Commodity price anomaly identification method, device and equipment and storage medium

Abstract

The application relates to a commodity price anomaly identification method, device, equipment and storage medium. The method comprises the following steps: in response to a pricing instruction for a target commodity, extracting sentiment information of a commodity brand to which the target commodity belongs from public opinion text; wherein the current pricing information of the target commodity is included in the pricing instruction; obtaining reference price information of the target commodity; wherein the reference price information comprises an official launch price, a historical transaction price and a historical successful pricing price of the target commodity; and according to the reference price information and the sentiment information, identifying whether the current pricing information is abnormal. In the target commodity pricing process, the sentiment information of the commodity brand to which the target commodity belongs and the reference price information of the target commodity included in the public opinion text are fully utilized, the pricing of the target commodity can be effectively identified whether it has speculation suspicion, and the accuracy of the commodity price anomaly identification result is improved.

Description

Technical Field

[0001] The present application relates to the field of computer processing technology, and in particular to a method, apparatus, device, and storage medium for identifying abnormal commodity prices. Background Art

[0002] On online platforms, sellers can set prices for their products, thereby enabling transactions. However, there are cases where some sellers hype up their products, causing prices to be far higher than the normal price fluctuation range, disrupting the normal trading order of the platform.

[0003] Related technologies can identify price anomalies by combining the cyclical characteristics of commodity prices. However, in practice, the cyclical changes in commodity prices are complex, and it is often difficult to detect cyclical patterns. Therefore, for those skilled in the art, identifying anomalies in commodity prices on online platforms has become a pressing technical problem. Summary of the Invention

[0004] Based on this, embodiments of the present application provide a method, apparatus, device, and storage medium for identifying abnormal commodity prices.

[0005] In a first aspect, an embodiment of the present application provides a method for identifying abnormal commodity prices, comprising:

[0006] In response to a pricing instruction for a target product, extracting sentiment information of the brand of the target product from the public opinion text; wherein the pricing instruction includes current pricing information of the target product;

[0007] Obtaining reference price information of the target product; wherein the reference price information includes the official selling price, historical transaction price, and historical successful pricing price of the target product;

[0008] Whether the current pricing information is abnormal is identified based on the reference price information and the sentiment information.

[0009] In a second aspect, an embodiment of the present application provides a device for identifying abnormal commodity prices, comprising:

[0010] An extraction module is configured to extract sentiment information of the brand of the target product from the public opinion text in response to a pricing instruction for the target product; wherein the pricing instruction includes current pricing information of the target product;

[0011] An acquisition module, configured to acquire reference price information of the target product; wherein the reference price information includes the official selling price, historical transaction price, and historical successful pricing price of the target product;

[0012] An identification module is used to identify whether the current pricing information is abnormal based on the reference price information and the emotional information.

[0013] In a third aspect, an embodiment of the present application provides an electronic device comprising: a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the method for identifying abnormal commodity prices provided in the first aspect of the embodiment of the present application are implemented.

[0014] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the method for identifying abnormal commodity prices provided in the first aspect of the embodiment of the present application are implemented.

[0015] The technical solution provided in the embodiment of the present application extracts sentiment information of the brand of the target product from the public opinion text in response to a pricing instruction for the target product, the pricing instruction including the current pricing information of the target product, and obtains reference price information of the target product. Based on the reference price information of the target product and the sentiment information of the brand of the target product, the solution identifies whether the current pricing information is abnormal. In the target product pricing process, the sentiment information of the brand of the target product and the reference price information of the target product included in the public opinion text are fully utilized to effectively identify whether the pricing of the target product is suspected of speculation, thereby improving the accuracy of the abnormal product price identification results. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A flowchart of a method for identifying abnormal commodity prices provided in an embodiment of the present application;

[0017] Figure 2 A flowchart of a product brand-based sentiment information extraction process provided in an embodiment of the present application;

[0018] Figure 3 Another flowchart of the method for identifying abnormal commodity prices provided in an embodiment of the present application;

[0019] Figure 4 A schematic diagram of the structure of a device for identifying abnormal commodity prices provided in an embodiment of the present application;

[0020] Figure 5 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0021] The present application will be further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are intended only to illustrate the present application and are not intended to limit the present application. It should also be noted that, for ease of description, the accompanying drawings only illustrate portions of the present application, not all of the structures.

[0022] In order to make the purpose, technical solutions and advantages of this application more clear, the technical solutions in the embodiments of this application are further described in detail through the following embodiments and in conjunction with the accompanying drawings. It should be understood that the specific embodiments described here are only used to explain this application and are not used to limit this application.

