Data index processing method and device

Abstract

The invention discloses a data index processing method and device, and relates to the technical field of computers. A specific embodiment of the method comprises the steps of obtaining logistics feature data of a user in a specified time period, wherein the logistics feature data comprises numerical features and text features; performing feature extraction on the numerical features to obtain first features, and performing semantic analysis on the text features to obtain second features of the preference degree of the user to the logistics service; determining a scene category corresponding to the user based on the first feature and the second feature; and inputting the first feature and the second feature into an attribution analysis model corresponding to the scene category to obtain a satisfaction attribution index of the user for the logistics service. According to the embodiment, the logistics customer historical behavior data and the large model extraction features are combined, the attribution classification model is trained, attribution classification of different customers is completed, and finally guidance verbal skills in different scenes are generated.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method and device for processing data indicators. Background Art

[0002] There are many B-end logistics customers, and their order placing behaviors are complex and changeable. We conduct attribution analysis on customers’ real-time status and historical behaviors, and understand information such as customers’ usage and satisfaction with logistics products. This guides front-line business personnel to make accurate operational measures, promote healthy cooperation between enterprises and customers, provide better services to customers, and create greater profits for enterprises.

[0003] In the process of realizing the present invention, the inventors found that there are at least the following problems in the prior art: In the existing B-end logistics customer attribution analysis method, the traditional manual method of visiting customers requires high labor costs and low efficiency. In order to obtain more realistic customer needs, multiple communications and visits are usually required, resulting in a large amount of resource waste. The current logistics scenarios are complex, the data sources are multiple, and the data quality is low. Customer order placement behavior is affected by many factors, and it is difficult to complete customer attribution analysis from a single perspective. In addition, when performing attribution analysis using methods such as statistical analysis and machine learning, data from different sources need to be integrated and processed. At this time, a large amount of feature engineering is required, which is difficult to process and the final model effect is usually not as expected, making it difficult to gain insight into the true needs of customers. Summary of the invention

[0004] In view of this, an embodiment of the present invention provides a method and device for processing data indicators, which can perform feature extraction based on the user's logistics feature data to characterize customer needs, save human resources and improve efficiency; extract user preference characteristics for logistics services through semantic analysis and sentiment analysis, which can be applied to a variety of data sources, complete customer satisfaction attribution analysis from multiple angles, improve data integration processing efficiency, and enhance model prediction effects; combine logistics customer historical behavior data and large model to extract features, train attribution classification models, complete attribution classification of different customers, and finally generate guidance words for different scenarios.

[0005] To achieve the above object, according to one aspect of an embodiment of the present invention, a method for processing a data indicator is provided, comprising: Obtaining the user's logistics feature data within a specified period of time, wherein the logistics feature data includes numerical features and text features; Extracting the numerical feature to obtain a first feature, and performing semantic analysis on the text feature to obtain a second feature of the user's preference for logistics services; Determine a scene category corresponding to the user based on the first feature and the second feature; The first feature and the second feature are input into the attribution analysis model corresponding to the scenario category to obtain an attribution index of the user's satisfaction with the logistics service.

[0006] Optionally, after obtaining the user's satisfaction attribution index for the logistics service, the method further includes: determining corresponding user service guidance words according to the satisfaction attribution index.

[0007] Optionally, the numerical features include continuous value features and discrete value features; extracting the first feature from the numerical features includes: extracting the third feature by changing the feature value distribution or performing feature value scaling operations on the continuous value features; extracting the fourth feature by performing discrete value encoding on the discrete value features; and obtaining the first feature by performing feature selection on the third feature and the fourth feature.

[0008] Optionally, a semantic analysis is performed on the text-type features to obtain a second feature of the user's preference for logistics services, including: constructing prompt words based on a set set of demand points and user sensitivity information, the prompt words are used to describe the feature information to be extracted by the large model; using the large model to perform text extraction and semantic analysis on the text-type features to obtain the user's demand points and the user's satisfaction with logistics services; and feature processing of the user's demand points and the user's satisfaction with logistics services to obtain a second feature of the user's preference for logistics services.

[0009] Optionally, determining the scene category corresponding to the user based on the first feature and the second feature includes: obtaining a threshold for dividing the user into different scene categories based on basic portrait information of the user; and determining the scene category corresponding to the user based on the first feature and the second feature, as well as the threshold for dividing the user into different scene categories.

