Take-out number identification method and user behavior identification method

Abstract

The invention relates to a take-out number identification method and a user behavior identification method. Comprising the following steps: acquiring call data corresponding to a to-be-identified number and internet data for target software, and acquiring software use data when multiple users use the target software; inputting the call data and the internet surfing data into a pre-trained target classification model, obtaining a classification result corresponding to the to-be-identified number, and determining the to-be-identified number as a pre-selected number when the classification result is a take-out number; the number of times of effective call-out is screened out from call data corresponding to the preselected number, the effective call-out is call-out in a preset scene, and call-out time corresponding to the effective call-out and software opening time in the software use data meet a first preset condition; and determining that the pre-selected number is a take-out number under the condition that the number of times of effective calling meets a second preset condition, thereby improving the identification accuracy of the take-out number.

Description

Technical Field

[0001] The present disclosure relates to the field of Internet technology, and in particular to a takeaway number recognition method and a user behavior recognition method. Background Art

[0002] In the related art, the number can be judged based on the features corresponding to the takeaway number to determine whether the number is a takeaway number. However, in actual applications, the features corresponding to the marketing number and the courier number are highly similar to the features corresponding to the takeaway number. Therefore, when identifying the courier number, it is easy to identify the marketing number or the courier number as a takeaway number. Summary of the invention

[0003] In view of this, the embodiments of the present disclosure provide a takeout number identification method and a user behavior identification method to solve the problems existing in the related art.

[0004] According to a first aspect of an embodiment of the present disclosure, a method for identifying a takeout number is provided, the method comprising: obtaining call data corresponding to a number to be identified and Internet access data for a target software, and obtaining software usage data when multiple users use the target software; inputting the call data and Internet access data into a pre-trained target classification model to obtain a classification result corresponding to the number to be identified, and determining that the number to be identified is a pre-selected number when the classification result is a takeout number, wherein the target classification model is obtained by training a preset classification model through training samples, and the training samples include sample call data and sample Internet access data corresponding to different sample numbers; filtering out the number of valid outgoing calls from the call data corresponding to the pre-selected number, wherein the valid outgoing calls are calls under a preset scenario, and the call out time corresponding to the valid outgoing calls and the software opening time in the software usage data meet a first preset condition; and determining that the pre-selected number is a takeout number when the number of valid outgoing calls meets a second preset condition.

[0005] A second aspect of an embodiment of the present disclosure provides a user behavior identification method, the method comprising: obtaining the user's call data and software usage data when using the target software, and determining an incoming takeout number and corresponding incoming data from the call data, wherein the takeout number is determined by the takeout number identification method of an embodiment of the present disclosure; matching the incoming data and the software usage data, and determining that the user's behavior of using the target software is a valid behavior when the incoming data and the software usage data match successfully.

[0006] According to a third aspect of the embodiments of the present disclosure, there is provided a takeaway number identification device, which is applied to the takeaway number identification method according to the first aspect, and the device comprises: an acquisition module, which is used to acquire call data corresponding to the number to be identified and Internet access data for the target software, and acquire software usage data when multiple users use the target software; a model processing module, which is used to input the call data and Internet access data into a pre-trained target classification model to obtain a classification result corresponding to the number to be identified, and determine that the number to be identified is a pre-selected number when the classification result is a takeaway number, wherein the target classification model is obtained by training a preset classification model through training samples, and the training samples include sample call data and sample Internet access data corresponding to different sample numbers; a screening module, which screens out the number of valid outgoing calls from the call data corresponding to the pre-selected number, wherein the valid outgoing calls are calls under a preset scenario, and the call out time corresponding to the valid outgoing calls and the software opening time in the software usage data meet a first preset condition; a determination module, which is used to determine that the pre-selected number is a takeaway number when the number of valid outgoing calls meets a second preset condition.

[0007] According to a fourth aspect of an embodiment of the present disclosure, a user behavior identification device is provided, which is applied to the user behavior identification device of the second aspect, and the device includes: an acquisition module, which is used to acquire the user's call data and software usage data when using the target software, and determine the incoming takeaway number and the corresponding incoming data from the call data, wherein the takeaway number is determined by the takeaway number identification method of an embodiment of the present disclosure; a determination module, which is used to match the incoming data and the software usage data, and when the incoming data and the software usage data match successfully, determine that the user's behavior of using the target software is a valid behavior.

[0008] According to a fifth aspect of an embodiment of the present disclosure, a computer device is provided, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the above-mentioned takeaway number identification method.

[0009] In a sixth aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided, on which a computer program / instruction is stored. When the computer program / instruction is executed by a processor, the steps of the above-mentioned takeaway number identification method are implemented.