[0023] It should be noted that the execution subject of the following method embodiment can be a device for identifying abnormal commodity prices, which can be implemented as part or all of an electronic device through software, hardware, or a combination of software and hardware. Optionally, the electronic device can be a smart phone, a tablet computer, an e-book reader, and a car terminal, etc. Of course, the electronic device can also be an independent server or a server cluster, etc. The embodiment of this application does not limit the specific type of electronic device. The following method embodiment is described by taking the execution subject as an electronic device as an example.

[0024] Figure 1 This is a flow chart of a method for identifying abnormal commodity prices provided in an embodiment of the present application. This embodiment relates to the specific process of how an electronic device identifies whether the current price of a target commodity is abnormal.

[0025] like Figure 1 As shown, the method may include:

[0026] S101: In response to a pricing instruction for a target product, extract sentiment information of the brand of the target product from a public opinion text.

[0027] Specifically, the aforementioned public opinion text can include user-generated content within the platform, product reviews, and social media platform public opinion text. For example, public opinion text can include comments and posts on forums, Weibo, shopping websites, and e-commerce platforms. For example, public opinion text can include user evaluation data of products on shopping networks.

[0028] In actual applications, users (e.g., sellers on a shopping platform) can trigger pricing instructions through corresponding controls displayed on the platform to set the price of the target product. The pricing instructions can include the user's current pricing information for the target product. For example, a user can set the price of a pair of shoes on the shopping platform.

[0029] After receiving the pricing instruction for the target product, the electronic device can obtain the public opinion text related to the target product, and extract the emotional information of the brand of the target product from the obtained public opinion text. The public opinion text contains entity information and emotional information. The entity information can be the name of the product or the brand, and the emotional information can represent the characteristics of the brand in a certain attribute, such as the characteristics in terms of quality, price and other attributes. In an optional embodiment, the entity words used to represent the entity information and the attribute words used to represent the emotional information can be determined in the public opinion text to obtain the emotional information of the brand of the target product. The emotional information can be used to effectively discover the recent popularity of the target product, and then predict whether the target product has a tendency to be hyped. For example, if the emotional information finds that the target product has been very popular recently, then the suspicion of hype about the target product is higher. On the contrary, if the emotional information finds that the popularity of the target product is not much different from usual, then the suspicion of hype about the target product is lower.

[0030] S102: Obtain reference price information of the target product.

[0031] Specifically, historical transaction logs are obtained and reference price information of the target product is extracted from the historical transaction logs, wherein the reference price information includes the official recommended price, historical transaction price, and historical successful pricing price of the target product. The reference price information is used to provide a reference basis for the current pricing information of the target product. Based on the difference between the current pricing information and the reference price information, it is possible to predict whether the current pricing information is abnormal. For example, the greater the deviation of the current pricing information from the reference price information, the higher the probability of abnormality of the current pricing information; conversely, the smaller the deviation of the current pricing information from the reference price information, the lower the probability of abnormality of the current pricing information.

[0032] S103: Identify whether the current pricing information is abnormal based on the reference price information and the sentiment information.

[0033] After obtaining the emotional information of the brand to which the target product belongs and the reference price information of the target product, the reference price information and the emotional information of the brand to which the target product belongs are comprehensively utilized to identify whether the current pricing information of the target product is abnormal. It can be understood that the above-mentioned reference price information, as a reference benchmark, provides a relatively reasonable price range for the target product, and the above-mentioned emotional information can reflect the hype trend of the brand to which the target product belongs. The greater the deviation of the current pricing information from the reference price information of the target product and the greater the speculation suspicion of the target product predicted by the emotional information, the greater the possibility that the current pricing information is an abnormal price. Conversely, the smaller the deviation of the current pricing information from the reference price information of the target product and the smaller the speculation suspicion of the target product predicted by the emotional information, the less likely the current pricing information is an abnormal price.

[0034] Of course, as an optional implementation method, corresponding weights can also be set for the above-mentioned reference price information and emotional information, and the above-mentioned reference price information, the above-mentioned emotional information, the weight corresponding to the above-mentioned reference price information, and the weight corresponding to the above-mentioned emotional information can be used to determine whether the current pricing information of the target product is abnormal.

[0035] In the target product pricing process, by introducing public opinion texts and making full use of the public opinion texts posted by users to mine the emotional information of the brand of the target product, the emotional information is used to predict whether the target product is suspected of speculation. By adding emotional information as one of the features to the abnormality identification process of the current pricing of the target product, the recognition results of the abnormal judgment of the current pricing information can be improved.