[0010] Optionally, the attribution analysis model corresponding to each scene category is trained in the following manner: for each scene category, the basic portrait information, the first feature and the second feature of the sample users under the scene category are obtained as training samples, and the training samples are labeled using the actual attribution results; based on the multi-classification model, the labeled training samples are used to perform model training to obtain the attribution analysis model corresponding to the scene category.

[0011] According to another aspect of an embodiment of the present invention, a device for processing data indicators is provided, including: A data acquisition module is used to acquire the user's logistics feature data within a specified period of time, wherein the logistics feature data includes numerical features and text features; A feature extraction module, configured to extract the numerical feature to obtain a first feature, and perform semantic analysis on the text feature to obtain a second feature of the user's preference for logistics services; A scene classification module, used to determine the scene category corresponding to the user based on the first feature and the second feature; An indicator processing module is used to input the first feature and the second feature into the attribution analysis model corresponding to the scenario category to obtain the attribution indicator of the user's satisfaction with the logistics service.

[0012] According to another aspect of an embodiment of the present invention, there is provided an electronic device, comprising: one or more processors; a storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method for processing data indicators provided by an embodiment of the present invention.

[0013] According to another aspect of an embodiment of the present invention, a computer-readable medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the method for processing data indicators provided by an embodiment of the present invention is implemented.

[0014] According to another aspect of an embodiment of the present invention, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the method for processing data indicators provided by an embodiment of the present invention is implemented.

[0015] An embodiment of the above invention has the following advantages or beneficial effects: by obtaining the logistics feature data of the user within a specified time period, the logistics feature data includes numerical features and text features; extracting the numerical features to obtain the first feature, and performing semantic analysis on the text features to obtain the second feature of the user's preference for logistics services; determining the scenario category corresponding to the user based on the first feature and the second feature; inputting the first feature and the second feature into the attribution analysis model corresponding to the scenario category to obtain the technical solution of the attribution index of the user's satisfaction with the logistics service, which can characterize customer needs based on the user's logistics feature data by feature extraction, saves human resources and improves efficiency; extracting the user's preference features for logistics services through semantic analysis and sentiment analysis can be applied to a variety of data sources, complete customer satisfaction attribution analysis from multiple angles, improve data integration processing efficiency, and enhance model prediction effect; combining logistics customer historical behavior data and large model to extract features, training attribution classification models, completing attribution classification of different customers, and finally generating guidance words in different scenarios.

[0016] The further effects of the above-mentioned non-conventional optional manner will be described below in conjunction with the specific implementation manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings are used to better understand the present invention and do not constitute an improper limitation of the present invention. Figure 1 is a schematic diagram of main steps of a method for processing data indicators according to an embodiment of the present invention; Figure 2 It is a schematic diagram of a large model feature extraction process according to an embodiment of the present invention; Figure 3 It is a schematic diagram of a processing flow of data indicators according to an embodiment of the present invention; Figure 4 is a schematic diagram of main modules of a device for processing data indicators according to an embodiment of the present invention; Figure 5 is an exemplary system architecture diagram to which embodiments of the present invention may be applied; Figure 6 It is a schematic diagram of the structure of a computer system of a terminal device or a server suitable for implementing an embodiment of the present invention. DETAILED DESCRIPTION

[0018] The following is a description of exemplary embodiments of the present invention in conjunction with the accompanying drawings, including various details of the embodiments of the present invention to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0019] It should be noted that the collection, collection, updating, analysis, processing, use, transmission, storage and other aspects of user personal information involved in the technical solution disclosed in the present invention are in compliance with the provisions of relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. Necessary measures are taken for user personal information to prevent illegal access to user personal information data and maintain the security of user personal information, network security and national security.

[0020] Figure 1 FIG. 1 is a schematic diagram of the main steps of the method for processing data indicators according to an embodiment of the present invention. Figure 1 As shown, the method for processing data indicators in the embodiment of the present invention mainly includes the following steps S101 to S104.

