Behavior analysis method and device of express user, electronic equipment and storage medium

Abstract

The application discloses a kind of express user's behavior analysis method, device, electronic equipment and storage medium, one aspect, based on Hbase database and kudu database, can long-term persistent mass express data and express business portrait is stored, and can respond to the query request of user behavior analysis in real time, with higher reliability, while, also can be expanded quickly when business volume increases, with good expansibility;Another aspect, the application can be based on the business behavior data of user, to generate each user's express business portrait, while receiving behavior analysis request, the business behavior data of Hbase database can be read, and behavior statistical data is generated, whereby, in combination with the express business portrait of user, the analysis of user express behavior can be realized, the express demand of user is known, so as to be able to provide data support for personalized service of express industry.

Description

Technical Field

[0001] This invention belongs to the field of information processing technology, and specifically relates to a method, device, electronic device, and storage medium for analyzing the behavior of express delivery users. Background Technology

[0002] Smart parcel lockers, parcel apps, and various parcel service websites are emerging phenomena that have arisen with the continuous development of the express delivery industry. Smart parcel lockers primarily assist couriers in sending and receiving parcels, serving a wide range of users including couriers and recipients. As 24-hour unattended parcel storage lockers, they interact with the service provider via IoT cards, representing the application of IoT technology in the modern logistics industry. Smart parcel lockers are the most effective solution for the last mile of e-commerce logistics, improving courier delivery efficiency, reducing delivery costs, and protecting user privacy and security, gradually gaining widespread recognition and promotion. Meanwhile, online parcel tools such as parcel apps and parcel service websites are used for online parcel tracking, allowing users to understand the logistics progress. They also allow online ordering of parcel services (such as sending packages), and, combined with the terminals of various courier companies, enable online ordering and courier pickup services, improving user experience and gaining widespread use.

[0003] Smart parcel lockers, parcel apps, and various parcel service websites serve as parcel delivery tools. With a massive amount of users storing, retrieving, and querying parcels daily, these activities generate enormous amounts of data. Effectively storing, querying, and analyzing this data presents a significant challenge. While traditional databases offer advantages such as support for integrity constraints and transactions, they cannot provide long-term storage for massive amounts of data and fail to meet the demands for efficient storage. Furthermore, the use of these tools often focuses solely on data storage without analysis, failing to identify individual users' parcel delivery needs and thus hindering personalized services within the parcel delivery industry. Therefore, a method for storing parcel data sustainably and efficiently, while also enabling data analysis of user behavior, is urgently needed. Summary of the Invention

[0004] The purpose of this invention is to provide a method, device, electronic device, and storage medium for analyzing the behavior of express delivery users, in order to solve the problems that existing databases cannot store express delivery data for a long time and efficiently, and that data analysis is not performed in applications, thus failing to provide data support for personalized services in the express delivery industry.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] In a first aspect, the present invention provides a method for analyzing and identifying the behavior of express delivery users, including:

[0007] Acquire business behavior data generated by each user when performing business operations on the express delivery business terminal, and store the business behavior data of each user in the HBase database;

[0008] Obtain the express delivery order data corresponding to each user's business behavior data within a preset time period, and use it as behavioral analysis data for each user;

[0009] The behavioral analysis data of each user is processed to extract key information, so as to obtain the express delivery business data of each user based on the extracted key information;

[0010] Based on the express delivery business data of each user, generate express delivery business tags for each user.

[0011] Based on each user's express delivery business tags, construct an express delivery business profile for each user and store the express delivery business profile for each user in the kudu database;

[0012] Receive a behavior analysis request, and based on the behavior analysis request, match the express delivery business profile of the target user in the Kudu database and match the business behavior data of the target user in the HBase database, wherein the target user is the user whose behavior is to be analyzed in the behavior analysis request;

[0013] Based on the target user's business behavior data, generate the target user's behavior statistics.

[0014] The express delivery business profile and behavioral statistics of the target users are visualized to complete the analysis of the target users' express delivery behavior.

[0015] Based on the above-disclosed content, when a user operates the express delivery service terminal, the present invention can obtain the business behavior data generated by each operation in real time and store it in the HBase database. The HBase database uses HDFS (distributed file system) to store the data, which has strong scalability and throughput. Therefore, it can store massive amounts of data for a long time, and will not cause a decrease in query performance when faced with the growth of massive amounts of data. Thus, the requirements for efficient storage of express delivery data can be met.

[0016] Meanwhile, this invention obtains corresponding express delivery order data based on user business behavior data, using it as behavioral analysis data for each user and extracting key information. Then, based on the extracted key information, it obtains each user's express delivery business data and generates express delivery business tags to construct an express delivery business profile for each user. Finally, upon receiving a behavior analysis request from express delivery operations personnel, it queries the corresponding user's business behavior data from the HBase database, performs data statistics to obtain the user's behavior statistics, and queries the user's express delivery business profile from the Kudu database. After the queries are completed, the behavior statistics and express delivery business profile are displayed to the operations personnel, who can then understand the user's express delivery needs based on the displayed data, thereby achieving personalized service.

[0017] In addition, this invention uses the Kudu database to store the express delivery business profiles of each user obtained from the analysis. The Kudu database supports partitioned viewing, has high scalability, and supports update operations and SQL query operations. Therefore, it can meet the storage needs of massive express delivery business profiles and can be read quickly during analysis.

[0018] Through the above design, on the one hand, this invention, based on HBase and Kudu databases, can persistently store massive amounts of express delivery data and express delivery business profiles for a long time, and can respond to query requests for user behavior analysis in real time, exhibiting high reliability. Furthermore, it can quickly expand its capacity as business volume increases, demonstrating excellent scalability. On the other hand, this invention can generate express delivery business profiles for each user based on their business behavior data. Upon receiving a behavior analysis request, it can read business behavior data from the HBase database and generate behavior statistics. Therefore, combined with the user's express delivery business profile, it is possible to analyze user express delivery behavior, understand user express delivery needs, and thus provide data support for personalized services in the express delivery industry.

[0019] In one possible design, key information extraction processing is performed on the behavioral analysis data of each user to obtain the express delivery business data of each user based on the extracted key information, including:

[0020] For each piece of behavior analysis data for each user, word segmentation is performed on each piece of behavior analysis data to obtain a word segmentation set for each user's behavior analysis data;

[0021] Calculate the importance factor value of each word in each word segmentation set for each user. The importance factor value of any word includes the word frequency value, word length value, and part-of-speech value of that word.

[0022] Based on the importance factor value of each word in each word segmentation set of each user, the weight value of each word in each word segmentation set of each user is calculated;

[0023] Based on the weight values ​​of each word in each word segmentation set for each user, the first keyword set corresponding to each piece of behavioral analysis data for each user is obtained;

[0024] Based on the first set of keywords corresponding to each behavioral analysis data of each user, words are combined for each behavioral analysis data of each user to obtain the second set of keywords corresponding to each behavioral analysis data of each user;

[0025] Merge the first keyword set and the second keyword set corresponding to each behavioral analysis data of each user to obtain the actual keyword set of each user's behavioral analysis data;

[0026] Duplicate keywords are removed from each user's actual keyword set, and the remaining keywords in each actual keyword set are used to obtain each user's express delivery business data.

[0027] Based on the above-disclosed content, this invention discloses a specific process for extracting key information. For any behavioral analysis data, it first performs word segmentation to obtain a word segmentation set. Then, it calculates the word frequency, word length, and part-of-speech value of each word in the segmentation set. Using the previously calculated parameters, it obtains the weight value of each word. Based on these weight values, a first keyword set for any behavioral analysis data can be obtained. Simultaneously, to avoid incomplete word strings caused by word segmentation, this invention also includes a word combination step. Based on the first keyword set, the words in the segmentation set are combined to obtain a second keyword set. Finally, the first and second keyword sets are merged, and duplicate keywords are removed to obtain the final keywords for any behavioral analysis data. Using this method, keywords are extracted from each behavioral analysis data point for each user. Based on the final keywords, the express delivery business data for each user can be obtained.