[0010] According to a seventh aspect of the present disclosure, a computer program product is provided, and when the computer program / instructions are executed by a processor, the steps of the above-mentioned takeaway number identification method are implemented.

[0011] The above-mentioned at least one technical scheme adopted in the embodiments of the present disclosure can achieve the following beneficial effects: obtaining call data corresponding to the number to be identified and Internet access data for the target software, and obtaining software usage data when multiple users use the target software; inputting the call data and Internet access data into a pre-trained target classification model to obtain a classification result corresponding to the number to be identified, and determining that the number to be identified is a pre-selected number when the classification result is a take-out number, wherein the target classification model is obtained by training a preset classification model through training samples, and the training samples include sample call data and sample Internet access data corresponding to different sample numbers; filtering out the number of valid outgoing calls from the call data corresponding to the pre-selected number, wherein the valid outgoing calls are calls under a preset scenario, and the call out time corresponding to the valid outgoing calls and the software opening time in the software usage data meet a first preset condition; and determining that the pre-selected number is a take-out number when the number of valid outgoing calls meets a second preset condition.

[0012] Based on this, the disclosed embodiment not only starts from the perspective of the number to be identified, obtains the call data corresponding to the number to be identified and the Internet access data for the target software, but also starts from the user side, combines the corresponding software usage data when the user uses the target software, and analyzes the data of different dimensions, so that the number to be identified can be analyzed more accurately to determine whether the number to be identified is a takeaway number. Specifically, firstly, a pre-trained target classification model is used to directly analyze the data corresponding to the number to be identified, and then a preliminary judgment is made on the number to be identified. When the target classification model determines that the number to be identified is a takeaway number, the software usage data corresponding to the user when using the target software is combined to further analyze the number to be identified, so as to determine whether the number to be identified is a takeaway number. It can be seen that the disclosed embodiment can cross-check the Internet access data and call data corresponding to the number to be identified with the software usage data corresponding to the user, thereby improving the accuracy of takeaway number recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The above and other purposes, features and advantages of the present disclosure will become more apparent by describing the embodiments of the present disclosure in more detail in conjunction with the accompanying drawings. The accompanying drawings are used to provide a further understanding of the embodiments of the present disclosure and constitute a part of the specification. Together with the embodiments of the present disclosure, they are used to explain the present disclosure and do not constitute a limitation of the present disclosure. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0014] Figure 1 A flowchart of a takeaway number identification method provided by an exemplary embodiment of the present disclosure;

[0015] Figure 2 A flowchart of a user behavior identification method provided by an exemplary embodiment of the present disclosure;

[0016] Figure 3 A flowchart of another takeout number identification method provided by an exemplary embodiment of the present disclosure;

[0017] Figure 4 A flowchart of another user behavior identification method provided by an exemplary embodiment of the present disclosure;

[0018] Figure 5 A schematic diagram of the structure of a takeaway number recognition device provided by an exemplary embodiment of the present disclosure;

[0019] Figure 6 A schematic diagram of the structure of a user behavior recognition device provided by an exemplary embodiment of the present disclosure;

[0020] Figure 7 A schematic diagram of the structure of an electronic device provided for an exemplary embodiment of the present disclosure;

[0021] Figure 8 A schematic diagram of the structure of a computer system provided for an exemplary embodiment of the present disclosure;

[0022] Fig. 9 A schematic diagram of a computer program product provided for an exemplary embodiment of the present disclosure. DETAILED DESCRIPTION

[0023] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein, which are instead provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0024] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.

[0025] The term "including" and its variations used in this document are open inclusions, that is, "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one other embodiment"; the term "some embodiments" means "at least some embodiments". Relevant definitions of other terms will be given in the description below. It should be noted that the concepts of "first", "second", etc. mentioned in this disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0026] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0027] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0028] At present, with the gradual saturation of traditional services such as voice and traffic of operators, new digital service businesses represented by rights and interests are increasingly valued and gradually become an important revenue growth point for operators. Among them, rights and interests, as an important digital service, can provide users with life rights and interests such as video membership, takeaway coupons, and local consumption discount coupons. At the same time, due to the rapid popularization of mobile Internet, the communications market has gradually entered a saturated state, and the value of traditional personal packages has limited room for improvement. Operators are actively promoting digital transformation, deepening their business areas, and accelerating the development of traffic-based digital services for individual users. Among them, with the continuous growth of the scale of the takeaway user group, takeaway rights and interests have become an important part of the operator's rights and interests digital services.