[0021] Step S101: Obtain the logistics feature data of the user within a specified time period, and the logistics feature data includes numerical features and text features. According to an embodiment of the present invention, in order to facilitate the processing of logistics feature data, the logistics feature data can be divided into numerical features and text features according to the data type of the logistics feature data, and feature processing is performed on the numerical features and text features respectively to better extract information that can characterize user characteristics. Among them, numerical features refer to structured numerical type features, such as: the number of contracted products, the number of orders in a historical period, and so on. Text features refer to unstructured expression features, such as: product description information, customer feedback information, etc. Among them, text features can also include, for example, sound data or picture data.

[0022] In an embodiment of the present invention, taking the logistics scenario as an example, the data processed by the customer feature construction module mainly includes customer basic information, contract information, historical order quantity characteristics, operation link characteristics, customer complaint claim information, visit feedback information and other data. Among them, the basic information of the customer includes, for example, the customer name, industry, enterprise type, office address and other information. The contract information includes, for example, the customer signing time, the number of contracted products, the contract price and other information. The historical order quantity characteristics record the detailed records of the customer's daily waybill. The operation characteristics include the timeliness rate, loss rate, and success rate of the collection and dispatch of each link in the waybill. The customer complaint claim information includes the problems encountered by customers in the process of using logistics products through system filling, offline feedback, email reporting, etc., as well as the claim form submitted for abnormal orders. The visit feedback information refers to the customer feedback information recorded by the front-line business personnel after communicating with the customer through telephone visits, door-to-door visits, etc., which may include the problems encountered by the customer, the required demands, the solution after communication and discussion, the next step and other text information. The above logistics feature data include numerical features and text features.

[0023] Step S102: extract the numerical feature to obtain a first feature, and perform semantic analysis on the text feature to obtain a second feature of the user's preference for logistics services. In the embodiment of the present invention, different methods are used to extract features from numerical features and text features, respectively, so that the features of different types of features are combined for feature extraction, and features that represent user information can be better extracted.

[0024] According to one embodiment of the present invention, the numerical features include continuous value features and discrete value features. Extracting the first feature from the numerical features includes: extracting the third feature by changing the distribution of feature values ​​or performing feature value scaling operations on the continuous value features; extracting the fourth feature by performing discrete value encoding on the discrete value features; and obtaining the first feature by performing feature selection on the third feature and the fourth feature. In the embodiment of the present invention, different methods are used to extract features from the continuous value features and the discrete value features, respectively, so that the feature representations thereof can be better extracted.

[0025] In an embodiment of the present invention, when feature processing is performed on a numerical feature, the following steps are mainly included: 1. Data cleaning: Since the data comes from multiple channels, it is necessary to recheck and screen the data and delete duplicate content and erroneous information; 2. Data denoising: Filter data containing abnormal continuous values ​​by using 3sigma denoising, binning, clustering and other methods; 3. Missing value filling: For numerical features, you can choose the maximum value, minimum value, average value or 0 value to fill; for discrete value features, you can choose the mode, empty characters, etc. to fill; 4. Feature extraction: construct a new feature subset by performing principal component analysis, linear discriminant analysis and other projection change methods on the existing feature set; 5. Data conversion: Continuous value features are transformed by logarithmization, min-max normalization, standardization, regularization and other methods to change the original data feature value distribution or scale the data; discrete value features are transformed by unique hot encoding, dictionary encoding, virtual encoding and other methods; 6. Feature selection: Feature selection is performed through filter methods, wrapper methods, and embedded methods to select features that can better represent user information, so as to increase the efficiency of model training prediction and reduce the risk of overfitting.

[0026] According to one of the embodiments of the present invention, a semantic analysis is performed on the text-type features to obtain a second feature of the user's preference for logistics services, which may specifically include: constructing prompt words based on a set set of demand points and user sensitivity information, the prompt words are used to describe the feature information to be extracted by the large model; using the large model to perform text extraction and semantic analysis on the text-type features to obtain the user's demand points and the user's satisfaction with logistics services; and performing feature processing on the user's demand points and the user's satisfaction with logistics services to obtain a second feature of the user's preference for logistics services.

[0027] In the embodiment of the present invention, a large model is used for text extraction and semantic analysis (sentiment analysis). Specifically, the large model is used to extract the customer's demand points from text data such as the customer's historical customer complaint claim form, the front-line business personnel's visit feedback form, and the customer's voice. At the same time, the text semantic analysis (sentiment analysis) is performed through the large model to obtain the customer's satisfaction with the current logistics service and the degree of attention to different points of concern. Among them, the customer's demand points need to be determined according to the business scenario, and a set of demand points can be pre-defined to facilitate matching the corresponding demand points from the text data.