[0028] In one possible design, based on the first set of keywords corresponding to each behavioral analysis data point of each user, words are combined to obtain the second set of keywords corresponding to each behavioral analysis data point of each user, including:

[0029] For any user’s behavior analysis data, determine whether the i-th word in the word segmentation set corresponding to the behavior analysis data belongs to the first keyword set, where i starts from 1;

[0030] If so, determine whether the (i-1)th word satisfies the first preset condition, wherein the first preset condition is that the (i-1)th word exists in the word segmentation set corresponding to any behavior analysis data and does not belong to the first keyword set;

[0031] If so, then the (i-1)th word and the ith word are merged sequentially to obtain the first combined word, and it is determined whether the first combined word conforms to the word combination rules;

[0032] If the first combination of words conforms to the word combination rules, then the first combination of words is used as the second candidate keyword, and the weight value of the second candidate keyword is the weight value of the i-th word;

[0033] Determine whether the (i+1)th word satisfies the second preset condition, wherein the second preset condition is that the (i+1)th word exists in the word segmentation set corresponding to any behavior analysis data and belongs to the first keyword set;

[0034] If the second preset condition is met, the i-th word and the (i+1)-th word are merged sequentially to obtain a second combined word, and it is determined whether the second combined word conforms to the word combination rules;

[0035] If the second combination word conforms to the word combination rules, then the second combination word is also used as the second candidate keyword, and the weight value of the second candidate keyword corresponding to the second combination word is the sum of the weight values ​​of the i-th word and the (i+1)-th word;

[0036] Increment i by 1 and re-determine whether the i-th word is in the first keyword set, so that when i is greater than n, a second candidate keyword set is obtained, where n is the total number of words in the word segmentation set corresponding to any behavior analysis data;

[0037] The second candidate keyword set is filtered to identify second candidate keywords that have a subset relationship, and the second candidate keywords that are subsets are extracted.

[0038] Using the second candidate keywords as a subset and the filtered second candidate keyword set, a second keyword set is formed for any behavior analysis data of any user.

[0039] Based on the above-disclosed content, this invention discloses a specific process for combining words in a word segmentation set according to a first keyword set. Specifically, for the i-th word in the word segmentation set, it is determined whether the i-th word belongs to the first keyword set. If so, it is then determined whether the (i-1)-th word exists in the word segmentation set and does not belong to the first keyword set. If so, they are merged to obtain a first combined word. Next, it is determined whether the first combined word satisfies the word combination rules. If it does, it is used as a second candidate keyword. Then, it is determined whether the (i+1)-th word satisfies a second preset condition. If it does, the i-th word and the (i+1)-th word are merged to obtain... Similarly, for the second combination word, it is also necessary to determine whether the second combination word meets the word combination rules. If it does, the second combination word is also considered as a second candidate keyword. At this point, the combination of a word and its adjacent words can be completed. Finally, i is incremented by 1, and word combination is performed in the aforementioned way. When i is greater than n, the combination of all words in the word segmentation set can be completed, and the second candidate keyword set is obtained. After obtaining the second candidate keyword set, word filtering is still required. That is, the second candidate keywords as subsets are extracted from the second candidate keyword set, and combined with the filtered second candidate keyword set to form the second keyword set of any behavioral analysis data.

[0040] In one possible design, if the i-th word does not belong to the first keyword set, the method further includes:

[0041] Determine whether the (i-1)th word and the (i+1)th word both belong to the first keyword set and both exist in the word segmentation set corresponding to any behavior analysis data;

[0042] If so, then the (i-1)th word, the ith word, and the (i+1)th word are merged sequentially to obtain the third combination word;

[0043] Determine whether the third combination of words conforms to the word combination rules;

[0044] If so, the third combination of words is also used as the second candidate keyword, wherein the weight value of the second candidate keyword corresponding to the third combination of words is the sum of the weight values ​​of the (i-1)th word and the (i+1)th word;

[0045] Increment i by 1 and re-determine whether the i-th word is in the first keyword set, so that when i is greater than n, a second candidate keyword set is obtained, where n is the total number of words in the word segmentation set corresponding to any behavior analysis data.

[0046] In one possible design, if the (i-1)th word exists in the word segmentation set corresponding to any of the behavior analysis data and belongs to the first keyword set, then the method further includes:

[0047] Determine whether the (i+1)th word satisfies the second preset condition;

[0048] If so, then the (i-1)th word, the ith word, and the (i+1)th word are merged sequentially to obtain the fourth combination word;

[0049] Determine whether the fourth combination of words conforms to the word combination rules;

[0050] If so, the fourth combination of words is also used as the second candidate keyword, wherein the weight value of the second candidate keyword corresponding to the fourth combination of words is the sum of the weight values ​​of the i-th word and the (i+1)-th word;

[0051] Increment i by 1 and re-determine whether the i-th word is in the first keyword set, so that when i is greater than n, a second candidate keyword set is obtained, where n is the total number of words in the word segmentation set corresponding to any behavior analysis data.

[0052] In one possible design, based on each user's express delivery business data, express delivery business tags for each user are generated, including:

[0053] Based on the express delivery business data of each user, the sending and receiving business data of each user are obtained. The sending business data includes the sending transportation method, the type of sending item, and the sending express company. The receiving business data includes the sender's name, the type of receiving item, the receiving item transportation method, and the transportation express company.

[0054] By analyzing the frequency of shipping methods, item types, and courier companies in each user's shipping data, we can obtain each user's preferred shipping methods, item types, and courier companies.

[0055] Frequency statistics are performed on the sender's name, type of received item, mode of transport, and courier company in each user's package receipt data to obtain each user's preferred sender's name, preferred type of received item, preferred mode of transport, and preferred courier company.

[0056] By utilizing each user's preferred shipping method, preferred item type, and preferred courier company, a shipping service tag is generated for each user. Similarly, by utilizing each user's preferred sender name, preferred item type, preferred shipping method, and preferred courier company, a receiving service tag is generated for each user. In this way, the shipping service tags and receiving service tags of each user are used to form the courier service tags of each user.

[0057] Accordingly, based on each user's express delivery service tags, a user's express delivery service profile is constructed, including:

[0058] Obtain each user's account information and generate account tags for each user based on that information;

[0059] Based on each user's account tags, shipping service tags, and receiving service tags, a courier service profile for each user is constructed.

[0060] Based on the above-disclosed content, this invention discloses a specific process for constructing a courier service profile for each user. First, the courier service data is divided to obtain each user's sending and receiving service data. Then, frequency statistics are performed on each piece of information in the aforementioned two types of data to obtain each user's preferred sending method, preferred sending item type, preferred sending courier company, preferred sender name, preferred receiving item type, preferred receiving item transportation method, and preferred courier company. The statistically obtained data serves as the tags for each user. Next, each user's account information is obtained, and account tags (including basic information such as user name, gender, and username) are generated based on the account information. Finally, using the aforementioned three tags, the courier service profile for each user can be constructed.

[0061] In one possible design, the express delivery service profile includes sender service tags and receiver service tags. After constructing the express delivery service profile for each user, the method further includes:

[0062] Based on the parcel delivery service tags in each user's express delivery service profile, parcel delivery service promotion information is generated for each user. The parcel delivery service promotion information includes at least one piece of service promotion data, and each piece of service promotion data includes: the parcel delivery express company, the transportation method, the transportation time, the type of item insurance, and the total price of the parcel delivery.

[0063] Based on the recipient service tags in the user's express delivery business profile, obtain the preferred product types of each user, and generate product promotion information for each user based on the preferred product types of each user;

[0064] The system sends parcel delivery service promotion information to each user's corresponding terminal and displays product promotion information to each user when they enter the main interface of the express delivery service terminal.

[0065] Based on the above-disclosed content, this invention can also utilize the parcel delivery service tags in the user's express delivery service profile to generate promotional information for the user's parcel delivery service. That is, based on the user's preferences in the parcel delivery service tags, different promotional data can be combined and sent to the user. At the same time, based on the parcel receiving service tags in the express delivery service profile, the user's preferred product types can be obtained, and product promotional information can be generated based on the preferred product types so that it can be displayed when the user enters the main interface of the express delivery service, thereby realizing the recommendation of the user's preferred products. Through the above design, personalized services for users can be realized, thereby improving the service capabilities of express delivery.

