Data processing method and related device

Abstract

The invention discloses a data processing method and a related device, and the method comprises the steps: obtaining store basic information and interaction information corresponding to a target store, and determining the features of surrounding interest points corresponding to the target store according to the store position information in the store basic information; wherein the store basic information is related to the target store, the interaction information is determined according to multimedia information which is uploaded by a first user and aims at the target store and can reflect feedback of a user side aiming at the target store, and the peripheral interest point features are used for indicating peripheral facilities corresponding to the target store. The condition that the user may go to the target store can be reflected, the target store is predicted based on the information, and the prediction result used for indicating the operation capability of the target store is obtained. Therefore, not only are the peripheral facilities of the store considered, but also the relevant information of the store is integrated, the interactive information capable of reflecting the user side and the information of different dimensions are integrated to more comprehensively predict the operating capacity of the store, and the accuracy of the prediction result is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to a data processing method and related device. BACKGROUND

[0002] A store is the front line of product sales and is one of the key sources of growth in sales performance for many brand owners. Generally, a brand owner can obtain the operating conditions of a store to control the product sales strategy for each store. For example, a brand owner can implement the latest products, marketing activities, sales policies, etc. to the downstream of the channel (i.e. the store), and make timely strategy adjustments according to the store's commodity sales, marketing activity effect, and other closed-loop conditions.

[0003] In commercial operations, brand owners rely on a comprehensive understanding of the geographical environment, business atmosphere, and operating capacity of the store to make various decisions about the store and activities, to ensure the accuracy and effectiveness of the evaluation of the operating capacity of the store. In related technologies, the operating capacity of a store is predicted according to the surrounding facilities of the store, wherein the surrounding facilities can also be referred to as the surrounding points of interest (POI) of the store. By combining the evaluation system of the surrounding POI and other environmental indicators of the store, the key elements in the geographical information are quantified, and a geographical environment theme score with commercial value and easy to understand is formed to predict the operating capacity of the store.

[0004] However, the way of predicting the operating capacity of a store according to the surrounding facilities of the store in related technologies has poor prediction accuracy, and it is difficult to accurately evaluate the operating capacity of the store. SUMMARY

[0005] To solve the above technical problems, the present application provides a data processing method and related device, which comprehensively integrates information of different dimensions to more comprehensively predict the operating capacity of a store, thereby improving the accuracy of the prediction result.

[0006] The embodiments of the present application disclose the following technical solutions:

[0007] In one aspect, the embodiments of the present application provide a data processing method, which comprises:

[0008] obtaining store basic information corresponding to a target store, and obtaining interaction information corresponding to the target store, wherein the store basic information comprises store location information corresponding to the target store and business information of the target store, and the interaction information is determined according to multimedia information uploaded by a first user for the target store;

[0009] determining a surrounding POI feature corresponding to the target store according to the store location information, wherein the surrounding POI feature is used to indicate surrounding facilities corresponding to the target store;

[0010] predict the target store based on the store basic information, the interaction information, and the surrounding point of interest feature, to obtain a prediction result, the prediction result being used to indicate the operating ability of the target store.

[0011] In another aspect, an embodiment of the present application provides a data processing apparatus, the apparatus comprising an acquisition unit, a determination unit, and a prediction unit:

[0012] The acquisition unit is configured to acquire store basic information corresponding to a target store, and acquire interaction information corresponding to the target store, the store basic information comprising store location information corresponding to the target store and business information of the target store, the interaction information being determined according to multimedia information uploaded by a first user for the target store;

[0013] The determination unit is configured to determine, according to the store location information, a surrounding point of interest feature corresponding to the target store, the surrounding point of interest feature being used to indicate surrounding facilities corresponding to the target store.

[0014] The prediction unit is configured to predict the target store based on the store basic information, the interaction information, and the surrounding point of interest feature, to obtain a prediction result, the prediction result being used to indicate the operating ability of the target store.

[0015] In yet another aspect, an embodiment of the present application provides a computer device, the computer device comprising a processor and a memory:

[0016] The memory is configured to store program code, and transmit the program code to the processor;

[0017] The processor is configured to execute the data processing method according to the instructions in the program code.