[0029] Usually, a complete food delivery service process requires two entities, one is the user who orders food delivery, and the other is the food delivery person. Specifically, first, the user needs to place an order for the food delivery product on the food delivery software, the food delivery person will accept the order, and then the food delivery person will deliver the food delivery product to the user. When the user receives the food delivery product, the food delivery service process is considered complete. During this process, the food delivery person and the user need to communicate through each other's numbers. However, usually the user does not make a note of the food delivery number, so the user often rejects the food delivery number as a harassing number. Therefore, how to accurately identify the food delivery number and prompt the user when the food delivery number is called out has become a key issue.

[0030] In the prior art, takeaway numbers can be identified by community through a trained takeaway number recognition algorithm, which can improve the outdoor accuracy of takeaway number community recognition. In this process, the K-Means clustering algorithm is mainly used to cluster the feature vectors collected through the base station and URL, and then identify the takeaway number. It should be understood that the K-means algorithm is a clustering analysis algorithm. Its main idea is to cluster with k points in space as the center and classify the objects closest to them. Each category will have a center point, which is defined according to the mean of the elements in the cluster. Uniform Resource Locator (URL) is the abbreviation of Uniform Resource Locator, also known as URL. It is used to specify the location and access method of resources on the network, and contains information such as the resource's protocol, host name, path, etc., which can uniquely identify resources on the Internet.

[0031] However, in a complete process corresponding to a valid takeaway order, it is also necessary to analyze the user's behavior information, that is, it is necessary to determine whether the user's behavior is a valid takeaway behavior. Only valid takeaway behavior will generate a valid takeaway order. Therefore, it is also necessary to analyze the user's behavior so that the takeaway number can be accurately identified when the user has a valid takeaway behavior.

[0032] In the prior art, detailed information of each takeaway order in the user account can be obtained, and each takeaway order purchased in the user account within the past month can be screened, and the corresponding food attributes in the detailed information of each takeaway order in the user account within the past month can be extracted, and the number of various food attributes in the takeaway orders in the user account within the past month can be counted, and the ratio of the number of various food attributes can be analyzed, and the maximum ratio of the number of food attributes can be screened. At the same time, the takeaway amount of each takeaway order in the user account within the past month can be counted, and the average takeaway amount of the takeaway orders in the user account within the past month can be analyzed, and the user's food preference coefficient for takeaway food can be comprehensively calculated, and the food information recommended by each merchant that meets the recommendation requirements in the takeaway platform can be compared and analyzed, and pushed. Alternatively, according to the location information of the first user, a base station within a preset range can be determined, and the second user to which the terminal that has information interaction with the base station belongs can be determined; the historical feature tag of the second user can be obtained, and the various behavior data of the second user can be obtained from the data source of the core network; according to the historical feature tag and each behavior data, the feature aggregation tag of the second user can be determined, and the feature aggregation tag is used to determine the behavior intention of the second user, thereby realizing the precise service operator information fixed-point push. It should be understood that both of the above-mentioned technologies imply analysis of user behavior, and the validity of the takeaway order can be determined through analysis of user behavior.

[0033] At present, the technologies used to identify takeaway numbers mainly include threshold matching and community clustering calculation. In the above methods, the threshold analysis method is used to identify takeaway numbers. It is necessary to analyze the communication characteristics of a large number of number samples, such as the number of dials and call frequency of the caller number field, in order to set the threshold to distinguish between takeaway numbers and other numbers. The selection of the threshold requires setting the field feature rules in advance, which is easy to cause the normal user numbers with relevant field features to be misjudged, and it is difficult to identify takeaway numbers that do not have number field features. When using community clustering methods to identify takeaway numbers, such as the K-Means clustering algorithm, such algorithms need to rely on experience to set the K value, but the sample K value will have certain differences under different regions, time and other conditions, which is easy to cause inaccurate identification. In addition, because the characteristic indicators of marketing numbers, express numbers and other numbers are relatively similar to those of takeaway numbers, the clustering calculation method is prone to misjudgment in the identification of takeaway numbers. At the same time, as the weak growth of traditional businesses such as operator traffic and voice becomes more and more prominent, new digital services such as rights and interests are urgently needed to drive the growth of large market revenue. Although operators have access to a lot of user information such as calls, text messages, data traffic, and Internet logs, they lack the means to identify users' specific consumption characteristics, making it difficult to build a complete user behavior profile and unable to support the precise marketing of digital services such as takeaway benefits.