[0036] After obtaining the recognition result of the current pricing information, corresponding control operations may be performed according to the recognition result of the current pricing information.

[0037] Specifically, if the current pricing information is determined to be an abnormal price, the user's pricing operation for the target product may be rejected, that is, the target product may not be priced according to the current pricing information. Furthermore, a prompt message indicating pricing failure may be output to the user, instructing the user to re-price the target product.

[0038] If the current pricing information is determined to be a normal price, the user's pricing operation for the target product can be approved, that is, the target product is priced according to the current pricing information, and the current pricing information of the target product can be displayed to facilitate subsequent product transactions. Furthermore, a prompt message indicating that the pricing is successful can be output to the user.

[0039] By executing corresponding control operations based on the identification results of current pricing information, accurate prevention and control of abnormal pricing of goods is achieved, thereby maintaining the normal trading order of the platform and preventing the risk of speculation in goods.

[0040] The method for identifying abnormal commodity prices provided in an embodiment of the present application extracts sentiment information of the target commodity's brand from public opinion text in response to a pricing instruction for a target commodity, the pricing instruction including the target commodity's current pricing information, and obtains reference price information of the target commodity. Based on the target commodity's reference price information and the sentiment information of the target commodity's brand, the method identifies whether the current pricing information is abnormal. In the target commodity pricing process, the sentiment information of the target commodity's brand and the reference price information of the target commodity included in the public opinion text are fully utilized to effectively identify whether the target commodity's pricing is suspected of speculation, thereby improving the accuracy of the results of identifying abnormal commodity prices.

[0041] In one embodiment, Figure 2As shown, extracting sentiment information of the brand of the target product from the public opinion text in S101 may include:

[0042] S201. Use crawler technology to obtain public opinion texts from various social platforms.

[0043] With the rapid development of Internet technology, it has become a carrier of vast amounts of information. Through the Internet, users can access a wide range of information, including news, e-commerce, and product reviews. Consequently, crawler technology has emerged. Crawler technology refers to programs or scripts that automatically crawl information on the Internet according to certain rules.

[0044] Since product information can be distributed across various social platforms, such as forums, Weibo, shopping websites, and various e-commerce platforms, crawler technology can be used to crawl product information content on various social platforms to obtain public opinion text. The aforementioned social platforms include both internal and external platforms.

[0045] S202: Pre-process the public opinion text to obtain a processed public opinion text.

[0046] After obtaining the public opinion text, it can be cleaned to filter out some content that is not related to the product brand, such as advertising posts, non-product brand content posts, etc.; further, abnormal characters in the public opinion text can be processed to provide a data basis for subsequent accurate analysis of the public opinion text, and avoid irrelevant content from affecting the public opinion analysis results.

[0047] S203: Input the processed public opinion text into a pre-trained brand sentiment analysis model to obtain sentiment information of the brand of the target product included in the public opinion text.

[0048] Among them, the architecture of the above-mentioned brand sentiment analysis model can be Bert+softmax and Bert+CRF. Bert is a model obtained by pre-training the two tasks of "Fill in the blank task" and "Next sentence prediction". It is only necessary to add a fully connected layer on the basis of Bert and determine the output dimension of the fully connected layer, so that the output of Bert can be mapped to the label set of the labeling problem. The output vector of a single token is then processed by softmax or CRF, and the value of each dimension represents the probability that the token is a certain label. Among them, CRF is a classic probabilistic graph model. In other words, adding softmax or CRF on the basis of the pre-trained model can realize the emotional judgment of product brands and the information labeling of product brand entities based on public opinion texts. Of course, the brand sentiment analysis model can also be implemented by other architectures, and this embodiment does not limit this.

[0049] Taking the brand sentiment analysis model as Bert+softmax and Bert+CRF architecture as an example, after obtaining the processed public opinion text, the processed public opinion text is input into the pre-trained brand sentiment analysis model, and the public opinion text is encoded through the Bert module in the brand sentiment analysis model to obtain the encoding vector corresponding to each token in the public opinion text. Then, the encoding vector is classified and mapped through softmax and CRF to extract the emotional information of the brand of the target product in the public opinion text.

[0050] Optionally, before using crawler technology to obtain public opinion texts from various social platforms, the brand sentiment analysis model can also be trained. Specifically, the brand sentiment analysis model can be trained by the following process:

[0051] Obtain a first training data set; wherein the first training data set includes multiple sample public opinion texts and sentiment information of the product brand included in each sample public opinion text; train a brand sentiment analysis model based on the first training data set.