[0018] As can be seen from the above technical solution, firstly, basic store information and interactive information corresponding to the target store can be obtained. Basic store information can include store location information and the target store's operational information. Next, based on the store location information, the characteristics of surrounding points of interest (POIs) corresponding to the target store can be determined. The basic store information is relevant to the target store itself, thus reflecting its operational capabilities. The interactive information is determined based on multimedia information uploaded by the first user regarding the target store, reflecting user feedback on the target store, and thus reflecting its operational capabilities. Furthermore, the surrounding POI characteristics can indicate the surrounding facilities of the target store, reflecting the likelihood of users visiting the target store, and thus reflecting its operational capabilities. Accordingly, based on the basic store information, interactive information, and surrounding POI characteristics, predictions can be made about the target store, yielding prediction results that can be used to indicate the target store's operational capabilities. Compared to related technologies that predict a store's operational capabilities based on its surrounding facilities, this application not only considers the store's surrounding facilities but also integrates relevant information about the store itself, as well as interactive information that reflects user interactions. By integrating information from different dimensions, a more comprehensive prediction of the store's operational capabilities can be made, which helps improve the accuracy of the prediction results and can better guide the store's relevant decisions. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A flowchart illustrating a data processing method provided in an embodiment of this application;

[0021] Figure 2 A schematic diagram of a data processing architecture provided in an embodiment of this application;

[0022] Figure 3 This is a structural diagram of a data processing apparatus provided in an embodiment of this application. Detailed Implementation

[0023] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0024] The methods used in related technologies to predict a store's operating capacity based on its surrounding facilities are limited by the fact that these facilities can only reflect a store's sales performance to a certain extent, resulting in data limitations and superficiality. It is precisely this limitation that makes it difficult to accurately predict a store's operating capacity solely based on geographical features such as its surrounding POIs.

[0025] To this end, this application provides a data processing method and related apparatus that not only considers the surrounding facilities of the store, but also integrates relevant information of the store itself, as well as interactive information that reflects the user side. In this way, by integrating information from different dimensions, the limitations of the singleness and superficiality of data are overcome, so as to more comprehensively predict the store's operating capabilities, which is conducive to improving the accuracy of the prediction results and thus better guiding the store's relevant decisions.

[0026] The data processing method provided in this application can be implemented using a computer device, which can be a terminal device or a server. The server can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services. Terminal devices include, but are not limited to, mobile phones, computers, smart voice interaction devices, smart home appliances, and in-vehicle terminals. Terminal devices and servers can be directly or indirectly connected via wired or wireless communication, and this application does not impose any limitations in this regard.

[0027] The following examples illustrate this in detail:

[0028] Figure 1 A flowchart of a data processing method provided in this application embodiment, using a server as an example of the aforementioned computer device, is used for illustration. The method includes S101-S103:

[0029] S101: Obtain the basic information of the target store and the interactive information of the target store.

[0030] The basic store information includes the target store's location and operating information. Location information indicates the store's geographical location, while operating information indicates its business performance. For example, location information might include the province, city, district, street, community, and address of the target store. Operating information might include operating hours, store size, channel type, whether it's a chain, geographical distribution, surrounding business district, popularity rating, map store rating, and products sold. In practice, basic store information can be obtained through publicly available data (such as the target store's registration information).

[0031] Furthermore, the interactive information is determined based on the multimedia information uploaded by the first user targeting the target store, reflecting the user's feedback regarding that target store. The first user can refer to the user who uploaded the multimedia information, such as a consumer, a brand representative, or a salesperson from the target store. The multimedia information can be at least one of the following: text comments about the target store, image information about the target store, etc.

[0032] The review text information is text-based, specifically referring to comments from a first-time user (e.g., a consumer) regarding the target store. The image information is images, specifically photos taken by a first-time user (e.g., a salesperson) about the target store, such as photos of the storefront, shelves, posters, and surrounding area. Typically, this image information reflects real-world information such as the store's decor, product display, layout, and marketing activities, thus reflecting the actual situation of the target store and can be used to predict its operational capabilities.

[0033] It should be noted that this application does not impose any limitations on the method of obtaining interactive information. For ease of understanding, this application provides the following illustrative example using multimedia information, including review text information and image information of the target store:

[0034] If multimedia information includes both types of information, then when determining interactive information, firstly, text semantic tags can be determined based on the comment text information, and image semantic tags can be determined based on the image information. The text semantic tags indicate the semantics expressed by the comment text information, reflecting the content it intends to convey, while the image semantic tags indicate the semantics expressed by the image information, reflecting the content it intends to convey. Next, if the semantics of the text semantic tags and image semantic tags do not overlap, it indicates that they convey different content. In this case, the text semantic tags and image semantic tags can be determined as interactive information. For example, they might express different aspects of a target store, such as the comment text information about the target store's products, and the image information about the target store's services.