[0066] Secondly, the present invention provides a device for analyzing the behavior of express delivery users, comprising:

[0067] The acquisition unit is used to acquire business behavior data generated by each user when performing business operations on the express delivery business terminal, and store the business behavior data of each user in the HBase database.

[0068] The data extraction unit is used to obtain the express delivery order data corresponding to each user's business behavior data within a preset time period, so as to serve as behavioral analysis data for each user.

[0069] The key information extraction unit is used to extract key information from the behavioral analysis data of each user, so as to obtain the express delivery business data of each user based on the extracted key information.

[0070] The tag generation unit is used to generate express delivery service tags for each user based on the express delivery service data of each user.

[0071] The profile building unit is used to build a profile of each user's express delivery business based on each user's express delivery business tags, and store each user's express delivery business profile in the kudu database;

[0072] An analysis unit is configured to receive a behavior analysis request, and based on the behavior analysis request, match the express delivery business profile of the target user in the Kudu database, and match the business behavior data of the target user in the HBase database, wherein the target user is the user whose behavior is to be analyzed in the behavior analysis request;

[0073] The analysis unit is also used to generate behavioral statistics data of the target user based on the target user's business behavior data;

[0074] The display unit is used to visualize the express delivery business profile and behavioral statistics of the target user in order to complete the express delivery behavior analysis of the target user.

[0075] Thirdly, the present invention provides another device for analyzing the behavior of express delivery users. Taking the device as an electronic device as an example, it includes a memory, a processor, and a transceiver that are connected in sequence. The memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the express delivery user behavior analysis method as described in the first aspect or any possible design in the first aspect.

[0076] Fourthly, the present invention provides a storage medium storing instructions that, when executed on a computer, perform the express delivery user behavior analysis method as described in the first aspect or any possible design of the first aspect.

[0077] Fifthly, the present invention provides a computer program product containing instructions that, when executed on a computer, cause the computer to perform the express delivery user behavior analysis method as described in the first aspect or any possible design of the first aspect. Attached Figure Description

[0078] Figure 1 A schematic diagram of the architecture of the express delivery user behavior analysis system provided by the present invention.

[0079] Figure 2 A flowchart illustrating the steps of the express delivery user behavior analysis method provided by the present invention;

[0080] Figure 3 This invention provides a schematic diagram illustrating the storage of business behavior data;

[0081] Figure 4 An interactive diagram illustrating the behavior analysis request query provided by this invention;

[0082] Figure 5 A schematic diagram of the structure of the express delivery user behavior analysis device provided by the present invention;

[0083] Figure 6 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0084] The present invention will be further described below with reference to the accompanying drawings and specific embodiments. It should be noted that while the description of these embodiments is intended to aid in understanding the invention, it does not constitute a limitation thereof. The specific structural and functional details disclosed herein are merely for describing exemplary embodiments of the invention. However, the invention can be embodied in many alternative forms and should not be construed as being limited to the embodiments described herein.

[0085] It should be understood that although the terms first, second, etc., may be used herein to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit, without departing from the scope of the exemplary embodiments of the invention.

[0086] It should be understood that the term "and / or" that may appear in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, B exists alone, and A and B exist simultaneously. The term " / and" that may appear in this document describes another relationship between related objects, indicating that two relationships can exist. For example, A / and B can mean: A exists alone, and A and B exist alone. In addition, the character " / " that may appear in this document generally indicates that the related objects before and after it are in an "or" relationship.

[0087] Example

[0088] See Figure 1 As shown, this application provides a system architecture. For example, a courier user behavior analysis system may include, but is not limited to, a data receiving layer, a data computing layer, a data storage layer, and a data output layer integrated within a server or behavior analysis terminal. The data receiving layer receives business behavior data transmitted from various front-end programs (i.e., courier service webpages) and various back-end programs (i.e., courier apps, smart parcel lockers). The data computing layer obtains corresponding courier order data based on the business behavior data, extracts key information, and then generates courier business tags based on the extracted key information to construct courier business profiles for each user. The data storage layer stores the business behavior data of each user in an H... The system consists of a base database and a kudu database storing the express delivery business profiles of each user. The data output layer acts as the interaction interface, interacting with the client of the operations and maintenance personnel. Upon receiving a behavior analysis request from the client, it extracts profile and business behavior data from the data storage layer, generates behavior statistics based on the extracted data, and then displays the profiles and behavior statistics, while also returning them to the client to complete the user behavior analysis. Therefore, this system can not only persistently store massive amounts of express delivery data and business profiles, but also analyze user express delivery behavior to understand user needs, thus providing data support for personalized services in the express delivery industry.

[0089] See Figure 2As shown, the behavior analysis method for express delivery users provided in the first aspect of this embodiment can be run on the behavior analysis terminal or server side, and the behavior analysis terminal can be, but is not limited to, a personal computer (PC), tablet computer, smartphone and / or personal digital assistant (PDA), etc. It is understood that the aforementioned execution subject does not constitute a limitation on the embodiment of this application. Accordingly, the operation steps of this method can be, but are not limited to, the steps S1 to S8 below.

[0090] S1. Obtain business behavior data generated by each user when performing business operations on the express delivery service terminal, and store the business behavior data of each user in the HBase database; in specific applications, the express delivery service terminal may include, but is not limited to: smart express lockers, express delivery apps, or express delivery service web pages. For example, when a user uses a smart express locker, express delivery app, and / or express delivery service web page, the above-mentioned devices send the corresponding business data for each operation to the server, and the server can use the received business data as the business behavior data of each user; furthermore, the business behavior data may include, but is limited to: sending data, receiving data, and logistics query data, etc.

[0091] In this embodiment, the HBase database uses HDFS to store data. Therefore, the database has strong scalability and throughput. Based on this advantage, this embodiment stores the aforementioned business behavior data. Thus, the database layout can store massive amounts of business behavior data for a long time, and will not cause a decrease in query performance when faced with the growth of massive amounts of business behavior data. Through the above design, the high-efficiency storage requirements of express delivery data can be met.

[0092] See Figure 3 As shown, each user's business behavior data is recorded as a row in a data table in the HBase database, and the data table is stored in an HDFS distributed file in the HBase database. In this way, the long-term storage of massive business behavior data can be achieved while ensuring data query efficiency.

[0093] In this embodiment, for example, the username or real name of the user logging into the express delivery business terminal is associated with the corresponding business behavior data so that the data can be searched later. Specifically, the username or real name can be written into the business behavior data for storage.

[0094] After acquiring and storing the business behavior data generated by each user's business operations on the express delivery business terminal, the data can be analyzed to build express delivery business profiles for each user. In this embodiment, the corresponding express delivery order data is first obtained based on each business behavior data. Then, key information in the express delivery order data is extracted to generate express delivery business tags. Finally, the express delivery business profile can be built based on the express delivery business tags. Specifically, the express delivery business profile building process is shown in steps S2 to S5 below.

[0095] S2. Obtain the express delivery order data corresponding to each user's business behavior data within a preset time period, as the behavioral analysis data for each user. In specific applications, business behavior data is extracted from the HBase database, and the corresponding express delivery order data is obtained based on the business behavior data corresponding to the user's operation. For example, if the business behavior data is parcel order data, then the corresponding parcel order can be obtained based on the parcel order data, which can be used as the express delivery order data. Similarly, if the business behavior data is parcel scanning data, then the corresponding express delivery order can be crawled based on the input or scanned parcel code. If the business behavior data is express delivery query data, then the corresponding express delivery order data can be obtained based on the input express delivery tracking number. Of course, when the business behavior data corresponds to other operations, the acquisition principle is the same as the above example, and will not be repeated here.

[0096] In this embodiment, example express order data may include, but is not limited to: sender's name, recipient's name, mode of transport, type of goods transported, name of the express company, tracking number, shipping time, precautions, remarks, and type of cargo insurance, etc.; and the preset time period may be, but is not limited to, 1 month, 3 months, or 6 months, etc. Of course, it can be specifically set according to actual use.