[0052] Specifically, multiple sample public opinion texts are obtained, and then the product brands and corresponding sentiment polarities in the sample public opinion texts are annotated to obtain the sentiment information of the product brands included in the sample public opinion texts. Next, the multiple sample public opinion texts are used as input, and the sentiment information of the product brands included in the multiple sample public opinion texts is used as the expected output. The actual output corresponding to the multiple sample public opinion texts is determined through the brand sentiment analysis model. The loss value of a preset loss function is determined based on the actual output and the expected output. The loss value is used to optimize the parameters of the brand sentiment analysis model until the loss value reaches the preset convergence condition and remains stable, thereby obtaining a trained brand sentiment analysis model.

[0053] Of course, after obtaining multiple sample public opinion texts, the sample public opinion texts can also be cleaned to filter out some content that is not related to the product brand, such as advertising posts, non-product brand content posts, etc.; further, the abnormal characters in the sample public opinion texts can also be processed to provide a data basis for the subsequent training of the brand sentiment analysis model, thereby ensuring the accuracy of the brand sentiment analysis model.

[0054] In this embodiment, since the brand sentiment analysis model is trained through a large amount of training data, the brand sentiment analysis model can better learn the representation of product brands and emotional information. In this way, by processing the public opinion text of the target product through the brand sentiment analysis model, the accuracy of sentiment analysis based on product brands can be improved.

[0055] In one embodiment, Figure 3As shown, the process of S103 may be:

[0056] S301: Input the current pricing information, reference price information, and sentiment information into a pre-trained anomaly detection model to obtain the anomaly probability of the current pricing information.

[0057] Specifically, the anomaly detection model can be a binary classification tree model, such as the LightGBM algorithm or the XGBoost algorithm. Furthermore, the anomaly detection model is trained based on a second training dataset. The second training dataset includes historical pricing information for multiple sample products, reference price information, sentiment information about the brands of the sample products, and label information indicating whether the historical pricing information is abnormal.

[0058] Among them, when training the anomaly detection model, the historical pricing information, reference price information, and emotional information of the brands of multiple sample products can be used as input, and the label information of whether the historical pricing information of multiple sample products is abnormal can be used as the expected output. The actual output of whether the historical pricing information of multiple sample products is abnormal is determined by the anomaly detection model, and the loss value of the preset loss function is determined based on the actual output and the expected output. The loss value is used to optimize the parameters of the anomaly detection model until the loss value reaches the preset convergence condition and remains stable, thereby obtaining the anomaly detection model.

[0059] After obtaining the trained anomaly detection model, the current pricing information, reference price information and sentiment information of the target product can be input into the anomaly detection model. The anomaly detection model can analyze the gap between the current pricing information and the reference price information, as well as the sentiment information of the brand of the target product, to predict whether the target product is suspected of speculation, thereby obtaining the anomaly probability of the current pricing information.

[0060] S302: Determine whether the abnormal probability of the current pricing information is greater than a preset threshold.

[0061] The preset threshold can be set based on actual needs. If the abnormal probability of the current pricing information is greater than the preset threshold, the following S303 is executed. If the abnormal probability of the current pricing information is less than or equal to the preset threshold, the following S304 is executed.

[0062] S303: Determine that the current pricing information is an abnormal price.

[0063] S304: Determine that the current pricing information is a normal price.

[0064] If the current pricing information is determined to be abnormal, the pricing operation for the target product can be rejected, meaning the target product cannot be priced according to the current pricing information indicated by the user; or the target product can be instructed to be repriced. If the current pricing information is determined to be normal, the pricing operation for the target product can be allowed, meaning the target product can be priced according to the current pricing information indicated by the user, thereby achieving precise control over abnormal pricing operations for products and maintaining normal trading order on the platform.

[0065] In this embodiment, because the anomaly detection model is trained using a large amount of training data, it is able to better learn the mapping relationship between various features and recognition results. Thus, by processing the current pricing information of the target product through the anomaly detection model, the accuracy of identifying price anomalies can be improved. Furthermore, by inputting sentiment information about the target product's brand into the anomaly detection model as a feature, it fully utilizes the public opinion context surrounding the target product, enabling the anomaly detection model to effectively identify whether the target product's pricing is suspected of speculation, thereby further improving the accuracy of identifying price anomalies.