[0034] In view of the above problems, the embodiments of the present disclosure provide a takeaway number identification method and a user behavior identification method. Specifically, when identifying the identification number, a pre-trained classification model (XGboost) can be used to classify and identify the identification number, so that the pre-selected number and other numbers can be accurately classified. In the process of identifying the pre-selected number, a series of decision tree models can be constructed and optimized to complete the refined identification of data features, so as to more accurately identify the takeaway number. At the same time, the embodiment of the present disclosure can also identify the user's activity behavior on the takeaway software through the DPI device of the operator's Internet, and combine the comprehensive judgment of the call access of the takeaway number to accurately identify the user group who frequently uses the takeaway platform to order food, thereby supporting the precision marketing of takeaway rights products. Operators can target different marketing resources based on the frequency and quantity of takeaway consumption of different groups of people, and improve the conversion efficiency of marketing resource delivery. It can be seen that the method of the embodiment of the present disclosure avoids the problem of large fluctuations in the recognition success rate under different application scenarios due to the reliance on experience to set the threshold or K value compared with traditional methods such as threshold matching and community clustering. At the same time, the effective incoming calls of the takeaway number are associated with the user's use of the takeaway software, and cross-checked, which can significantly improve the recognition accuracy of the takeaway number. It should be understood that XGBoost, or extreme gradient boosting, is a scalable distributed gradient boosting decision tree machine learning library. XGBoost provides parallel tree boosting capabilities and is an advanced machine learning library for regression, classification, and ranking problems. Deep packet inspection (DPI) equipment is an application-layer-based traffic detection and control technology that is often used by enterprises and Internet service providers to identify and prevent network attacks, track user behavior, block malware, and monitor network traffic.

[0035] Figure 1 A flowchart of a method for identifying a takeout number provided by an exemplary embodiment of the present disclosure. Since takeout numbers have the characteristics of outgoing calls during a specific time period and long-term logins to the takeout software, the present disclosure embodiment can use the XGboost classification model to accurately identify the takeout number based on the communication behavior between the takeout number and the user and the takeout deliveryman's login characteristics to the takeout software, such as Figure 1 As shown, the specific steps include:

[0036] S101, obtaining call data corresponding to the number to be identified and Internet access data for the food delivery software from the operator, and determining characteristic variables of the food delivery number from the call data and Internet access data, wherein the characteristic variables are used to describe basic characteristics of the food delivery number under different backgrounds.

[0037] In some embodiments, in combination with the basic features of the food delivery number, the selected feature variables may include the number of incoming calls, the number of outgoing calls, the frequency of outgoing calls, the frequency of incoming calls, the average call duration, the number of contacted users, the number of outgoing calls during the dining time period, the proportion of outgoing calls during the dining time period, the number of silent calls / missed calls, the call back rate, and the online duration of the food delivery software, etc. The specific reasons for the selection are as follows:

[0038] As for the number of incoming calls and outgoing calls, the deliveryman needs to dial the user's number in the process of delivering the takeaway order, and then notify the user to pick up the takeaway. Since the deliveryman needs to constantly obtain takeaway orders, the deliveryman's number has relatively more outgoing calls and fewer incoming calls compared to normal users. At the same time, due to the delivery distance and efficiency of the deliveryman, the monthly outgoing calls of the takeaway number are relatively small compared to a type of number such as fraudulent numbers. Based on this, the embodiment of the present disclosure analyzes the number of incoming calls and outgoing calls corresponding to the takeaway number as characteristic variables.

[0039] Regarding outgoing and incoming call frequencies, in general, the number of takeout orders during meal time is large, and a delivery person may receive multiple takeout orders at the same time. Therefore, the outgoing frequency of the takeout number is high during meal time. After the user hangs up the takeout number, there will be a corresponding SMS to remind the user to pick up the takeout, so the incoming frequency of the takeout number is generally low.

[0040] Regarding the average call duration, compared with normal numbers, the takeaway numbers have the characteristics of relatively shorter call duration and relatively longer total call duration. It should be understood that the longer call duration means that the total call duration of the takeaway numbers is longer than the total call duration of the normal numbers every day or week.

[0041] Regarding the number of users contacted, compared with other numbers, the number of users contacted by the takeaway number shows a certain relationship with the number of orders received by the deliverymen. The number of users contacted every day or every month is higher than that of other normal numbers.

[0042] As for the number of outbound calls during the dining time period and the proportion of outbound calls during the dining time period, the number of outbound calls and the proportion of outbound calls during the dining time period can significantly distinguish between takeout numbers and express numbers. Therefore, takeout numbers can be identified based on this feature. Specifically, the number of outbound calls of numbers during the peak time interval of takeout orders can be counted. For example, the number of outbound calls of takeout numbers during the peak ordering time areas such as lunch 10:00-13:59 or dinner 17:00-20:59 can be counted each month, and then compared with the number of outbound calls for the entire month, the proportion of outbound calls during the dining time period can be determined. Usually, takeout numbers call out more during the dining time period, and the proportion of outbound calls during the dining time period is larger.

[0043] Regarding the number of silent / missed call transfers, since takeout numbers are usually anonymous numbers, the hang-up rate of takeout numbers is relatively high compared to normal numbers.