[0035] Correspondingly, if the semantics of text semantic tags and image semantic tags contradict each other, it indicates that they both intend to express the same aspect of information regarding the target store, such as the store's area. Therefore, to ensure data accuracy, the higher confidence level can be selected as the interaction information to combine the confidence levels of both. Specifically, firstly, the number of first tags corresponding to the text semantic tags and the first number of collected comment text information corresponding to the text semantic tags can be obtained, as well as the number of second tags corresponding to the image semantic tags and the second number of collected image information corresponding to the image semantic tags. Next, the first confidence level corresponding to the text semantic tags can be determined based on the number of first tags and the first number of collected images, and the second confidence level corresponding to the image semantic tags can be determined based on the number of second tags and the second number of collected images. Finally, if the first confidence level is greater than the second confidence level, the text semantic tag is determined as the interaction information; if the first confidence level is less than the second confidence level, the image semantic tag is determined as the interaction information. Based on this, for information expressing the same aspect, confidence levels are considered, and the one with higher confidence is selected as the interaction information to ensure prediction accuracy.

[0036] The number of tags can refer to the total number of tags. For example, the first tag number can refer to the total number of text semantic tags, and the second tag number can refer to the total number of image semantic tags. For instance, if the text semantic tag is "the target store has a large store area" and the image semantic tag is "the target store has a small store area", then the two are semantically contradictory. Based on all the comment text information, 11 of these text semantic tags are extracted, and based on all the image information, 12 of these image semantic tags are extracted. That is, the number of first tags is 11, and the number of second tags is 12.

[0037] The collection quantity indicates the total number of information items extracted with the tag. For example, the first collection quantity refers to the total number of comment text information items with the text semantic tag extracted, and the second collection quantity refers to the total number of image information items with the image semantic tag extracted. In other words, the collection quantity reflects the number of data source channels. For example, if all four comment text information items from the first user can extract the text semantic tag "the target store has a large store area", then the first collection quantity is four. Similarly, if all four image information items taken by the first user can extract the image semantic tag "the target store has a small store area", then the second collection quantity is four.

[0038] In practical applications, a larger number of tags indicates that the information about the target store in that aspect is more likely to be reflected by a larger number of semantic tags. Therefore, the confidence level of semantic tags can be determined based on the number of tags and the amount of data collected. In the example above, the first confidence level is lower than the second confidence level, so "the target store has a small store area" can be used as the final interactive information to participate in the subsequent prediction of the target store's operating capabilities.

[0039] It should also be noted that this application does not impose any limitations on how to determine whether the semantics of text semantic tags and image semantic tags overlap. For ease of understanding, the following methods are provided as examples in the embodiments of this application:

[0040] In practical applications, the collected image information typically includes multiple images. Preprocessing can begin, including resizing the images to fit the model input, normalizing pixel values ​​to eliminate differences in lighting and contrast, and performing image enhancement if necessary to improve image quality. After preprocessing, the image sequence v of the target store is obtained.

[0041] Similarly, the collected comment text information typically includes multiple comments. It can be preprocessed to clean the text data, including removing unnecessary elements such as spaces, stop words, punctuation, and numbers, and performing word segmentation. The collected comment text information refers to comments gathered from various stores, with the aim of extracting general text tags relevant to describing each store. Specifically, based on a pre-trained model like Bidirectional Encoder Representations from Transformers (BERT), keyword extraction, sentiment analysis, and topic word extraction can be performed on each comment text. After deduplicating all user comment texts, a text sequence t for each store is obtained. This text sequence t can also be called the text tag library for each store, which can include multiple general text tags that can be directly used for each store later. This approach addresses the issue of some stores having only a small number of comment texts, resulting in insufficient extraction of rich text tags.

[0042] Next, for the input image sequence v, the powerful visual feature extraction capability of the image encoder in the pre-trained multimodal pre-trained neural network (Contrastive Language-Image Pre-Training, CLIP) is utilized to process each input image v. i Transform into the corresponding feature token sequence, denoted as Where n is the dimension of the image features. In practical applications, the ViT (Vision Transformer) model can be used to process the image first to better extract its semantics, and then processed by the CLIP model to map the information from these two different modalities, images and text, to the same semantic space.

[0043] Furthermore, for the input text sequence t, the text input t is processed using the pre-trained text encoder in CLIP. i Transform it into the corresponding feature token sequence, denoted as Here, n represents the dimension of the text features. Based on this, encoding processing of images and text is implemented. In practical applications, the processed features can be embedded vectors. Furthermore, both are encoded based on the CLIP model, thus mapping text and images, which originally belong to different types of information, to the same semantic space.

[0044] Next, based on the similarity calculation module, the cosine distance between the image features and text features can be calculated as a similarity score. Based on this, a symmetric similarity matrix can be obtained between the text feature sequence t and the image feature sequence v. Specifically, the similarity score can be determined using the following formula:

[0045]

[0046] In the above formula, sim(v i ,t i ) is used to represent v i With t i The similarity between them, v i t is used to represent the i-th image feature. i The term is used to represent the i-th text feature, and <·> is used to represent inner product calculation.