[0066] Figure 4 This is a schematic diagram of a device for identifying abnormal commodity prices provided in an embodiment of the present application. Figure 4 As shown, the device may include: an extraction module 401 , an acquisition module 402 and an identification module 403 .

[0067] Specifically, the extraction module 401 is configured to extract sentiment information of the brand of the target product from the public opinion text in response to a pricing instruction for the target product; wherein the pricing instruction includes current pricing information of the target product;

[0068] The acquisition module 402 is used to obtain reference price information of the target product; wherein the reference price information includes the official recommended price, historical transaction price, and historical successful pricing price of the target product;

[0069] The identification module 403 is used to identify whether the current pricing information is abnormal based on the reference price information and the sentiment information.

[0070] The device for identifying abnormal commodity prices provided in an embodiment of the present application extracts sentiment information of the brand of the target commodity from a public opinion text in response to a pricing instruction for a target commodity, the pricing instruction including the current pricing information of the target commodity, and obtains reference price information of the target commodity. Based on the reference price information of the target commodity and the sentiment information of the brand of the target commodity, the device identifies whether the current pricing information is abnormal. In the target commodity pricing process, the sentiment information of the brand of the target commodity and the reference price information of the target commodity included in the public opinion text are fully utilized to effectively identify whether the pricing of the target commodity is suspected of speculation, thereby improving the accuracy of the results of identifying abnormal commodity prices.

[0071] Based on the above embodiment, optionally, the extraction module 401 is specifically used to use crawler technology to obtain public opinion texts from various social platforms; preprocess the public opinion texts to obtain processed public opinion texts; input the processed public opinion texts into a pre-trained brand sentiment analysis model to obtain the sentiment information of the brand of the target product included in the public opinion text.

[0072] Based on the above embodiment, optionally, the device further includes: a training module.

[0073] Specifically, the training module is used to obtain a first training data set before using crawler technology to obtain public opinion texts from various social platforms; train the brand sentiment analysis model based on the first training data set; wherein, the first training data set includes multiple sample public opinion texts and the emotional information of the product brand included in each sample public opinion text.

[0074] Based on the above embodiment, optionally, the identification module 403 is specifically used to input the current pricing information, the reference price information and the sentiment information into a pre-trained anomaly detection model to obtain the abnormal probability of the current pricing information; when the abnormal probability is greater than a preset threshold, the current pricing information is determined to be an abnormal price; when the abnormal probability is less than or equal to the preset threshold, the current pricing information is determined to be a normal price.

[0075] Based on the above embodiment, optionally, the anomaly detection model is a binary classification tree model, and the anomaly detection model is trained based on a second training data set; wherein, the second training data set includes historical pricing information of multiple sample products, reference price information, sentiment information of the product brands to which the multiple sample products belong, and label information of whether the historical pricing information is abnormal.

[0076] Based on the above embodiment, optionally, the device further includes: a processing module.

[0077] Specifically, the processing module is used to perform corresponding control operations according to the recognition result of the current pricing information.

[0078] Based on the above embodiment, optionally, the public opinion text includes at least one of the following:

[0079] Original content generated by users within the platform, product evaluation information, and public opinion texts on social platforms.

[0080] Figure 5 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application is shown in FIG. Figure 5 As shown, the device includes a processor 510, a memory 520, an input device 530, and an output device 540; the number of the processor 510 in the device can be one or more. Figure 5 A processor 510 is taken as an example; the processor 510, memory 520, input device 530 and output device 540 in the device can be connected via a bus or other means. Figure 5 The bus connection is taken as an example.

[0081] The memory 520, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules, such as the program instructions / modules corresponding to the method for identifying abnormal commodity prices in the embodiments of the present application (for example, the extraction module 401, acquisition module 402, and identification module 403 in the device for identifying abnormal commodity prices). The processor 510 executes the software programs, instructions, and modules stored in the memory 520 to execute the various functional applications and data processing of the above-mentioned device, thereby implementing the above-mentioned method for identifying abnormal commodity prices.

[0082] The memory 520 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system and applications required for at least one function; the data storage area may store data created based on the use of the terminal, etc. In addition, the memory 520 may include a high-speed random access memory and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some instances, the memory 520 may further include a memory remotely located relative to the processor 510, and these remote memories may be connected to the device / terminal / electronic device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0083] The input device 530 may be used to receive input digital or character information and generate key signal input related to user settings and function control of the above-mentioned device. The output device 540 may include a display device such as a display screen.