[0097] After obtaining the express delivery order data for each user within a preset time period, the express delivery order data can be used as behavioral analysis data, and key information can be extracted to generate express delivery business tags based on the extracted key information. In this embodiment, the key information extraction step is as shown in step S3 below.

[0098] S3. Perform key information extraction processing on the behavioral analysis data of each user to obtain the express delivery business data of each user based on the extracted key information. In specific applications, firstly, each behavioral analysis data of each user is segmented into words to obtain a segmented set of each behavioral analysis data. Then, keywords are extracted from the segmented set to obtain a first keyword set corresponding to each behavioral analysis data. Next, based on the first keyword set, the words in the segmented set are combined to obtain a second keyword set. Finally, the two are merged and duplicate keywords are deleted to obtain the actual keyword set of the corresponding behavioral analysis data. After obtaining the actual keyword set, it can be used as the express delivery business data. Optionally, the keyword extraction process may include, but is not limited to, the following steps S31 to S37.

[0099] S31. For each piece of behavior analysis data of each user, perform word segmentation on each piece of behavior analysis data to obtain a word segmentation set for each piece of behavior analysis data of each user; in specific implementation, examples may include, but are not limited to, using a dictionary-based word segmentation algorithm to perform word segmentation on each piece of behavior analysis data or a statistical machine learning algorithm to perform word segmentation on each piece of behavior analysis data, such as using the SVM (support vector machine) machine learning algorithm for word segmentation, or using the Hanlp word segmentation tool for word segmentation.

[0100] In this embodiment, the words in the word segmentation set are sorted according to the word segmentation order of the behavior analysis data.

[0101] After obtaining the word segmentation set of each user's behavior analysis data, keywords can be extracted. In this embodiment, the importance factor value of each word in the word segmentation set is calculated first. Then, based on the importance factor value of each word, the weight of each word is calculated. Finally, keywords can be extracted based on the weight. The extraction process is shown in steps S32, S33 and S34 below.

[0102] S32. Calculate the importance factor value of each word in each word segmentation set for each user, wherein the importance factor value of any word includes the word frequency value, word length value, and part-of-speech value of that word; in specific applications, the calculation formulas for each parameter in the importance factor value are disclosed below:

[0103] First, the word frequency value of any word can be calculated using the following formula (1):

[0104] (1)

[0105] In equation (1), This represents the first word in the word segmentation set corresponding to any behavior analysis data. The word frequency value of each word. Indicates the first The number of times a word appears in its corresponding word segmentation set.

[0106] Secondly, the word length value for any given word can be calculated using the following formula (2):

[0107] (2)

[0108] In the above formula (2), This represents the first word in the word segmentation set corresponding to any behavior analysis data. The word length value of each word. Indicates the first The length of each word For the first Each word corresponds to the maximum word length of all words in the word segmentation set; in this embodiment, the example word length is counted by the number of characters.

[0109] Finally, in this embodiment, any word has a different part-of-speech value depending on its part of speech. For example, the word... The first word is either a name or a verb, and its corresponding part-of-speech value is 0.8. The first word is an adjective or adverb, with a corresponding part-of-speech value of 0.6, while if the second word is an adjective or adverb... If a word is classified as another part of speech (i.e., other parts of speech besides nouns, verbs, adjectives, and adverbs), the corresponding part of speech value is 0. Furthermore, for any word in the word segmentation set, the part of speech can be obtained by, but is not limited to, rule-based part of speech tagging methods, statistical model-based part of speech tagging methods (such as Hidden Markov Models, Conditional Random Field Models, etc.), or deep learning-based part of speech tagging methods (such as using LSTM+CRF deep learning models for part of speech recognition).

[0110] After obtaining the importance factor values ​​of each word in each word segmentation set for each user, the weight values ​​of each word can be calculated based on the importance factor values, as shown in step S33 below.

[0111] S33. Based on the importance factor values ​​of each word in each word segmentation set of each user, calculate the weight value of each word in each word segmentation set of each user; in specific applications, different proportional coefficients can be assigned to the aforementioned word frequency value, word length value, and part-of-speech value, so that each importance factor value is multiplied by the corresponding proportional coefficient and the sum is obtained to obtain the weight value of each word; optionally, for example, the proportional coefficient of word frequency value can be, but is not limited to, 1.5, the proportional coefficient of word length value can be, but is not limited to, 0.8, and the proportional coefficient of part-of-speech value can be, but is not limited to, 1.1.

[0112] After obtaining the weight values ​​of each word in the word segmentation set, keywords can be extracted based on the weight values, as shown in step S34 below.

[0113] S34. Based on the weight values ​​of each word in each segmentation set for each user, obtain the first keyword set corresponding to each piece of behavioral analysis data for each user. In specific applications, for any segmentation set, words with parts of speech such as nouns, verbs, or adjectives can be filtered out first. Then, the remaining words are sorted in descending order of weight value. Finally, based on the total number of words n in any segmentation set, the number of first keywords is calculated. That is, if n*5%<20, then the extracted keywords are n*5%, and if n*5%>=20, then the extracted keywords are 20. Thus, based on the calculated number of keywords, keywords can be extracted according to the sorting order of weight values, and the extracted keywords can form the first keyword set. For example, if the calculated number of keywords is 5, then the top 5 words in the sorting order are extracted to form the first keyword set.

[0114] Because incomplete word strings are easily generated during word segmentation, such as the shipper's name being segmented into "shipper" and "name", this embodiment also includes a secondary keyword extraction step to improve the accuracy of the extracted keywords. This involves combining the words in the word segmentation set based on the keywords extracted in the first step, so as to obtain a second keyword set after the word combination is completed. The word combination process is shown in step S35 below.

[0115] S35. Based on the first keyword set corresponding to each behavioral analysis data of each user, combine words for each behavioral analysis data of each user to obtain the second keyword set corresponding to each behavioral analysis data of each user; in specific applications, take the word segmentation set corresponding to any behavioral analysis data of any user as an example to illustrate the word combination process, as shown in the following steps S35a to S35j.

[0116] S35a. For any behavior analysis data of any user, determine whether the i-th word in the word segmentation set corresponding to the behavior analysis data belongs to the first keyword set, where i starts from 1; in specific applications, it is necessary to determine whether the i-th word in the word segmentation set corresponding to any behavior analysis data is the first keyword in the first keyword set. If so, proceed to the following step S35b.

[0117] S35b. If yes, determine whether the (i-1)th word satisfies the first preset condition, wherein the first preset condition is that the (i-1)th word exists in the word segmentation set corresponding to any of the behavior analysis data and does not belong to the first keyword set; in specific application, it is to determine whether the (i-1)th word is non-empty, that is, i-1 must be greater than 0, and equal to 0 means that it does not exist in the aforementioned word segmentation set. For example, for the first word in the word segmentation set, there is no preceding word, so the first preset condition is not satisfied; while for the second word, there is a preceding word, so it is necessary to determine whether the preceding word belongs to the first keyword set. If it does, then proceed to the following step S35c.

[0118] S35c. If so, then the (i-1)th word and the ith word are merged sequentially to obtain the first combined word, and it is determined whether the first combined word conforms to the word combination rules; specifically, if the (i-1)th word is the shipper and the ith word is the name, then the first combined word is the shipper's name. At the same time, in order to make the combined words a meaningful combined keyword, it is also necessary to determine whether the first combined word conforms to the word combination rules. Only after conforming to the word combination rules can it be used as a second candidate keyword.

[0119] In this embodiment, for a compound word formed by combining two words, the parts of speech of the two words must be one of the following: adjective + noun, verb + adverb, verb + noun, noun + noun, and noun + verb; while for a compound word formed by combining three words, the parts of speech of the three words must be one of the following: noun + noun + noun, noun + adjective + noun, noun + preposition + noun, verb + noun + noun, noun + noun + verb, noun + verb + noun, adjective + noun + verb, and noun + conjunction + noun.

[0120] If the first combination of words conforms to the aforementioned word combination rules, then the first combination of words can be used as a second candidate keyword, as shown in step S35d below.