[0047] For a target store, for example, if we can identify B image features and B text features, where i is greater than or equal to 1 and less than or equal to B, where B is a positive integer, then we can use symmetric cross-entropy loss based on similarity scores to train the model's parameters. Specifically:

[0048]

[0049] L = L v2t +L t2v

[0050] In the above formula, L v2t Used to represent the loss from image features to text features, i.e., video-to-text loss, L t2v The loss is used to represent the loss from text features to image features, i.e., the text-to-video loss. The total loss is L, which is the sum of the losses from these two parts. Also, j takes values ​​greater than or equal to 1 and less than or equal to B, and i and j can take the same value.

[0051] During training, by minimizing the loss L, the model is able to find the text feature with the highest similarity to the image feature from the text feature sequence for the input image features, and assign the found text label with the highest similarity to the image as the image semantic label.

[0052] S102: Based on the store location information, determine the characteristics of the surrounding points of interest corresponding to the target store.

[0053] The surrounding points of interest (POIs) feature is used to indicate the surrounding facilities of the target store. In practical applications, the surrounding facilities of the target store can also be referred to as surrounding POIs. It should be noted that this application does not impose any limitations on the method for determining the surrounding POI features. For ease of understanding, the following methods are provided as examples in this application's embodiments:

[0054] In one possible implementation, firstly, based on the store's location information, various types of surrounding facilities corresponding to the target store can be determined. Next, the number of facilities for each type and the distance from each facility to the target store can be determined. Understandably, the number of facilities and their distance to the target store both affect the likelihood of nearby users visiting that store. Therefore, based on the number of facilities for each type and the distance from each facility to the target store, the characteristics of surrounding points of interest can be determined. Based on this, by comprehensively considering the number of different types of surrounding facilities and their distance to the store, the accuracy of the surrounding point of interest characteristics can be improved, thereby enhancing the accuracy of predicting operational capabilities.

[0055] In practical applications, different types of facilities offer varying benefits to users reaching their target stores. Therefore, one possible approach is to comprehensively consider the correlation between facility types and target stores to determine the weights of each type of facility, thereby identifying more accurate characteristics of surrounding points of interest.

[0056] In practical implementation, the weight of each type of surrounding facility can be determined based on the store type of the target store and the facility type of each type. The weight indicates the degree of correlation between each type of surrounding facility and the target store, and the weight is positively correlated with the degree of correlation. Next, the characteristics of surrounding points of interest can be determined based on the number of facilities corresponding to each type of surrounding facility, the weight of each type of surrounding facility, and the distance of each surrounding facility to the target store. Based on this, considering the correlation between various types of facilities and the target store, a weight is assigned to each type of facility. The higher the weight of the surrounding facility, the more numerous it is, and the shorter the distance to the store, indicating that the user is more likely to go to the target store, and correspondingly, the higher the value of the surrounding point of interest characteristic.

[0057] In practical applications, if the first type of surrounding facilities are associated with the target store, and the second type of surrounding facilities are mutually exclusive with the target store, then the weight of the first type of surrounding facilities is greater than the weight of the second type of surrounding facilities.

[0058] For example, various surrounding facilities can include different amenities such as residences, shopping, restaurants, tourism, hospitals, education, living facilities, transportation facilities, workplaces, and entertainment facilities. If the target store mainly sells products related to children, then the greater the relevance of facilities such as hospitals and residences to the target store, the higher their weight can be assigned. Conversely, if the target store's competing facilities are mutually exclusive with it, their weight will be lower and can even be set to negative numbers. In other words, the weight value can be a number between -1 and 1.

[0059] In another possible implementation, the number of Points of Interest (POIs) can be considered to determine the store POI index corresponding to the target store, and the distance to the store can be considered to determine the distance index corresponding to the target store. That is, the aforementioned surrounding point of interest characteristics can include both the store POI index and the distance index. The store POI index focuses more on the number of POIs, while the distance index focuses more on the distance to the store. Based on this, the surrounding facilities of the target store are evaluated from both the quantity and distance dimensions, resulting in a more comprehensive assessment.

[0060] S103: Based on the store's basic information, interaction information, and features of surrounding points of interest, predict the target store and obtain the prediction results.

[0061] Among these, basic store information is relevant to the target store itself, thus reflecting its operational capabilities. Interactive information, determined based on multimedia information uploaded by the first user regarding the target store, reflects user feedback and thus also reflects its operational capabilities. Furthermore, surrounding points of interest characteristics indicate nearby facilities and suggest potential user visits to the target store, further reflecting its operational capabilities. Accordingly, predictions can be made based on these different levels of information, and the resulting predictions can be used to indicate the target store's operational capabilities.