[0044] As for the callback rate, the takeaway number will usually place the customer's items at the designated location through the platform or text message after the number is hung up. Therefore, the callback rate of the takeaway number is relatively low.

[0045] As for the online time of food delivery apps, the online time of users accessing food delivery apps such as Meituan and Ele.me can be obtained through the operator's Internet / mobile Internet DPI equipment. However, due to work reasons, food delivery drivers need to log in to food delivery apps in real time during delivery. Therefore, the daily or monthly login time for food delivery apps is much longer than that of the general population.

[0046] Among the above-mentioned characteristic variables, the characteristic variables corresponding to the call data may include the number of incoming calls, the number of outgoing calls, the frequency of outgoing calls, the frequency of incoming calls, the average call duration, the number of contacted users, the number of outgoing calls during the dining time period, the proportion of outgoing calls during the dining time period, the number of silent / missed calls, and the call-back rate, while the characteristic variables corresponding to the Internet data may include the online time of the deliveryman on the food delivery software. Based on this, a characteristic variable table can be constructed, as shown in Table 1, including different characteristic variables and the corresponding characteristic identifiers of the characteristic variables in the model:

[0047] Table 1 Characteristic variables table

[0048]

[0049]

[0050] S102, determining the specific data corresponding to each characteristic variable, and constructing a data set according to the corresponding specific data, and then using the data set to train a preset classification model to obtain a target classification model.

[0051] In some embodiments, the specific data corresponding to the above-mentioned feature variables can be obtained from the operator, and then a data set can be constructed based on the specific data of different feature variables corresponding to different takeaway numbers, and then a certain amount of sample data can be obtained from the data set, and 70% of the data in the sample data is taken to construct a training set, and 30% of the data is taken to construct a test set. Based on this, the preset classification model can be trained using the training and test sets, and the model parameters can be iteratively tuned under the constraints of accuracy, precision, and recall to obtain the target classification model. Among them, the preset classification model can be an XGboost classification model. The embodiment of the present disclosure limits the initial values ​​corresponding to each parameter in the XGboost classification model, as shown in Table 2:

[0052] Table 2 Initial value comparison table

[0053]

[0054]

[0055] Among them, learning rate represents the learning rate; colsample_bytree represents the proportion of features that each decision tree will use; max_depth represents an important hyperparameter; it defines the maximum depth of the decision tree; subsample represents subsample, which is a subsampled sample extracted from the sample; objective represents the target parameter in the model learning task parameter, which specifies the target of the training task; n_estimators represents the number of trees; eval_metric represents the evaluation function, which has no effect on the training of the model, but evaluates the model effect after the model training is completed; early_stopping_rounds represents the early stopping strategy. It should be understood that in XGBoost, when objective = 'multi:softprob', the model will output a probability distribution for each instance of the multi-classification problem. AUC (Area Under Curve) is used to measure the ranking ability. The AUC value is the area under the curve, which is a probability value. The larger the better. mlogloss represents the loss function in the multi-classification problem, also known as the cross entropy loss function, which measures the difference between the probability distribution predicted by the model and the probability distribution of the true label.

[0056] S103, inputting data corresponding to a plurality of numbers to be identified under different characteristic variables into a target classification model to obtain a plurality of pre-selected numbers.

[0057] Specifically, according to the multiple characteristic variables determined in S101, data corresponding to multiple numbers to be identified under different characteristic variables can be obtained from the operator, and these data can be input into the target classification model for judgment, and the numbers to be identified confirmed by the target classification model can be used as pre-selected numbers to wait for further identification. Here, a pre-selected number set V can also be constructed based on the pre-selected numbers to facilitate data storage and retrieval.

[0058] S104: When the outgoing call data of the pre-selected number matches the software usage data corresponding to the user using the food delivery software, the pre-selected number is determined to be a food delivery number.

[0059] In some embodiments, the delivery person may also dial the wrong number, that is, the time when the takeout number is dialed does not match the time when the user places the order. Therefore, it is not only necessary to analyze the characteristics of the takeout number, but also to further confirm the takeout number in combination with the software usage data when the user uses the takeout software, that is, only the corresponding pre-selected number when the user effectively uses the takeout software is the takeout number. Based on this, the embodiment of the present disclosure also needs to verify whether the user's behavior of using the takeout software is valid.