[0028] Figure 2 The following is a schematic diagram of a large model feature extraction process according to an embodiment of the present invention. When a large model is used for feature extraction, the following steps are mainly included.

[0029] 1. Under the customer dimension, summarize and organize text data that can reflect customer problems, such as customer complaint claims, visit feedback information, and customer voice. The processed text data must include fields such as the customer's unique identification ID, time, type, problem description, and problem impact.

[0030] 2. Construct prompts based on the current business context, describe the feature information that needs to be extracted by the big model, and determine the input and output formats. In the attribution analysis of logistics customers, the big model is mainly used for text extraction and semantic analysis (sentiment analysis). The following examples show the prompts designed when using the big model to perform the above tasks under two different types of data.

[0031] Among them, when using the big model to extract text from the feedback information of front-line business personnel, the prompt words designed are as follows:

[0032] When using a large model to perform semantic analysis (sentiment analysis) on customer voices, the prompt words designed are as follows:

[0033] 3. Fill the text information in each data sample of the customer dimension into the above-mentioned prompt word template, then call the large model interface, pass in the prompt word containing the customer text information, and obtain the output result of the large model.

[0034] 4. Further feature processing is performed on the output results of the large model. Statistics are collected on the number of customers' attention to different demand points and the degree of sensitivity to the issue during the cycle according to the dimensions of day, week, month, etc. The second feature of the customer's preference for logistics services extracted by the large model is constructed through the mean, extreme value, sum, etc.

[0035] Step S103: Determine the scene category corresponding to the user based on the first feature and the second feature. Different users correspond to different features and different scene categories. The scene category describes the real-time status of the user.

[0036] According to one of the embodiments of the present invention, determining the scene category corresponding to the user based on the first feature and the second feature may specifically include: obtaining a threshold for dividing the user into different scene categories based on basic portrait information of the user; determining the scene category corresponding to the user based on the first feature and the second feature, as well as the threshold for dividing the user into different scene categories.

[0037] In an embodiment of the present invention, the logistics scene classification module can identify the user's real-time status based on the user's basic portrait information, order quantity change trend, and the user's preference characteristics for logistics services extracted by the big model, and classify the user's scene in combination with business needs. Among them, the scene categories corresponding to the user include the following three categories: Reduction scenario: The customer's order volume shows an abnormal weekly or monthly decline; Potential tapping scenarios: The customer's order volume is stable, there is growth potential, or the customer is about to enter the peak shipping season; Churn scenario: The customer has been active in placing orders in the past, but has not placed any orders in recent period of time.

[0038] When constructing the user's logistics characteristics, we obtained the user's daily order situation. Now we can determine whether the user's order volume has declined, or has not issued any orders, or is relatively stable. Among them, we can set the corresponding reduction scenario when the user's order volume declines according to the specific business scenario, the order volume fluctuation does not exceed a certain amount to be stable, and so on.

[0039] Step S104: inputting the first feature and the second feature into the attribution analysis model corresponding to the scenario category to obtain an attribution index of the user's satisfaction with the logistics service.

[0040] According to one of the embodiments of the present invention, the attribution analysis model corresponding to each scene category is obtained by training in the following manner: for each scene category, the basic portrait information, the first feature and the second feature of the sample user under the scene category are obtained as training samples, and the training samples are labeled using the actual attribution results; based on the multi-classification model, the labeled training samples are used to perform model training to obtain the attribution analysis model corresponding to the scene category.

[0041] Specifically, in an embodiment of the present invention, after the large model extracts features and determines the scenario category in which the user is located, combined with business needs, attribution analysis can be performed on the customer to find the starting point that can improve the customer experience and increase the number of orders issued by the customer. Attribution analysis can be performed on the customer by training an attribution analysis model for each scenario category. First, by integrating the basic portrait information of the sample user, the trend of the number of orders, and the characteristics of the user's preference for logistics services extracted by the large model, the actual attribution results of historical front-line personnel feedback are used as category labels to construct training samples for the attribution analysis model. Then, a multi-classification model for attribution analysis is constructed, and specifically, GBDT (Gradient Boosting Decision Tree), XGBoost ((eXtreme Gradient Boosting, distributed gradient boosting library), lightGBM (Light Gradient Boosting Machine, gradient boosting decision tree model) and other models can be used to train the attribution data training samples of historical users. After that, the model effect is tuned. After the model effect is tuned, attribution analysis can be performed based on the user's real-time logistics characteristics to obtain the user's satisfaction attribution index in the current scenario.