[0062] Compared to related technologies that predict a store's operational capabilities based on its surrounding facilities, this application not only considers the store's surrounding facilities but also integrates relevant information about the store itself, as well as interactive information that reflects user interactions. By integrating information from different dimensions, a more comprehensive prediction of the store's operational capabilities can be made, which helps improve the accuracy of the prediction results and can better guide the store's relevant decisions.

[0063] It should be noted that this application does not impose any limitations on the method of making predictions to obtain the prediction results. For ease of understanding, the following embodiments of this application are provided as examples:

[0064] In another possible implementation, the accessibility of the target store can be comprehensively considered to further improve the accuracy of the prediction. That is, in specific implementations, S103 may also include:

[0065] First, based on the store's location information, the reachable area of ​​the target store can be determined. Second, the time required for a user to reach the target store from any location within this reachable area must be less than or equal to a preset time. Here, "second user" refers to any user; the reachable area identifies the geographical area within which users can reach the store within the preset time. It's understood that a larger reachable area indicates a higher likelihood of more users visiting the store, thus reflecting the store's operational capacity. Therefore, predictions about target stores can be made based on basic store information, interaction information, surrounding points of interest characteristics, and the reachable area. Based on this, and by integrating the reachable area information into the aforementioned data, the prediction becomes more comprehensive and accurate.

[0066] The preset duration can be set according to actual conditions, and this application does not impose any limitations. For example, it can be set to 15 minutes, such as 15 minutes of walking, 15 minutes of public transportation, 15 minutes of driving, etc. Alternatively, different preset durations can be set for different modes of transportation, such as 15 minutes of walking, 20 minutes of public transportation, 25 minutes of driving, etc.

[0067] When determining the reachable area of ​​a store, the initial reachable area can be determined first, centered on the target store, based on its location information. The time required for a second user to reach the target store from any location within this initial reachable area must be less than or equal to a preset time. In other words, the initial reachable area is determined based on the time required for the second user to reach the store. Next, the built-up area of ​​the city where the target store is located can be obtained. Since the built-up area accurately indicates the actual geographical situation, the initial reachable area and the built-up area can be overlaid to determine the overlapping area as the reachable area of ​​the store.

[0068] Based on this, the determined store reach area not only meets the requirement that the user's visit time is within the preset time, but also represents a real geographical area, thus more accurately reflecting the reach of the target store.

[0069] In practical applications, the latitude and longitude information of the initial area boundary can be obtained. Typically, this data is provided as a list of coordinate points, representing the vertices of the boundary. Then, the latitude and longitude points of the boundary are connected to form a closed polygon, and the area of ​​this polygon is calculated. Correspondingly, the area of ​​the overlapping area can be calculated to visually reflect the size of the store's reachable area using area.

[0070] In one possible implementation, the aforementioned surrounding facilities are located within the target area, which in turn is located within the store's reachable area, and the target store itself is located within the target area. In other words, for the target store, the store's reachable area is a relatively large area, while the surrounding area is a smaller area, but both revolve around the target store. Thus, considering the surrounding facilities closer to the target store and other reachable factors within a larger area allows for a more accurate prediction of the store's operational capabilities.

[0071] As can be seen from the above technical solution, firstly, basic store information and interactive information corresponding to the target store can be obtained. Basic store information can include store location information and the target store's operational information. Next, based on the store location information, the characteristics of surrounding points of interest (POIs) corresponding to the target store can be determined. The basic store information is relevant to the target store itself, thus reflecting its operational capabilities. The interactive information is determined based on multimedia information uploaded by the first user regarding the target store, reflecting user feedback on the target store, and thus reflecting its operational capabilities. Furthermore, the surrounding POI characteristics can indicate the surrounding facilities of the target store, reflecting the likelihood of users visiting the target store, and thus reflecting its operational capabilities. Accordingly, based on the basic store information, interactive information, and surrounding POI characteristics, predictions can be made about the target store, yielding prediction results that can be used to indicate the target store's operational capabilities. Compared to related technologies that predict a store's operational capabilities based on its surrounding facilities, this application not only considers the store's surrounding facilities but also integrates relevant information about the store itself, as well as interactive information that reflects user interactions. By integrating information from different dimensions, a more comprehensive prediction of the store's operational capabilities can be made, which helps improve the accuracy of the prediction results and can better guide the store's relevant decisions.