[0060] In actual applications, the complete process of a takeaway order includes the user opening the takeaway software, using the takeaway software, the delivery person delivering the food and receiving the order, the delivery person calling out the pickup number / text message, and the user receiving the takeaway, which means the takeaway order is complete. Here, the user opening the takeaway software and using the takeaway software are combined with the takeaway number / text message calling action, and one by one confirms whether the pre-selected numbers in the pre-selected number set V are takeaway numbers. It should be understood that the user's behavior of using the takeaway software here may include the user's behavior of opening the takeaway software and the behavior of using the takeaway software. Here, the behavior of using the takeaway software is the user's behavior of placing an order on the takeaway software. The judgment process is based on the following assumptions:

[0061] Assumption 1: To select the right fast food and complete the ordering process, if the user opens the food delivery app for more than 0.5 minutes, the user's use of the food delivery app is deemed valid; if the user opens the food delivery app for less than 0.5 minutes, the user's use of the food delivery app is deemed invalid. It should be understood that the 0.5 hour here can be adjusted according to actual conditions.

[0062] Assumption 2: When the user's use of the food delivery app is effective, the user will receive at least one incoming call from the food delivery number / text message after placing an order, and the time interval between food delivery to the user is within 1.2 hours. It should be understood that here it is assumed that a valid food delivery order is generated when the user opens the food delivery app, and the 1.2 hours can be adjusted according to actual conditions.

[0063] For ease of understanding, taking one month as an example, the present disclosure provides a valid takeaway number indicator symbol and meaning description as a reference, see Table 3 for details:

[0064] Table 3: Valid takeaway number indicator symbols and meanings

[0065]

[0066]

[0067] Among them, X1 represents the time when the user first opens the food delivery app in the month; X n Indicates the time when the user opens the food delivery app for the nth time in the month; X Nrepresents the time when the user opens the food delivery app for the Nth time in the month; Y1 represents the duration of the user opening the food delivery app for the first time in the month; Y n Indicates the duration of the user opening the food delivery app for the nth time in the month; Y N Indicates the duration of the user opening the food delivery app for the Nth time in the month; XX1 indicates the time when the user effectively opens the food delivery app for the first time in the month; XX n Indicates the time when the user effectively opens the food delivery app for the nth time in the month; XX N Indicates the time when the user effectively opens the food delivery app for the Nth time in the month; Z iA Indicates the time when the i-th pre-selected number in V calls the number / text message of user A; wherein Vi represents the i-th pre-selected number, i represents the sequence number of the pre-selected number, i=1,2…

[0068] Based on this, we can confirm that the pre-selected number is a takeaway number by following the following three steps based on the above assumptions. The specific steps are as follows:

[0069] Step 1: Identify the behavior of user A opening the food delivery software.

[0070] Specifically, if user A opens the food delivery app for the nth time, the duration is Y n If it is greater than 0.5 minutes, the user's use of the food delivery app is deemed valid. At this time, the corresponding user's opening time for the food delivery app needs to be counted into {XX n It should be understood that the duration of opening the food delivery software is the usage time of the food delivery software, and the time threshold of 0.5 minutes can be set according to actual conditions and is not limited here.

[0071] Step 2: If the user's use of the food delivery software is valid, determine the valid outgoing call from multiple outgoing calls corresponding to the pre-selected number according to the software usage data corresponding to the user. It should be understood that the software usage data includes the software opening time.

[0072] Specifically, we first need to traverse and search the software opening time set corresponding to the user. If there is a software opening time X in the set, n Make 0 <Z iA -X n ≤1.2h, that is, when the difference between the outgoing call time of a pre-selected number and the software opening time is greater than 0 and less than 1.2h, confirm Z iA The corresponding outgoing call is a valid outgoing call. Here, a valid outgoing call is a call corresponding to the takeaway number delivering this order when the use of the takeaway software is valid. At the same time, the specific time threshold can be set according to the situation and is not limited here.

[0073] Step 3: When the ratio of the number of valid outgoing calls corresponding to the pre-selected number to the total number of outgoing calls is greater than a preset threshold, the pre-selected number is determined to be the target number.

[0074] Specifically, the corresponding valid outgoing call times during the peak ordering period can be traversed, and if the ratio of the valid outgoing call times of the valid takeaway number to the total outgoing call times is ≥ 80%, the pre-selected number is confirmed to be a takeaway number. It should be understood that the setting of the specific threshold can be adjusted according to the actual situation.

[0075] In some embodiments, the complete process of a takeaway order on the user side may include the steps of opening the takeaway software, browsing products and placing an order, receiving an incoming call from a takeaway number, and obtaining the takeaway. Here, the user's behavior of opening the takeaway software can be combined with the takeaway number / SMS access behavior to determine whether the user's behavior of using the takeaway software is a valid behavior. It should be understood that the user's behavior of using the takeaway software here may include the user's behavior of opening the takeaway software and the behavior of using the takeaway software, and the user's behavior of using the takeaway software is valid, which requires that both the user's behavior of opening the takeaway software and the user's behavior of using the takeaway software are valid. For ease of understanding, taking one month as an example, the present disclosure embodiment provides a user behavior indicator and a description of its meaning as a reference, see Table 4 for details:

[0076] Table 4 User behavior indicators and their meanings

[0077]

[0078] Among them, A1 represents the time when the user opens the food delivery app for the first time in the month; A m The time when the user opens the food delivery app for the mth time in the month; A M The time when the user opens the food delivery app for the Mth time in the month; B1 represents the duration of the user opening the food delivery app for the first time in the month; B m Indicates the duration of the user opening the food delivery app for the mth time in the month; B M Indicates the duration of opening the food delivery app for the Mth time in the month; AA1 indicates the time when the user effectively opens the food delivery app for the first time in the month; AA M Indicates the time when the user effectively opens the food delivery app for the mth time in the month; AA M represents the time when the user opens the food delivery app for the Mth time in the month; C1 represents the time when the user's food delivery number makes the first incoming call in the month; C k represents the time of the kth incoming call from the user’s takeaway number in the current month; C K Indicates the Kth incoming call time of the user's takeaway number in the current month.

[0079] Figure 2 A flowchart of a user behavior identification method provided by an exemplary embodiment of the present disclosure.

[0080] S201, obtaining the user's software usage data and user call data from the operator, and determining whether the user's use of the food delivery software is a valid behavior based on the software usage data.

[0081] Specifically, if the duration B of the user opening the food delivery app for the mth time m If the time is greater than 0.5 minutes, the use of the food delivery software is deemed valid. At this time, the corresponding user's opening time of the food delivery software needs to be counted into {AA m It should be understood that the duration of opening the food delivery software is the usage time of the food delivery software, and the time threshold of 0.5 minutes can be set according to actual conditions and is not limited here.

[0082] S202, if the use of the food delivery software is valid, determine whether the use of the food delivery software by the user is a valid behavior. It should be understood that the software use data includes the software opening time.

[0083] Specifically, for the mth time of effectively opening the food delivery software, the time AA m , traverse and search the takeaway number incoming call time set {C k}, if there is a takeaway number whose kth incoming call time is such that 0 <Z k -X n ≤1.2h, the user's use of the food delivery software is deemed valid, that is, a valid food delivery order is generated. Otherwise, the user's use of the food delivery software is deemed invalid. It should be understood that the determination of the food delivery number here can be determined by the food delivery number identification method provided in the embodiment of the present disclosure. At the same time, the specific time threshold can be set according to the situation and is not limited here.

[0084] S203: Determine whether the user is a heavy takeout user based on the number of takeout orders corresponding to the user.

[0085] Specifically, we can traverse the effective opening time set of the user's food delivery software {XX n}, and according to the identification method for the effectiveness of user behavior in S202, the total number of takeaway orders generated by the user in the month W is counted, and if W ≥ 10, the user is identified as a heavy user of takeaway. Among them, the value corresponding to the total number of heavy takeaway users can be adjusted according to actual conditions.

[0086] The takeaway number identification method and user behavior identification method provided in the embodiments of the present disclosure may be executed by a terminal or by a chip applied to the terminal.

[0087] Exemplarily, the above-mentioned terminals may include one or more of mobile phones, tablet computers, wearable devices, vehicle-mounted devices, laptop computers, ultra-mobile personal computers (UMPCs), netbooks, personal digital assistants (PDAs), and wearable devices based on augmented reality (AR) and / or virtual reality (VR) technology, and may also include but not be limited to remote control devices, wearable devices, street lamps, home appliances and other smart terminals, which are not specifically limited in the embodiments of the present disclosure.

[0088] Figure 3 The following is a flow chart of another method for identifying a takeout number provided by an exemplary embodiment of the present disclosure. Figure 3 As shown, specifically including:

[0089] S301, acquiring call data corresponding to the number to be identified and Internet access data for the target software, and acquiring software usage data when multiple users use the target software.

[0090] In some embodiments, the call data corresponding to the number to be identified can be directly obtained from the operator side. At the same time, since the number to be identified needs to access the network provided by the operator when using the software, the Internet access data of the number to be identified can also be obtained from the operator side. The call data corresponding to the number to be identified may include the number of incoming calls, the number of outgoing calls, the frequency of outgoing calls, the frequency of incoming calls, etc., and the Internet access data may include the usage time of the number to be identified and the target software, which may specifically include the opening time and the usage time.

[0091] The software usage data of users using the target software can also be obtained from the operator side. When users use the target software, they need to access the network provided by the operator and then open the target software. Therefore, the software usage data generated when users use the target software can be obtained on the operator side, which may include the software opening time and the software usage duration.

[0092] S302, input the call data and Internet access data into a pre-trained target classification model to obtain a classification result corresponding to the number to be identified, and when the classification result is a take-out number, determine that the number to be identified is a pre-selected number, wherein the target classification model is obtained by training a preset classification model through training samples, and the training samples include sample call data and sample Internet access data corresponding to different sample numbers.