[0042] According to an embodiment of the present invention, after obtaining the user's satisfaction attribution index for logistics services, the method further includes: determining corresponding user service guidance words according to the satisfaction attribution index. When generating guidance words based on the user's scenario category and attribution analysis results, it can be achieved by designing a fixed template. The following examples of words are given for several scenarios.

[0043] 1. Reduction scenario: The customer's recent order volume has declined because they are not satisfied with the price. The customer may switch to other couriers. You can adjust the discount appropriately to get more discounts for the customer. 2. Potential tapping scenario: The customer is a newly signed customer who has a large demand for placing orders recently. He is not familiar with the system operation process and needs guidance as soon as possible; 3. Churn scenario: The customer has recently experienced several incidents of lost shipping orders, resulting in the customer not placing any orders. It is necessary to follow up and complete the claims settlement for the customer as soon as possible.

[0044] Finally, the user service guidance scripts and key information extracted from the big model are sent to the business side to provide business guidance to front-line business personnel, solve customer problems more accurately and efficiently, and achieve a sustainable and healthy cooperative relationship with B-end logistics customers.

[0045] Figure 3 FIG. 1 is a schematic diagram of a processing flow of data indicators according to an embodiment of the present invention. Figure 3As shown, in an embodiment of the present invention, when processing data indicators, first, the logistics feature data of the user within a specified time period is obtained, such as data including basic customer information, signed contract information, historical order quantity characteristics, operation link characteristics, customer complaint and claim information, visit feedback information, etc.; then, the logistics feature data of the user is processed according to numerical features and text features respectively, and the numerical features are extracted to obtain the first feature, and the text features are extracted and semantically analyzed (sentiment analysis) based on a large model to obtain the second feature; then, the scenario category corresponding to the user is determined based on the first feature and the second feature, and the scenario category includes, for example: reduction scenario, potential tapping scenario and loss scenario; finally, through the attribution analysis model of the scenario category corresponding to the user, attribution analysis is performed and user service guidance words are generated.

[0046] Figure 4 Schematic diagram of the main modules of the device for processing data indicators according to an embodiment of the present invention. Figure 4 As shown, the device 400 for processing data indicators according to the embodiment of the present invention mainly includes: a data acquisition module 401 , a feature extraction module 402 , a scene classification module 403 and an indicator processing module 404 .

[0047] The data acquisition module 401 is used to acquire the logistics characteristic data of the user within a specified period of time, wherein the logistics characteristic data includes numerical characteristics and text characteristics; A feature extraction module 402 is used to extract the numerical feature to obtain a first feature, and to perform semantic analysis on the text feature to obtain a second feature of the user's preference for logistics services; A scene classification module 403, configured to determine a scene category corresponding to the user based on the first feature and the second feature; The index processing module 404 is used to input the first feature and the second feature into the attribution analysis model corresponding to the scenario category to obtain the attribution index of the user's satisfaction with the logistics service.

[0048] According to an embodiment of the present invention, the indicator processing module 404 may also be used to: after obtaining the user's satisfaction attribution indicator for the logistics service, determine the corresponding user service guidance words according to the satisfaction attribution indicator.

[0049] According to another embodiment of the present invention, the numerical features include continuous value features and discrete value features; the feature extraction module 402 can also be used to: obtain a third feature by performing feature value distribution change or feature value scaling operation on the continuous value feature to perform feature extraction; obtain a fourth feature by performing discrete value encoding on the discrete value feature to perform feature extraction; obtain the first feature by performing feature selection on the third feature and the fourth feature.

[0050] According to another embodiment of the present invention, the feature extraction module 402 can also be used to: construct prompt words based on a set of demand points and user sensitivity information, wherein the prompt words are used to describe the feature information to be extracted by the large model; use the large model to perform text extraction and semantic analysis on the text-type features to obtain the user's demand points and the user's satisfaction with the logistics service; perform feature processing on the user's demand points and the user's satisfaction with the logistics service to obtain a second feature of the user's preference for the logistics service.