[0072] The above embodiments have provided a detailed description of the data processing method and related apparatus provided in this application. For a further understanding, the embodiments of this application also provide, for example... Figure 2 The diagram shown illustrates a data processing architecture, specifically:

[0073] The acquisition and calculation of POIs around the store are used to determine the aforementioned features of surrounding points of interest. The acquisition and calculation of the store's reachability circle are used to determine the aforementioned reachable area of ​​the store. Basic store information includes the aforementioned store location information, business information, etc. Store matching tags can be used to determine the aforementioned interactive information, specifically based on text and image modalities. This includes two parts: acquiring and preprocessing store review text, then determining store tag feature descriptions and extracting corresponding text features; and acquiring and preprocessing images of the store's interior and exterior, processing them based on flat subgraph sequences and linear projection layers, and processing them based on the ViT model to extract corresponding image features. Finally, the text and image features are processed based on the CLIP model to determine matching text-image pairs. It is understood that all of this content can be found in the detailed descriptions of the aforementioned related embodiments, and will not be repeated here.

[0074] In practical applications, store tags can be compiled based on the specific circumstances of each store. As an example, store tags can be found in Table 1 below:

[0075] Table 1. Examples of store labels

[0076] Operating information label Door head high door face wide area large display neat poster clear, etc. Location information label Pedestrian-intensive intersection, roadside, alley, inside the mall, market, and community Business district information label High-end community, villa area, and high-rise buildings Type information label General store, specialty store, supermarket, convenience store, and shopping mall Product information label SKU on the shelf, product category, and brand Accessible circle label 15-minute walk, 15-minute public transportation, and 15-minute drive POI around the label School, community, hospital, office building, mall, scenic spot, hotel, and factory

[0077] In Table 1 above, the business district information tag, surrounding POI tag, and type information tag can all be used to obtain information such as which POIs are nearby, their quantity, and distance, which can guide the determination of the aforementioned surrounding points of interest characteristics. The accessibility tag can be used to indicate the set accessibility range that needs to be considered, and can guide the determination of the aforementioned store accessibility area. The location information tag can guide the acquisition of the aforementioned store location information, and the business information tag and product information tag can guide the acquisition of the aforementioned business information, etc.

[0078] Furthermore, the scorecard model can be used to make predictions based on input data and output predicted scores. These scores can be used to indicate a store's operational capabilities in a more intuitive way. For a specific store, basic store information, interaction information, surrounding points of interest features, and the store's reachable area can be input into the scorecard model to output a score for that store, thus intuitively indicating its operational capabilities.

[0079] In practical applications, variable screening can be used to identify the most significant indicators affecting store operational capabilities through statistical methods. This primarily involves univariate feature selection and machine learning-based methods. Furthermore, the model development phase can include three parts: variable segmentation, WOE (Weight of Evidence) transformation of variables, and logistic regression estimation. Based on these, more accurate weights can be determined. The model can also be evaluated to assess its discriminative power, predictive ability, and stability, generating a model evaluation report to determine its usability. Finally, scorecards can be generated, using methods such as logistic regression coefficients and WOE to determine store operational capability scores, converting the logistic model into a standardized scoring format. Finally, a scoring system can be established (deployed online), and an automatic scoring system for store operational capabilities can be built based on the generated scorecards. This allows for timely and flexible adjustments to store sales strategies and marketing activities.

[0080] It is understood that this basically corresponds to the method embodiment, so relevant details can be found in the description of the method embodiment.

[0081] Figure 3 This is a structural diagram of a data processing apparatus provided in an embodiment of this application. The apparatus includes an acquisition unit 301, a determination unit 302, and a prediction unit 303.

[0082] The acquisition unit 301 is used to acquire basic store information corresponding to the target store and interactive information corresponding to the target store. The basic store information includes the store location information and business information of the target store. The interactive information is determined based on the multimedia information of the target store uploaded by the first user.

[0083] The determining unit 302 is used to determine the surrounding point of interest features corresponding to the target store based on the store location information, and the surrounding point of interest features are used to indicate the surrounding facilities corresponding to the target store.

[0084] The prediction unit 303 is used to predict the target store based on the store's basic information, the interaction information, and the features of the surrounding points of interest, and to obtain a prediction result, which is used to indicate the operating capacity of the target store.

[0085] In one possible implementation, the prediction unit is further configured to:

[0086] Based on the store location information, the reachable area of ​​the target store is determined, and the time required for the second user to reach the target store from any location within the reachable area is less than or equal to a preset time.

[0087] Based on the store's basic information, the interaction information, the features of surrounding points of interest, and the store's reachable area, the target store is predicted to obtain the prediction result.

[0088] In one possible implementation, the prediction unit is further configured to:

[0089] Based on the store location information, an initial area range corresponding to the target store is determined with the target store as the center. The time required for the second user to reach the target store from any location within the initial area range is less than or equal to the preset time.

[0090] Obtain the urban built-up area of ​​the city where the target store is located;

[0091] The initial area range is superimposed with the urban built-up area range, and the overlapping area range is determined as the reachable area range of the store.