[0121] S35d. If the first combined word conforms to the word combination rules, then the first combined word is used as the second candidate keyword, and the weight value of the second candidate keyword is the weight value of the i-th word; in specific applications, taking the aforementioned example, the first combined word is the name of the shipper, and both words are names. Therefore, the name of the shipper can be used as a second candidate keyword, and the weight value of the first combined word is the weight value of the name.

[0122] After determining whether the word preceding the i-th word can be combined, it is also necessary to determine whether the word following the i-th word can be combined. The determination process is shown in step S35e below.

[0123] S35e. Determine whether the (i+1)th word satisfies the second preset condition, wherein the second preset condition is that the (i+1)th word exists in the word segmentation set corresponding to any behavior analysis data and belongs to the first keyword set; in specific applications, it is still necessary to determine whether the (i+1)th word is non-empty, that is, i+1 must be less than or equal to n (n is the total number of words in the word segmentation set corresponding to any behavior analysis data). If it is non-empty (that is, it exists in the word segmentation set), then it is also necessary to determine whether it is the first keyword in the first keyword set. If so, word merging can be performed, as shown in step S35f below; at the same time, in this embodiment, if the (i-1)th word in step S35b does not exist in the word segmentation set corresponding to any behavior analysis data, then step S35e is executed directly, and if the (i+1)th word does not exist in the word segmentation set, it means that the last word in the word segmentation set has been determined. At this time, the word combination processing process can be ended.

[0124] S35f. If the second preset condition is met, the i-th word and the (i+1)-th word are merged sequentially to obtain a second combined word, and it is determined whether the second combined word conforms to the word combination rule; in this embodiment, the word combination rule can be referred to the aforementioned example, and its judgment principle is the same as the aforementioned step S35d principle, and will not be repeated here.

[0125] After determining that the second combination of words conforms to the word combination rules, it can also be used as a second candidate keyword, as shown in step S35g below.

[0126] S35g. If the second combined word conforms to the word combination rules, then the second combined word is also used as the second candidate keyword, and the weight value of the second candidate keyword corresponding to the second combined word is the sum of the weight values ​​of the i-th word and the (i+1)-th word; in specific applications, the weight of the second combined word as the second candidate keyword is the sum of the weights of the two words.

[0127] In this embodiment, if the (i+1)th word exists in the word segmentation set corresponding to any of the behavior analysis data and does not belong to the first keyword set, the i-th word can be merged with the (i+1)th word to obtain the second combined word. However, at this time, the weight of the second combined word is the weight of the i-th word, and not the sum of the weight values ​​of the i-th word and the (i+1)th word in the aforementioned step S34g.

[0128] Based on the aforementioned steps, the combination of the i-th word and its adjacent words can be completed. After the combination is completed, the next word in the word segmentation set can be combined with its adjacent words until all words in the word segmentation set have been traversed. The loop process is shown in step S35h below.

[0129] S35h. Increment i by 1 and re-determine whether the i-th word is in the first keyword set, so that when i is greater than n, a second candidate keyword set is obtained, where n is the total number of words in the word segmentation set corresponding to any behavior analysis data.

[0130] In this embodiment, if the i-th word in step S35a does not belong to the first keyword set, the method further includes the following steps A to D.

[0131] A. Determine whether the (i-1)th word and the (i+1)th word both belong to the first keyword set and both exist in the word segmentation set corresponding to any behavior analysis data.

[0132] B. If so, then combine the (i-1)th word, the ith word, and the (i+1)th word in sequence to obtain the third combination word.

[0133] C. Determine whether the third combination of words conforms to the word combination rules; In this embodiment, the judgment rule of step C can refer to the combination rule of three words in the aforementioned word combination rules, and its principle is the same as that of step S35d, so it will not be repeated here.

[0134] D. If so, the third combined word is also used as a second candidate keyword, wherein the weight value of the second candidate keyword corresponding to the third combined word is the sum of the weight values ​​of the (i-1)th word and the (i+1)th word; in this embodiment, after determining that the third combined word conforms to the word combination rules, it can also be used as a second candidate keyword. At this time, the word combination process of the i-th word ends, and the word combination of the next word needs to be carried out, that is, the aforementioned step S35h is executed.

[0135] Similarly, in step S35b, if the (i-1)th word exists in the word segmentation set corresponding to any of the behavior analysis data and belongs to the first keyword set, then the method further includes the following steps E to H.

[0136] E. Determine whether the (i+1)th word satisfies the second preset condition.

[0137] F. If so, then the (i-1)th word, the ith word, and the (i+1)th word are merged in sequence to obtain the fourth combination word.

[0138] G. Determine whether the fourth combination of words conforms to the word combination rules; in this embodiment, the principle of step G is the same as that of step D mentioned above, and will not be repeated here.

[0139] H. If so, the fourth combined word is also used as the second candidate keyword, wherein the weight value of the second candidate keyword corresponding to the fourth combined word is the sum of the weight values ​​of the i-th word and the (i+1)-th word; in this embodiment, the principle of the aforementioned steps E to H is the same as that of the aforementioned steps A to D, which is to merge the (i-1)-th word, the i-th word, and the (i+1)-th word. The only difference is that the weight values ​​of the combined second candidate keywords are different. The larger the weight value, the more important the word is, which is an important basis for subsequent keyword extraction.

[0140] In this embodiment, if the combined words obtained by the aforementioned steps do not conform to the word combination rules, they are discarded and not used as the second candidate keywords, and the next step is directly performed.

[0141] After the aforementioned steps S35a~S35h, steps A~D and steps E~H, the combination of all words in the word segmentation set corresponding to any behavior analysis data can be completed, resulting in a second candidate keyword set composed of multiple second candidate keywords. In this embodiment, in order to remove second candidate keywords with the same meaning and reduce the interference of irrelevant information on keyword extraction, a keyword screening step is also set, as shown in step S35i below.

[0142] S35i. Perform keyword filtering on the second candidate keyword set, filter out second candidate keywords with subset relationships, and extract the second candidate keywords that are subsets. In specific applications, second candidate keywords with subset relationships are those in which one second candidate keyword contains another second candidate keyword. For example, if one second candidate keyword is "shipper's name" and the other second candidate keyword is "shipper's name Zhang Moumou", then "shipper's name Zhang Moumou" is a subset of "shipper's name". In this case, it is necessary to extract "shipper's name Zhang Moumou". Of course, the extraction process for other second candidate keywords with subset relationships is the same as the example above, and will not be repeated here.

[0143] After extracting the second candidate keywords as a subset, the remaining second candidate keywords in the second candidate keyword set can be combined to form the second keyword set of any user's behavior analysis data, as shown in step S35j below.

[0144] S35j. Using the second candidate keywords as a subset and the filtered second candidate keyword set, a second keyword set is formed for any behavior analysis data of any user.

[0145] After obtaining the second keyword set of any behavior analysis data of any user, the first keyword set can be combined to obtain the actual keyword set of any behavior analysis data. Of course, the first keyword set and the second keyword set are extracted for each behavior analysis data of other users using the aforementioned steps. After extraction, the keyword sets can be merged, as shown in steps S36 and S37 below.

[0146] S36. Merge the first keyword set and the second keyword set corresponding to each behavioral analysis data of each user to obtain the actual keyword set of each behavioral analysis data of each user.

[0147] S37. Remove duplicate keywords from each user's actual keyword set to obtain each user's express delivery business data based on the remaining keywords in each actual keyword set. In specific applications, since the first keyword set and the second keyword set may contain the same keywords, after merging to obtain the actual keyword set, duplicate keywords need to be deleted. After deletion, the remaining keywords in the actual keyword set can be used as each user's express delivery business data. Alternatively, in this embodiment, keywords with a weight value greater than a preset threshold (preset by the maintenance personnel) can also be selected as the corresponding user's express delivery business data. Of course, both of the aforementioned keyword extraction methods can be used and can be specifically set according to actual use, which is not limited here.

[0148] In this embodiment, each user's express delivery data corresponds to the behavioral analysis data; that is, the number of behavioral analysis data points corresponds to the number of express delivery data points.