[0051] According to another embodiment of the present invention, the scene classification module 403 can also be used to: obtain the user's classification threshold for different scene categories based on the user's basic portrait information; determine the scene category corresponding to the user based on the first feature and the second feature, and the user's classification threshold for different scene categories.

[0052] According to another embodiment of the present invention, the attribution analysis model corresponding to each scene category is obtained by training in the following manner: for each scene category, the basic portrait information, the first feature and the second feature of the sample user under the scene category are obtained as training samples, and the training samples are labeled using the actual attribution results; based on the multi-classification model, the labeled training samples are used to perform model training to obtain the attribution analysis model corresponding to the scene category.

[0053] Figure 5 An exemplary system architecture 500 is shown to which the method for processing a data indicator or the apparatus for processing a data indicator according to an embodiment of the present invention can be applied.

[0054] like Figure 5 As shown, system architecture 500 may include terminal devices 501, 502, 503, a network 504 and a server 505. Network 504 is used to provide a medium for communication links between terminal devices 501, 502, 503 and server 505. Network 504 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0055] Users can use terminal devices 501, 502, 503 to interact with server 505 through network 504 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 501, 502, 503, such as shopping applications, data processing applications, indicator analysis applications, instant messaging tools, email clients, social platform software, etc. (only as examples).

[0056] The terminal devices 501 , 502 , and 503 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, and the like.

[0057] Server 505 may be a server that provides various services, such as a background management server that provides support for websites browsed by users using terminal devices 501, 502, and 503 (for example only). The background management server may obtain the user's logistics feature data within a specified period of time for the received data indicator processing request and other data, and the logistics feature data includes numerical features and text features; extract the numerical features to obtain a first feature, and perform semantic analysis on the text features to obtain a second feature of the user's preference for logistics services; determine the scenario category corresponding to the user based on the first feature and the second feature; input the first feature and the second feature into the attribution analysis model corresponding to the scenario category to obtain the user's satisfaction attribution index for logistics services, and feed back the processing results (such as the user's satisfaction attribution index for logistics services - for example only) to the terminal device.

[0058] It should be noted that the method for processing data indicators provided in the embodiment of the present invention is generally executed by the server 505 , and accordingly, the device for processing data indicators is generally set in the server 505 .

[0059] It should be understood that Figure 5 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0060] Reference below Figure 6 , which shows a schematic diagram of the structure of a computer system 600 of a terminal device or a server suitable for implementing an embodiment of the present invention. Figure 6 The terminal device or server shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0061] like Figure 6 As shown, the computer system 600 includes a central processing unit (CPU) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage part 608 into a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the system 600 are also stored. The CPU 601, the ROM 602, and the RAM 603 are connected to each other through a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0062] The following components are connected to the I / O interface 605: an input section 606 including a keyboard, a mouse, etc.; an output section 607 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 608 including a hard disk, etc.; and a communication section 609 including a network interface card such as a LAN card, a modem, etc. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to the I / O interface 605 as needed. A removable medium 611, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 610 as needed, so that a computer program read therefrom is installed into the storage section 608 as needed.

[0063] In particular, according to the embodiments disclosed in the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present invention include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 609, and / or installed from the removable medium 611. When the computer program is executed by the central processing unit (CPU) 601, the above-mentioned functions defined in the system of the present invention are executed.

[0064] It should be noted that the computer-readable medium shown in the present invention may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present invention, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0065] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0066] The units or modules involved in the embodiments of the present invention may be implemented by software or hardware. The units or modules described may also be provided in a processor. For example, they may be described as: a processor including a data acquisition module, a feature extraction module, a scene classification module, and an index processing module. Among them, the names of these units or modules do not constitute limitations on the units or modules themselves in certain circumstances. For example, the index processing module may also be described as "a module for inputting the first feature and the second feature into the attribution analysis model corresponding to the scene category to obtain the user's satisfaction attribution index for logistics services."