[0084] In one embodiment, a computer-readable storage medium is further provided, on which a computer program is stored. When the computer program is executed by a processor, it is used to perform a method for identifying abnormal commodity prices. The method includes:

[0085] In response to a pricing instruction for a target product, extracting sentiment information of the brand of the target product from the public opinion text; wherein the pricing instruction includes current pricing information of the target product;

[0086] Obtain reference price information of the target product; wherein the reference price information includes the official recommended price, historical transaction price, and historical successful pricing price of the target product;

[0087] According to the reference price information and the sentiment information, it is identified whether the current pricing information is abnormal.

[0088] Of course, the computer-readable storage medium provided in the embodiment of the present application, when its computer program is executed, is not limited to the method operations described above, but can also execute the relevant operations in the commodity price abnormality identification method provided in any embodiment of the present application.

[0089] Through the above description of the implementation methods, those skilled in the art can clearly understand that the present application can be implemented with the help of software and necessary general-purpose hardware, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0090] It is worth noting that in the embodiment of the above-mentioned search device, the various units and modules included are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application.

[0091] Note that the above are only preferred embodiments of the present application and the technical principles employed. Those skilled in the art will understand that the present application is not limited to the specific embodiments described herein, and that various obvious changes, readjustments, and substitutions can be made by those skilled in the art without departing from the scope of protection of the present application. Therefore, although the present application has been described in more detail through the above embodiments, the present application is not limited to the above embodiments and may include many other equivalent embodiments without departing from the scope of the present application. The scope of the present application is determined by the scope of the appended claims.

Claims

1. A method for identifying abnormal commodity prices, characterized in that: include: In response to a pricing instruction for a target product, extracting sentiment information of the brand to which the target product belongs from the public opinion text; wherein the pricing instruction includes current pricing information of the target product, and the sentiment information is used to indicate a hype trend of the brand to which the target product belongs; Obtain reference price information of the target product; wherein the reference price information includes the official recommended price, historical transaction price, and historical successful pricing price of the target product; Whether the current pricing information is abnormal is identified based on the degree of deviation between the current pricing information and the reference price information and the hype trend of the target commodity predicted by the sentiment information.

2. The method according to claim 1, characterized in that Extracting sentiment information of the brand of the target product from the public opinion text includes: Use crawler technology to obtain public opinion texts from various social platforms; Preprocessing the public opinion text to obtain a processed public opinion text; The processed public opinion text is input into a pre-trained brand sentiment analysis model to obtain the sentiment information of the brand of the target product included in the public opinion text.

3. The method according to claim 2, characterized in that Before using crawler technology to obtain public opinion texts from various social platforms, it also includes: Obtain a first training data set; wherein the first training data set includes a plurality of sample public opinion texts and sentiment information of a product brand included in each sample public opinion text; The brand sentiment analysis model is trained based on the first training data set.

4. The method according to claim 1, wherein The identifying, based on the reference price information and the sentiment information, whether the current pricing information is abnormal includes: Inputting the current pricing information, the reference price information, and the sentiment information into a pre-trained anomaly detection model to obtain an anomaly probability of the current pricing information; When the abnormal probability is greater than a preset threshold, determining that the current pricing information is an abnormal price; When the abnormal probability is less than or equal to the preset threshold, the current pricing information is determined to be a normal price.

5. The method according to claim 4, characterized in that The anomaly detection model is a binary classification tree model, and the anomaly detection model is trained based on a second training data set; The second training data set includes historical pricing information of multiple sample products, reference price information, sentiment information of product brands to which the multiple sample products belong, and label information indicating whether the historical pricing information is abnormal.

6. The method according to any one of claims 1 to 5, characterized in that Also includes: A corresponding control operation is performed according to the recognition result of the current pricing information.

7. The method according to any one of claims 1 to 5, characterized in that The public opinion text includes at least one of the following: Original content generated by users within the platform, product evaluation information, and public opinion texts on social platforms.

8. A device for identifying abnormal commodity prices, characterized in that: include: an extraction module configured to extract, in response to a pricing instruction for a target product, sentiment information of the brand to which the target product belongs from the public opinion text; wherein the pricing instruction includes current pricing information of the target product, and the sentiment information is used to indicate a hype trend of the brand to which the target product belongs; An acquisition module, configured to acquire reference price information of the target product; wherein the reference price information includes the official suggested price, historical transaction price, and historical successful pricing price of the target product; An identification module is used to identify whether the current pricing information is abnormal based on the degree of deviation between the current pricing information and the reference price information and the hype trend of the target commodity predicted by the sentiment information.

9. An electronic device, characterized in that: include: A memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images