is within the scope of the present invention and forms different embodiments. For example, in the following claims, any one of the claimed embodiments can be used in any combination.

[0095] Each component embodiment of the present invention can be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art should understand that a microprocessor or a digital signal processor (DSP) can be used in practice to implement some or all of the functions of some or all of the components according to the embodiments of the present invention. The present invention can also be implemented as a device or apparatus program (for example, a computer program and a computer program product) for executing part or all of the methods described herein. Such a program implementing the present invention can be stored on a computer-readable medium, or can be in the form of one or more signals. Such signals can be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form.

[0096] It should be noted that the above embodiments are illustrative of the present invention and not restrictive thereof, and alternative embodiments can be designed by those skilled in the art without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in the claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present invention can be implemented by means of hardware including several different elements and by means of a suitably programmed computer. In the unit claims listing several devices, several of these devices can be embodied by the same item of hardware. The use of the words first, second, and third, etc. does not denote any order. These words can be interpreted as names. The steps in the above embodiments, unless otherwise specified, should not be construed as limiting the order of execution.

Claims

1. A risk control method, comprising: Obtaining item description information corresponding to the item to be detected; Performing text disassembling processing on the item description information to obtain a plurality of disassembled texts; Performing random combination on the plurality of disassembled texts to obtain a combined text; Performing filtering processing on the plurality of disassembled texts by single character or single word to obtain the filtered disassembled texts; Performing text parsing on the combined text and the filtered disassembled texts, and calculating an analysis dimension risk score value according to the text parsing result; Performing word retrieval matching on the combined text and the filtered disassembled texts, and calculating a retrieval dimension risk score value according to the matching result; Comparing and matching the combined text and the filtered disassembled texts with a thesaurus of the same type of items, and calculating a verification dimension risk score value according to the matching result; Determining a risk control plan for the item to be detected according to the analysis dimension risk score value, the retrieval dimension risk score value, and the verification dimension risk score value; Wherein, the performing word retrieval matching on the combined text and the filtered disassembled texts, and calculating a retrieval dimension risk score value according to the matching result further includes: Performing semantic analysis on the combined text and the filtered disassembled texts, and screening out texts with unclear semantics; Performing word retrieval matching on the texts with unclear semantics and a second thesaurus, and determining a first matching quantity of the texts with unclear semantics that match the second thesaurus; performing network word retrieval matching on the texts with unclear semantics, and determining a second matching quantity of the texts with unclear semantics that are preset type words; Matching the sum of the first matching quantity and the second matching quantity with a quantity interval corresponding to the retrieval dimension risk score value to obtain the retrieval dimension risk score value; The performing text parsing on the combined text and the filtered disassembled texts, and calculating an analysis dimension risk score value according to the text parsing result further includes: Matching the combined text and the filtered disassembled texts with a first thesaurus, and calculating a second analysis dimension risk score value according to the matching result; The calculating a verification dimension risk score value according to the matching result further includes: Counting a third matching quantity of the combined text and the filtered disassembled texts that match the words in the thesaurus of the same type of items; Matching the third matching quantity with a quantity interval corresponding to the verification dimension risk score value to obtain the verification dimension risk score value.

2. The method according to claim 1, wherein, The performing text parsing on the combined text and the filtered disassembled texts, and calculating an analysis dimension risk score value according to the text parsing result further includes: Comparing the combined text and the filtered disassembled texts with item element text information, and calculating a first analysis dimension risk score value according to the comparison result.

3. The method according to claim 1 or 2, wherein The performing text parsing on the combined text and the filtered disassembled texts, and calculating an analysis dimension risk score value according to the text parsing result further includes: Performing speech recognition processing on the combined text and the filtered disassembled texts to obtain a converted text; Matching the converted text with a first thesaurus, and calculating a third analysis dimension risk score value according to the matching result.

4. The method according to claim 1 or 2, wherein, The determination of the risk control plan for the item to be detected based on the parsing dimension risk score value, the retrieval dimension risk score value, and the verification dimension risk score value further includes: Performing a weighted sum processing on the parsing dimension risk score value, the retrieval dimension risk score value, and the verification dimension risk score value to obtain a risk control score value corresponding to the item to be detected; Matching the risk control score value with the risk control matching score range corresponding to the risk control plan to obtain the risk control plan corresponding to the item to be detected.

5. A risk control device, comprising: An acquisition module, adapted to acquire item description information corresponding to an item to be detected; A disassembling module, adapted to perform text disassembling processing on the item description information to obtain a plurality of disassembled texts; A processing module, adapted to randomly combine the plurality of disassembled texts to obtain combined texts; and perform filtering processing on the plurality of disassembled texts by single character or single word to obtain filtered disassembled texts; A text parsing module, adapted to perform text parsing on the combined texts and the filtered disassembled texts, and calculate a parsing dimension risk score value according to the text parsing result; A retrieval matching module, adapted to perform word retrieval matching according to the combined texts and the filtered disassembled texts, and calculate a retrieval dimension risk score value according to the matching result; A verification matching module, adapted to perform verification matching of the combined texts and the filtered disassembled texts with a thesaurus of items of the same type, and calculate a verification dimension risk score value according to the matching result; A determination module, adapted to determine the risk control plan for the item to be detected according to the parsing dimension risk score value, the retrieval dimension risk score value, and the verification dimension risk score value; The retrieval matching module is further adapted to: perform semantic analysis on the combined texts and the filtered disassembled texts, and screen out texts with unclear semantics; Perform word retrieval matching on the texts with unclear semantics and a second thesaurus, and determine a first matching quantity of the texts with unclear semantics matching the second thesaurus; perform network word retrieval matching according to the texts with unclear semantics, and determine a second matching quantity of the texts with unclear semantics being preset type words; Match the sum of the first matching quantity and the second matching quantity with the quantity range corresponding to the retrieval dimension risk score value to obtain the retrieval dimension risk score value; The text parsing module is further adapted to: match the combined texts and the filtered disassembled texts with a first thesaurus, and calculate a second parsing dimension risk score value according to the matching result; The verification matching module is further adapted to: count a third matching quantity of the combined texts and the filtered disassembled texts matching the words in the thesaurus of items of the same type; match the third matching quantity with the quantity range corresponding to the verification dimension risk score value to obtain the verification dimension risk score value.

6. A computing device, comprising: A processor, a memory, a communication interface, and a communication bus, where the processor, the memory, and the communication interface complete communication with each other through the communication bus; The memory is used to store at least one executable instruction, and the executable instruction causes the processor to perform the operations corresponding to the risk control method according to any one of claims 1-4.

7. A computer storage medium storing at least one executable instruction that causes a processor to perform operations corresponding to the risk control method according to any one of claims 1-4.

Images