[0067] As another aspect, the present invention also provides a computer-readable medium, which may be included in the device described in the above embodiment; or it may exist independently without being assembled into the device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by a device, the device includes: obtaining the logistics feature data of the user within a specified period of time, the logistics feature data including numerical features and text features; extracting the first feature from the numerical feature, and performing semantic analysis on the text feature to obtain the second feature of the user's preference for logistics services; determining the scene category corresponding to the user based on the first feature and the second feature; inputting the first feature and the second feature into the attribution analysis model corresponding to the scene category to obtain the user's satisfaction attribution index for logistics services.

[0068] According to the technical solution of the embodiment of the present invention, by obtaining the logistics feature data of the user within a specified time period, the logistics feature data includes numerical features and text features; extracting the numerical features to obtain the first feature, and performing semantic analysis on the text features to obtain the second feature of the user's preference for logistics services; determining the scenario category corresponding to the user based on the first feature and the second feature; inputting the first feature and the second feature into the attribution analysis model corresponding to the scenario category to obtain the technical solution of the attribution index of the user's satisfaction with the logistics service, feature extraction can be performed based on the user's logistics feature data to characterize customer needs, saving human resources and improving efficiency; extracting the user's preference features for logistics services through semantic analysis and sentiment analysis can be applied to a variety of data sources, completing customer satisfaction attribution analysis from multiple angles, improving data integration processing efficiency, and enhancing model prediction effect; combining logistics customer historical behavior data and large model to extract features, training attribution classification models, completing attribution classification of different customers, and finally generating guidance words in different scenarios.

[0069] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may occur depending on design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for processing data indicators, characterized in that: include: Obtaining the user's logistics feature data within a specified period of time, wherein the logistics feature data includes numerical features and text features; Extracting the numerical feature to obtain a first feature, and performing semantic analysis on the text feature to obtain a second feature of the user's preference for logistics services; Determine a scene category corresponding to the user based on the first feature and the second feature; The first feature and the second feature are input into the attribution analysis model corresponding to the scenario category to obtain an attribution index of the user's satisfaction with the logistics service.

2. The method according to claim 1, characterized in that After obtaining the user's satisfaction attribution index for the logistics service, it also includes: The corresponding user service guidance words are determined according to the satisfaction attribution index.

3. The method according to claim 1, characterized in that The numerical features include continuous value features and discrete value features; Extracting the numerical feature to obtain a first feature includes: The third feature is obtained by performing a feature value distribution change or a feature value scaling operation on the continuous value feature to perform feature extraction; The fourth feature is obtained by performing discrete value encoding on the discrete value feature to perform feature extraction; The first feature is obtained by performing feature selection on the third feature and the fourth feature.

4. The method according to claim 1, characterized in that: The second feature of the user's preference for logistics services is obtained by semantically analyzing the text feature, including: Constructing prompt words based on a set of demand points and user sensitivity information, wherein the prompt words are used to describe feature information to be extracted from the large model; Use a large model to perform text extraction and semantic analysis on the text-based features to obtain the user's demand points and the user's satisfaction with the logistics service; The user's demand points and the user's satisfaction with the logistics service are processed to obtain a second feature of the user's preference for the logistics service.

5. The method according to claim 1, characterized in that Determining the scenario category corresponding to the user based on the first feature and the second feature includes: Obtaining the user's classification thresholds for different scene categories based on the user's basic portrait information; The scene category corresponding to the user is determined based on the first feature and the second feature, and a threshold for the user to classify different scene categories.

6. The method according to claim 1 or 5, characterized in that: The attribution analysis model corresponding to each scene category is trained in the following way: For each scenario category, basic portrait information, first features, and second features of sample users in the scenario category are obtained as training samples, and the training samples are labeled using actual attribution results; Based on the multi-classification model, the labeled training samples are used to perform model training to obtain the attribution analysis model corresponding to the scene category.

7. A device for processing data indicators, characterized in that: include: A data acquisition module is used to acquire the user's logistics feature data within a specified period of time, wherein the logistics feature data includes numerical features and text features; A feature extraction module, configured to extract the numerical feature to obtain a first feature, and perform semantic analysis on the text feature to obtain a second feature of the user's preference for logistics services; A scene classification module, used to determine the scene category corresponding to the user based on the first feature and the second feature; An indicator processing module is used to input the first feature and the second feature into the attribution analysis model corresponding to the scenario category to obtain the attribution indicator of the user's satisfaction with the logistics service.

8. An electronic device, characterized in that: include: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 6.

9. A computer readable medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images