[0149] In this embodiment, the extracted keywords are equivalent to the key information in the aforementioned express order data, such as the name of the sender, the name of the recipient, the mode of transport, the type of goods transported, and the name of the express delivery company. After obtaining the express business data of each user, the express business screen can be constructed, as shown in steps S4 and S5 below.

[0150] S4. Based on the express delivery business data of each user, generate express delivery business tags for each user; in specific applications, the process of constructing the block delivery business tags is as shown in steps S41 to S44 below.

[0151] S41. Based on each user's express delivery business data, obtain each user's sending and receiving business data. The sending business data includes the sending transportation method, the type of item being sent, and the sending courier company. The receiving business data includes the sender's name, the type of item being received, the receiving transportation method, and the transport courier company. In specific applications, the key information extracted above, namely the sender's name in the express delivery business data, can be used to determine whether the key information in any behavioral analysis data is sending or receiving key information. If the sender's name is the name corresponding to the user's account logged into the express delivery business terminal, then it is considered sending key information; otherwise, it is considered receiving key information. Thus, according to the aforementioned rules, the express delivery business data corresponding to each user's behavioral analysis data can be divided into sending business data and receiving business data.

[0152] After the courier service data of each user is divided, a courier service tag can be generated based on the courier service data, and a receiver service tag can be generated based on the receiver service data, as shown in steps S42, S43 and S44 below.

[0153] S42. Calculate the frequency of shipping methods, item types, and courier companies in each user's shipping data to obtain each user's preferred shipping methods, item types, and courier companies.

[0154] S43. Perform frequency statistics on the sender's name, type of received item, mode of transport, and courier company in each user's receipt business data to obtain each user's preferred sender's name, preferred type of received item, preferred mode of transport, and preferred courier company.

[0155] In this embodiment, step S42 involves counting the frequency of transportation methods (such as car, air or train), types of items sent, and courier companies used in all the shipping data of each user. For example, within a preset time period, the number of times the car was used for shipping is 3, the number of times the air was used is 15, and the number of times the train was used is 5. The number of times each type of item was sent includes: 3 times for electronic products, 2 times for fresh produce, and 18 times for household goods. The number of times each courier company was used includes: 15 times for courier company A, 5 times for courier company C, and 3 times for courier company B.

[0156] Therefore, based on the aforementioned frequency statistics, we can determine each user's preferred shipping method, preferred item type, and preferred courier company. In practical applications, these preferences can be sorted from highest to lowest frequency, with the top three preferences being taken as the user's shipping preferences.

[0157] Similarly, step S43 counts the frequency of the transportation method (such as car, air or train), the type of item received, and the courier company used in all the receiving business data of each user. The statistical principle is the same as the example in step S42 above, and will not be repeated here. After the statistics are completed, they can also be sorted in descending order of usage frequency, and the top 3 in the sorted order can be regarded as the user's receiving preference.

[0158] Finally, based on the results obtained in steps S42 and S43, the sending and receiving service tags for each user can be generated, as shown in step S44 below.

[0159] S44. Generate a shipping service tag for each user by utilizing their preferred shipping method, preferred item type, and preferred courier company. Similarly, generate a receiving service tag for each user by utilizing their preferred sender name, preferred item type, preferred shipping method, and preferred courier company. These shipping and receiving service tags are then used to form the user's overall courier service tag. In practical applications, the user's preferred shipping method, preferred item type, and preferred courier company can be used as the shipping service tag, while the user's preferred sender name, preferred item type, preferred shipping method, and preferred courier company can be used as the receiving service tag.

[0160] After obtaining the express delivery service tags for each user, the express delivery service profile can be constructed, as shown in step S5 below.

[0161] S5. Based on the express delivery service tags of each user, construct the express delivery service profile of each user and store the express delivery service profile of each user in the kudu database; in specific applications, the express delivery service profile construction process is as shown in steps S51 and S52 below.

[0162] S51. Obtain the account information of each user and generate account tags for each user based on the account information. In specific applications, the account information may include, but is not limited to: username, real name, date of birth, gender, occupation, valid ID number and hobbies. The aforementioned account information can form the user's account tags, that is, the account tags are the user's own tags.

[0163] After obtaining the user's account tags, the delivery and receiving service tags of each user can be combined to generate a courier service profile for each user, as shown in step S52 below.

[0164] S52. Based on each user's account tags, sending service tags, and receiving service tags, construct a courier service profile for each user; in specific applications, the courier service profile can be stored in image format, but is not limited to, so that it can be directly displayed in image format when querying later, so as to clearly display the user's profile information.

[0165] In this embodiment, the present invention uses the Kudu database to store the express delivery business profiles of each user obtained from the analysis. The Kudu database supports partitioned viewing, has high scalability, and supports update operations and SQL query operations. Therefore, it can meet the storage requirements of massive express delivery business profiles and can be quickly read during analysis.

[0166] After completing the construction of the express delivery business profile for each user, the user's express delivery behavior can be analyzed and displayed based on the client of the operation and maintenance personnel, as shown in steps S6, S7 and S8 below.

[0167] S6. Receive a behavior analysis request, and based on the behavior analysis request, match the target user's express delivery business profile in the Kudu database and the target user's business behavior data in the HBase database, wherein the target user is the user whose behavior is to be analyzed in the behavior analysis request; in a specific application, for example, the behavior analysis request is sent from the client of the operation and maintenance personnel to the server, and the example behavior analysis request may include, but is not limited to, the real name or username of the user to be analyzed, thereby enabling data to be searched in the Kudu database and the HBase database based on the user's real name or username.

[0168] After retrieving the target user's business behavior data from the HBase database, data statistics can be performed based on the business behavior data to obtain the target user's behavior statistics, as shown in step S7 below.

[0169] S7. Based on the target user's business behavior data, generate the target user's behavior statistics; in specific applications, this can include, but is not limited to, counting the number of times the target user sends or receives packages, the number of devices used, etc., and finally, outputting the statistics in the form of a statistical table.

[0170] After obtaining the target user's behavioral statistics, they can be returned to the operations and maintenance personnel's client along with the target user's express delivery business profile and displayed visually to complete the analysis of the target user's express delivery behavior, as shown in step S8 below.

[0171] S8. Visualize the express delivery business profile and behavioral statistics of the target user to complete the express delivery behavior analysis of the target user; in specific applications, the interaction process between the operation and maintenance personnel's client and this server can be found in [reference needed]. Figure 4 As shown, at the same time, operations and maintenance personnel can learn about the target user's express delivery preferences and needs based on express delivery business profiles and behavioral statistics, thereby enabling personalized services to users.

[0172] Therefore, through the express delivery user behavior analysis method described in detail in steps S1 to S8 above, this invention, based on HBase and Kudu databases, can persistently store massive amounts of express delivery data and express delivery business profiles, and can respond to user behavior analysis query requests in real time, exhibiting high reliability. Furthermore, it can quickly expand its capacity as business volume increases, demonstrating excellent scalability. In addition, this invention can generate express delivery business profiles for each user based on their business behavior data, and upon receiving a behavior analysis request, it can also read business behavior data from the HBase database and generate behavior statistics. Thus, combined with the user's express delivery business profile, it is possible to analyze user express delivery behavior, understand user express delivery needs, and thereby provide data support for personalized services in the express delivery industry.

[0173] In one possible design, the second aspect of this embodiment, based on the first aspect of the embodiment, generates delivery service promotion information and product promotion information according to the user's express delivery user profile, and pushes them to the user to achieve personalized service for the user. The push process is as follows: steps S9 to S11.

[0174] S9. Based on the parcel delivery service tags in each user's courier service profile, generate parcel delivery service promotion information for each user. This parcel delivery service promotion information includes at least one piece of service promotion data, and each piece of service promotion data includes: the courier company for parcel delivery, the mode of transport, the transport time, the type of insurance for the item, and the total price for the parcel delivery. In specific applications, as previously explained, the top three items in the parcel delivery service data—the mode of transport, the type of item being delivered, and the courier company—are used as the user's parcel delivery service tags. Therefore, corresponding service promotion data can be generated according to the sorting order, thus forming the parcel delivery service promotion information.