[0092] In one possible implementation, the surrounding facilities are located within a target area, the target area is within the reachable area of ​​the store, and the target store is located within the target area.

[0093] In one possible implementation, if the multimedia information includes review text information and image information of the target store, the acquisition unit is further configured to:

[0094] The text semantic tags are determined based on the comment text information, and the image semantic tags are determined based on the image information;

[0095] If the semantics of the text semantic tag and the image semantic tag do not overlap, the text semantic tag and the image semantic tag are determined as the interaction information.

[0096] In one possible implementation, the acquisition unit is further configured to:

[0097] If the semantics of the text semantic tag and the image semantic tag are contradictory, obtain the first number of tags corresponding to the text semantic tag and the first number of collected comment text information to which the text semantic tag belongs, and obtain the second number of tags corresponding to the image semantic tag and the second number of collected image information to which the image semantic tag belongs;

[0098] A first confidence level corresponding to the text semantic label is determined based on the first number of labels and the first number of acquisitions, and a second confidence level corresponding to the image semantic label is determined based on the second number of labels and the second number of acquisitions.

[0099] If the first confidence level is greater than the second confidence level, the text semantic tag is determined to be the interaction information;

[0100] If the first confidence level is less than the second confidence level, the image semantic label is determined as the interaction information.

[0101] In one possible implementation, the determining unit is further configured to:

[0102] Based on the store location information, determine the various types of surrounding facilities corresponding to the target store;

[0103] Determine the number of facilities corresponding to each type of surrounding facility, and determine the distance from each of the surrounding facilities to the target store;

[0104] The features of the surrounding points of interest are determined based on the number of facilities corresponding to the various types of surrounding facilities and the distance of each surrounding facility from the target store.

[0105] In one possible implementation, the determining unit is further configured to:

[0106] Based on the store type of the target store and the facility type of each type of surrounding facility, a weight is determined for each type of surrounding facility. The weight is used to indicate the degree of association between each type of surrounding facility and the target store, and the weight is positively correlated with the degree of association.

[0107] The features of the surrounding points of interest are determined based on the number of facilities corresponding to each of the various types of surrounding facilities, the weight of each of the various types of surrounding facilities, and the distance of each of the surrounding facilities to the target store.

[0108] As can be seen from the above technical solution, firstly, basic store information and interactive information corresponding to the target store can be obtained. Basic store information can include store location information and the target store's operational information. Next, based on the store location information, the characteristics of surrounding points of interest (POIs) corresponding to the target store can be determined. The basic store information is relevant to the target store itself, thus reflecting its operational capabilities. The interactive information is determined based on multimedia information uploaded by the first user regarding the target store, reflecting user feedback on the target store, and thus reflecting its operational capabilities. Furthermore, the surrounding POI characteristics can indicate the surrounding facilities of the target store, reflecting the likelihood of users visiting the target store, and thus reflecting its operational capabilities. Accordingly, based on the basic store information, interactive information, and surrounding POI characteristics, predictions can be made about the target store, yielding prediction results that can be used to indicate the target store's operational capabilities. Compared to related technologies that predict a store's operational capabilities based on its surrounding facilities, this application not only considers the store's surrounding facilities but also integrates relevant information about the store itself, as well as interactive information that reflects user interactions. By integrating information from different dimensions, a more comprehensive prediction of the store's operational capabilities can be made, which helps improve the accuracy of the prediction results and can better guide the store's relevant decisions.

[0109] In another aspect, embodiments of this application provide a computer device, the computer device including a processor and a memory:

[0110] The memory is used to store program code and transmit the program code to the processor;

[0111] The processor is used to execute the data processing method provided in the above embodiments according to the instructions in the program code.

[0112] The computer device may include a terminal device or a server, and the aforementioned data processing device may be configured in the computer device.

[0113] In another aspect, embodiments of this application also provide a storage medium for storing a computer program for executing the data processing method provided in the above embodiments.

[0114] In addition, this application also provides a computer program product including instructions, which, when run on a computer, causes the computer to execute the data processing method provided in the above embodiments.

[0115] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium can be at least one of the following media: read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk, etc., and other media capable of storing program code.

[0116] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0117] It should be noted that, in this document, relational terms such as "first" and "second," if present, are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0118] The data processing method and related apparatus provided in the embodiments of this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are only for the purpose of helping to understand the method of this application. At the same time, those skilled in the art will recognize that there will be changes in the specific implementation methods and application scope based on the method of this application.

[0119] In summary, the content of this specification should not be construed as limiting this application. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the protection scope of this application. Furthermore, based on the implementation methods provided in the above aspects, this application can be further combined to provide more implementation methods.

Claims

1. A data processing method, characterized in that, The method includes: The system obtains basic store information corresponding to the target store and interactive information corresponding to the target store. The basic store information includes the store location information and the business information of the target store. The interactive information is determined based on the multimedia information of the target store uploaded by the first user. Based on the store location information, the surrounding points of interest features corresponding to the target store are determined, and the surrounding points of interest features are used to indicate the surrounding facilities corresponding to the target store. Based on the store's basic information, the interaction information, and the features of the surrounding points of interest, a prediction is made about the target store to obtain a prediction result, which is used to indicate the target store's operational capabilities.

2. The method according to claim 1, characterized in that, The information is based on the store's basic information, the interaction information, and the features of surrounding points of interest. The target stores are predicted, and the prediction results are obtained, including: Based on the store location information, the reachable area of ​​the target store is determined, and the time required for the second user to reach the target store from any location within the reachable area is less than or equal to a preset time. Based on the store's basic information, the interaction information, the features of surrounding points of interest, and the store's reachable area, the target store is predicted to obtain the prediction result.

3. The method according to claim 2, characterized in that, The step of determining the reachable area of ​​the target store based on the store location information includes: Based on the store location information, an initial area range corresponding to the target store is determined with the target store as the center. The time required for the second user to reach the target store from any location within the initial area range is less than or equal to the preset time. Obtain the urban built-up area of ​​the city where the target store is located; The initial area range is superimposed with the urban built-up area range, and the overlapping area range is determined as the reachable area range of the store.

4. The method according to claim 2, characterized in that, The surrounding facilities are located within the target area, the target area is within the reachable area of ​​the store, and the target store is located within the target area.

5. The method according to claim 1, characterized in that, If the multimedia information includes comment text information and image information related to the target store, obtaining the interaction information corresponding to the target store includes: The text semantic tags are determined based on the comment text information, and the image semantic tags are determined based on the image information; If the semantics of the text semantic tag and the image semantic tag do not overlap, the text semantic tag and the image semantic tag are determined as the interaction information.

6. The method according to claim 5, characterized in that, The method further includes: If the semantics of the text semantic tag and the image semantic tag are contradictory, obtain the first number of tags corresponding to the text semantic tag and the first number of collected comment text information to which the text semantic tag belongs, and obtain the second number of tags corresponding to the image semantic tag and the second number of collected image information to which the image semantic tag belongs; A first confidence level corresponding to the text semantic label is determined based on the first number of labels and the first number of acquisitions, and a second confidence level corresponding to the image semantic label is determined based on the second number of labels and the second number of acquisitions. If the first confidence level is greater than the second confidence level, the text semantic tag is determined to be the interaction information; If the first confidence level is less than the second confidence level, the image semantic label is determined as the interaction information.

7. The method according to any one of claims 1-6, characterized in that, The step of determining the surrounding points of interest features corresponding to the target store based on the store location information includes: Based on the store location information, determine the various types of surrounding facilities corresponding to the target store; Determine the number of facilities corresponding to each type of surrounding facility, and determine the distance from each of the surrounding facilities to the target store; The features of the surrounding points of interest are determined based on the number of facilities corresponding to the various types of surrounding facilities and the distance of each surrounding facility from the target store.

8. The method according to claim 7, characterized in that, The step of determining the features of the surrounding points of interest based on the number of facilities corresponding to the various types of surrounding facilities and the distance of each surrounding facility to the target store includes: Based on the store type of the target store and the facility type of each type of surrounding facility, a weight is determined for each type of surrounding facility. The weight is used to indicate the degree of association between each type of surrounding facility and the target store, and the weight is positively correlated with the degree of association. The features of the surrounding points of interest are determined based on the number of facilities corresponding to each of the various types of surrounding facilities, the weight of each of the various types of surrounding facilities, and the distance of each of the surrounding facilities to the target store.

9. A data processing apparatus, characterized in that, The device includes an acquisition unit, a determination unit, and a prediction unit: The acquisition unit is used to acquire basic store information corresponding to the target store and interactive information corresponding to the target store. The basic store information includes the store location information and the business information of the target store. The interactive information is determined based on the multimedia information of the target store uploaded by the first user. The determining unit is used to determine the surrounding point of interest features corresponding to the target store based on the store location information, and the surrounding point of interest features are used to indicate the surrounding facilities corresponding to the target store. The prediction unit is used to predict the target store based on the store's basic information, the interaction information, and the features of the surrounding points of interest, and to obtain a prediction result, which is used to indicate the target store's operational capabilities.

10. A computer device, characterized in that, The computer device includes a processor and memory: The memory is used to store program code and transmit the program code to the processor; The processor is configured to execute the method described in any one of claims 1-8 according to the instructions in the program code.

Images