[0175] To elaborate on the examples mentioned above:

[0176] The top-ranked courier company is Company A, the top-ranked transportation method is air transport, and the top-ranked item type is household goods. Therefore, the generated business promotion data would be: Courier Company: Company A; Transportation Method: Air Transport; Transportation Time: 1 day (within 500km), 1-2 days (500-1000km), 2-3 days (1000-2000km), 3-5 days (over 2000km); Item Insurance Type (Loss Insurance); and Total Shipping Price (within 500km, Loss Insurance, 30 RMB; 500-1000km, Loss Insurance, 60 RMB; 1000-2000km, Loss Insurance, 100 RMB; over 2000km, Loss Insurance, 200 RMB). Of course, the second and third ranked tags can also generate business promotion data according to the same principle. Finally, the aforementioned three pieces of business promotion data constitute the shipping business promotion information.

[0177] In this embodiment, the insurance type for an item can be adapted according to the item type. For example, an insurance type lookup table is stored on the server, which records the insurance types corresponding to each item type. When needed, the item type can be searched for.

[0178] Similarly, based on the package receiving service tags in the express delivery business profile, the preferred product types of each user can be obtained, so as to generate product promotion information for each user based on the preferred product types. The generation process is shown in step S10 below.

[0179] S10. Based on the receiving service tags in the user's express delivery service profile, obtain the preferred product types of each user, and generate product promotion information for each user based on their preferred product types. In specific applications, as previously explained, the top three preferred receiving item types are used as one of the tags in the receiving service tags. Therefore, the preferred receiving item types in the profile can be used as preferred product types, and the product promotion information can be, but is not limited to, discount information for the aforementioned preferred product types, which can be obtained by associating with e-commerce platforms. For example, if the preferred receiving item types in the profile are electronic products, fitness equipment, and beauty products, then the product promotion information for electronic products can include, but is not limited to, the name, price, product image, discount information, and discount period of the electronic products, and the display format can be, but is not limited to, images. Of course, the content of the product promotion information for fitness equipment and beauty products is the same as the previous example, and will not be repeated here.

[0180] After obtaining the parcel delivery service promotion information and product promotion information from each user, the information can be pushed out, as shown in step S11 below.

[0181] S11. Send the parcel delivery service promotion information of each user to the corresponding terminal of each user, and display the corresponding product promotion information of each user when each user enters the main interface of the express delivery service terminal; in specific applications, the parcel delivery service promotion information may be sent to the corresponding terminal of each user via SMS, or sent to the corresponding account of each user in the express delivery service terminal via notification; and the product promotion information is displayed in the form of an image when the user enters the main interface of the express delivery service terminal.

[0182] Thus, through the aforementioned steps S9 to S11, the present invention can realize personalized services for users based on the user's express delivery business cycle, thereby improving the service capabilities of express delivery.

[0183] like Figure 5 As shown, the third aspect of this embodiment provides a hardware device for implementing the express delivery user behavior analysis method described in the first and second aspects of the embodiments, comprising:

[0184] The acquisition unit is used to acquire business behavior data generated by each user when performing business operations on the express delivery business terminal, and to store the business behavior data of each user in the HBase database.

[0185] The data extraction unit is used to obtain the express delivery order data corresponding to each user's business behavior data within a preset time period, so as to serve as behavioral analysis data for each user.

[0186] The key information extraction unit is used to extract key information from the behavioral analysis data of each user, so as to obtain the express delivery business data of each user based on the extracted key information.

[0187] The tag generation unit is used to generate express delivery service tags for each user based on the express delivery service data of each user.

[0188] The profile building unit is used to build a profile of each user's express delivery business based on each user's express delivery business tags, and store the profile of each user's express delivery business in the kudu database.

[0189] An analysis unit is configured to receive a behavior analysis request, and based on the behavior analysis request, match the target user's express delivery business profile in the Kudu database and match the target user's business behavior data in the HBase database, wherein the target user is the user whose behavior is to be analyzed in the behavior analysis request.

[0190] The analysis unit is also used to generate behavioral statistics data of the target user based on the target user's business behavior data.

[0191] The display unit is used to visualize the express delivery business profile and behavioral statistics of the target user in order to complete the express delivery behavior analysis of the target user.

[0192] The working process, working details and technical effects of the device provided in this embodiment can be found in the first and second aspects of the embodiment, and will not be repeated here.

[0193] like Figure 6 As shown, the fourth aspect of this embodiment provides another device for analyzing the behavior data of express delivery users. Taking the device as an electronic device as an example, it includes: a memory, a processor, and a transceiver that are connected in sequence. The memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the behavior analysis method for express delivery users as described in the first and / or second aspects of the embodiments.

[0194] For specific examples, the memory may include, but is not limited to, random access memory (RAM), read-only memory (ROM), flash memory, first-in-first-out (FIFO) memory, and / or first-in-last-out (FILO) memory, etc.; specifically, the processor may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor may be implemented using at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), PLA (Programmable Logic Array). The processor may also include a main processor and a coprocessor. The main processor, also known as the CPU (Central Processing Unit), is used to process data in the wake-up state; the coprocessor is a low-power processor used to process data in the standby state.

[0195] In some embodiments, the processor may integrate a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the screen. For example, the processor may not be limited to microprocessors of the STM32F105 series, reduced instruction set computer (RISC) microprocessors, x86 architecture processors, or processors with integrated neural network processing units (NPUs). The transceiver may be, but is not limited to, a Wi-Fi transceiver, a Bluetooth transceiver, a General Packet Radio Service (GPRS) transceiver, a ZigBee (a low-power LAN protocol based on the IEEE 802.15.4 standard) transceiver, a 3G transceiver, a 4G transceiver, and / or a 5G transceiver. Furthermore, the device may also include, but is not limited to, a power module, a display screen, and other necessary components.

[0196] The working process, working details and technical effects of the electronic device provided in this embodiment can be found in the first or second aspect of the embodiment, and will not be repeated here.

[0197] The fifth aspect of this embodiment provides a storage medium that stores instructions containing the express delivery user behavior analysis method described in the first and / or second aspects of the embodiments. That is, the storage medium stores instructions that, when the instructions are run on a computer, execute the express delivery user behavior analysis method as described in the first and / or second aspects.

[0198] The storage medium refers to a carrier for storing data, which may include, but is not limited to, floppy disks, optical disks, hard disks, flash memory, USB flash drives, and / or memory sticks. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0199] The working process, working details and technical effects of the storage medium provided in this embodiment can be found in the first and / or second aspects of the embodiment, and will not be repeated here.

[0200] The sixth aspect of this embodiment provides a computer program product containing instructions that, when executed on a computer, cause the computer to perform the behavior analysis method for express delivery users as described in the first and / or second aspects of the embodiments, wherein the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device.

[0201] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for analyzing the behavior of express delivery users, characterized in that, include: Acquire business behavior data generated by each user when performing business operations on the express delivery business terminal, and store the business behavior data of each user in the HBase database; Obtain the express delivery order data corresponding to each user's business behavior data within a preset time period, and use it as behavioral analysis data for each user; The behavioral analysis data of each user is processed to extract key information, so as to obtain the express delivery business data of each user based on the extracted key information; Based on the express delivery business data of each user, generate express delivery business tags for each user. Based on each user's express delivery business tags, construct an express delivery business profile for each user and store the express delivery business profile for each user in the kudu database; Receive a behavior analysis request, and based on the behavior analysis request, match the express delivery business profile of the target user in the Kudu database and match the business behavior data of the target user in the HBase database, wherein the target user is the user whose behavior is to be analyzed in the behavior analysis request; Based on the target user's business behavior data, generate the target user's behavior statistics. The express delivery business profile and behavioral statistics of the target users are visualized to complete the analysis of the express delivery behavior of the target users; The behavioral analysis data of each user is processed to extract key information, and based on the extracted key information, the express delivery business data of each user is obtained, including: For each piece of behavior analysis data for each user, word segmentation is performed on each piece of behavior analysis data to obtain a word segmentation set for each user's behavior analysis data; Calculate the importance factor value of each word in each word segmentation set for each user. The importance factor value of any word includes the word frequency value, word length value, and part-of-speech value of that word. Based on the importance factor value of each word in each word segmentation set of each user, the weight value of each word in each word segmentation set of each user is calculated; Based on the weight values ​​of each word in each word segmentation set for each user, the first keyword set corresponding to each piece of behavioral analysis data for each user is obtained; Based on the first set of keywords corresponding to each behavioral analysis data of each user, words are combined for each behavioral analysis data of each user to obtain the second set of keywords corresponding to each behavioral analysis data of each user; Merge the first keyword set and the second keyword set corresponding to each behavioral analysis data of each user to obtain the actual keyword set of each user's behavioral analysis data; Duplicate keywords are removed from each user's actual keyword set, and the remaining keywords in each actual keyword set are used to obtain each user's express delivery business data.

2. The method as described in claim 1, characterized in that, Based on the first set of keywords corresponding to each behavioral analysis data point of each user, word combinations are performed on each behavioral analysis data point of each user to obtain the second set of keywords corresponding to each behavioral analysis data point of each user, including: For any user’s behavior analysis data, determine whether the i-th word in the word segmentation set corresponding to the behavior analysis data belongs to the first keyword set, where i starts from 1; If so, determine whether the (i-1)th word satisfies the first preset condition, wherein the first preset condition is that the (i-1)th word exists in the word segmentation set corresponding to any behavior analysis data and does not belong to the first keyword set; If so, then the (i-1)th word and the ith word are merged sequentially to obtain the first combined word, and it is determined whether the first combined word conforms to the word combination rules; If the first combination of words conforms to the word combination rules, then the first combination of words is used as the second candidate keyword, and the weight value of the second candidate keyword is the weight value of the i-th word; Determine whether the (i+1)th word satisfies the second preset condition, wherein the second preset condition is that the (i+1)th word exists in the word segmentation set corresponding to any behavior analysis data and belongs to the first keyword set; If the second preset condition is met, the i-th word and the (i+1)-th word are merged sequentially to obtain a second combined word, and it is determined whether the second combined word conforms to the word combination rules; If the second combination word conforms to the word combination rules, then the second combination word is also used as the second candidate keyword, and the weight value of the second candidate keyword corresponding to the second combination word is the sum of the weight values ​​of the i-th word and the (i+1)-th word; Increment i by 1 and re-determine whether the i-th word is in the first keyword set, so that when i is greater than n, a second candidate keyword set is obtained, where n is the total number of words in the word segmentation set corresponding to any behavior analysis data; The second candidate keyword set is filtered to identify second candidate keywords that have a subset relationship, and the second candidate keywords that are subsets are extracted. Using the second candidate keywords as a subset and the filtered second candidate keyword set, a second keyword set is formed for any behavior analysis data of any user.

3. The method as described in claim 2, characterized in that, If the i-th word does not belong to the first keyword set, the method further includes: Determine whether the (i-1)th word and the (i+1)th word both belong to the first keyword set and both exist in the word segmentation set corresponding to any behavior analysis data; If so, then the (i-1)th word, the ith word, and the (i+1)th word are merged sequentially to obtain the third combination word; Determine whether the third combination of words conforms to the word combination rules; If so, the third combination of words is also used as the second candidate keyword, wherein the weight value of the second candidate keyword corresponding to the third combination of words is the sum of the weight values ​​of the (i-1)th word and the (i+1)th word; Increment i by 1 and re-determine whether the i-th word is in the first keyword set, so that when i is greater than n, a second candidate keyword set is obtained, where n is the total number of words in the word segmentation set corresponding to any behavior analysis data.

4. The method as described in claim 2, characterized in that, If the (i-1)th word exists in the word segmentation set corresponding to any of the behavior analysis data and belongs to the first keyword set, then the method further includes: Determine whether the (i+1)th word satisfies the second preset condition; If so, then the (i-1)th word, the ith word, and the (i+1)th word are merged sequentially to obtain the fourth combination word; Determine whether the fourth combination of words conforms to the word combination rules; If so, the fourth combination of words is also used as the second candidate keyword, wherein the weight value of the second candidate keyword corresponding to the fourth combination of words is the sum of the weight values ​​of the i-th word and the (i+1)-th word; Increment i by 1 and re-determine whether the i-th word is in the first keyword set, so that when i is greater than n, a second candidate keyword set is obtained, where n is the total number of words in the word segmentation set corresponding to any behavior analysis data.

5. The method as described in claim 1, characterized in that, Based on each user's express delivery business data, express delivery business tags are generated for each user, including: Based on the express delivery business data of each user, the sending and receiving business data of each user are obtained. The sending business data includes the sending transportation method, the type of sending item, and the sending express company. The receiving business data includes the sender's name, the type of receiving item, the receiving item transportation method, and the transportation express company. By analyzing the frequency of shipping methods, item types, and courier companies in each user's shipping data, we can obtain each user's preferred shipping methods, item types, and courier companies. Frequency statistics are performed on the sender's name, type of received item, mode of transport, and courier company in each user's package receipt data to obtain each user's preferred sender's name, preferred type of received item, preferred mode of transport, and preferred courier company. By utilizing each user's preferred shipping method, preferred item type, and preferred courier company, a shipping service tag is generated for each user. Similarly, by utilizing each user's preferred sender name, preferred item type, preferred shipping method, and preferred courier company, a receiving service tag is generated for each user. In this way, the shipping service tags and receiving service tags of each user are used to form the courier service tags of each user. Accordingly, based on each user's express delivery service tags, a user's express delivery service profile is constructed, including: Obtain each user's account information and generate account tags for each user based on that information; Based on each user's account tags, shipping service tags, and receiving service tags, a courier service profile for each user is constructed.

6. The method as described in claim 1, characterized in that, The express delivery service profile includes sending service tags and receiving service tags. After constructing the express delivery service profile for each user, the method further includes: Based on the parcel delivery service tags in each user's express delivery service profile, parcel delivery service promotion information is generated for each user. The parcel delivery service promotion information includes at least one piece of service promotion data, and each piece of service promotion data includes: the parcel delivery express company, the transportation method, the transportation time, the type of item insurance, and the total price of the parcel delivery. Based on the recipient service tags in the user's express delivery business profile, obtain the preferred product types of each user, and generate product promotion information for each user based on the preferred product types of each user; The system sends parcel delivery service promotion information to each user's corresponding terminal and displays product promotion information to each user when they enter the main interface of the express delivery service terminal.

7. A device for analyzing the behavior of express delivery users, characterized in that, The device is used to perform the behavior analysis method for express delivery users according to any one of claims 1 to 6, wherein the device comprises: The acquisition unit is used to acquire business behavior data generated by each user when performing business operations on the express delivery business terminal, and store the business behavior data of each user in the HBase database. The data extraction unit is used to obtain the express delivery order data corresponding to each user's business behavior data within a preset time period, so as to serve as behavioral analysis data for each user. The key information extraction unit is used to extract key information from the behavioral analysis data of each user, so as to obtain the express delivery business data of each user based on the extracted key information. The tag generation unit is used to generate express delivery service tags for each user based on the express delivery service data of each user. The profile building unit is used to build a profile of each user's express delivery business based on each user's express delivery business tags, and store each user's express delivery business profile in the kudu database; An analysis unit is configured to receive a behavior analysis request, and based on the behavior analysis request, match the express delivery business profile of the target user in the Kudu database, and match the business behavior data of the target user in the HBase database, wherein the target user is the user whose behavior is to be analyzed in the behavior analysis request; The analysis unit is also used to generate behavioral statistics data of the target user based on the target user's business behavior data; The display unit is used to visualize the express delivery business profile and behavioral statistics of the target user in order to complete the express delivery behavior analysis of the target user.

8. An electronic device, characterized in that, include: A memory, a processor, and a transceiver are sequentially connected in communication, wherein the memory is used to store computer programs, the transceiver is used to send and receive messages, and the processor is used to read the computer programs and execute the behavior analysis method for express delivery users as described in any one of claims 1 to 6.

9. A storage medium, characterized in that, The storage medium stores instructions that, when executed on a computer, perform the behavior analysis method for express delivery users as described in any one of claims 1 to 6.

Images