[0093] In some embodiments, when identifying the number to be identified, the number to be identified can be analyzed using a pre-trained target classification model to determine whether the number to be identified is a takeaway number. That is, the call data and Internet access data corresponding to the number to be identified can be input into the pre-trained target classification model, and the call data and Internet access data corresponding to the number to be identified can be analyzed by the target classification model to determine whether the number to be identified is a takeaway number, and if the number to be identified is a takeaway number, the number to be identified is determined to be a pre-selected number.

[0094] For example, call data can be the number of outgoing calls, and Internet access data can be the duration of use of the target software. Generally, the number of outgoing calls from a takeaway number is higher than that of a normal user, and it takes a long time to log in to the target software. Therefore, it is possible to determine whether the number to be identified is a takeaway number based on the number of outgoing calls and the duration of use. Based on this, it is possible to determine whether the number to be identified is a takeaway number based on the characteristics of the number to be identified.

[0095] In some embodiments, the embodiments of the present disclosure may also construct a training sample, wherein the training sample may include sample call data and sample Internet access data corresponding to different sample numbers. Based on this, the training sample may be used to train a preset classification model to obtain a target classification model.

[0096] Specifically, the sample call data can include the number of incoming calls, the number of outgoing calls, the frequency of outgoing calls, the frequency of incoming calls, the average call duration, the number of contacted users, the number of outgoing calls during the dining time period, the ratio of outgoing calls during the dining time period, the number of silent / missed calls, and the corresponding data of the callback rate, and the sample online data can include the online time of the deliveryman when using the food delivery software. It should be understood that in actual applications, the sample call data and sample online data can also include other data.

[0097] In practical applications, specific data corresponding to sample call data and sample Internet access data can be obtained from the operator, and then training samples can be constructed based on sample call data and sample Internet access data corresponding to different takeaway numbers, and then a certain amount of sample data can be obtained from the training samples, and 70% of the data in the sample data can be used to construct a training set, and 30% of the data can be used to construct a test set. Based on this, the training set and the test set can be used to train the preset classification model, and under the constraints of accuracy, precision, and recall, the model parameters can be iteratively tuned to obtain the target classification model. Among them, the preset classification model can be an XGboost classification model. The embodiment of the present disclosure limits the initial values ​​corresponding to each parameter in the XGboost classification model, as shown in Table 2:

[0098] Table 2 Initial value comparison table

[0099]

[0100]

[0101] Among them, learning rate represents the learning rate; colsample_bytree represents the proportion of features that each decision tree will use; max_depth represents an important hyperparameter; it defines the maximum depth of the decision tree; subsample represents subsample, which is a subsampled sample extracted from the sample; objective represents the target parameter in the model learning task parameter, which specifies the target of the training task; n_estimators represents the number of trees; eval_metric represents the evaluation function, which has no effect on the training of the model, but evaluates the model effect after the model training is completed; early_stopping_rounds represents the early stopping strategy. It should be understood that in XGBoost, when objective = 'multi:softprob', the model will output a probability distribution for each instance of the multi-classification problem. AUC (Area Under Curve) is used to measure the ranking ability. The AUC value is the area under the curve, which is a probability value. The larger the better. mlogloss represents the loss function in the multi-classification problem, also known as the cross entropy loss function, which measures the difference between the probability distribution predicted by the model and the probability distribution of the true label.

[0102] S303, filtering out the number of valid outgoing calls from the call data corresponding to the preselected number, wherein the valid outgoing calls are outgoing calls under the preset scenario, and the outgoing call time corresponding to the valid outgoing calls and the software opening time in the software usage data meet the first preset condition. It should be understood that the preset scenario can be a food delivery scenario, in which the valid outgoing calls are the outgoing calls made by the food delivery man to the user's phone number on the food delivery order after receiving the food delivery order; the preset scenario can be an errand running scenario, in which the valid outgoing calls are the outgoing calls made by the errand runner to the user's phone number on the errand running order after receiving the errand running order.

[0103] In some embodiments, the call data corresponding to the preselected number includes not only the call data corresponding to the takeaway order, but also the call data corresponding to other calls. Moreover, when the preselected number is a takeaway number, one preselected number may correspond to multiple takeaway orders. Therefore, in order to analyze the preselected number more accurately, it is also necessary to determine the number of valid outgoing calls for the takeaway order in the preselected number based on the call data, and then the preselected number can be further determined based on the number of valid outgoing calls of the preselected number.

[0104] Specifically, the call time corresponding to each call included in the call data and the software opening time corresponding to the user opening the target software can be analyzed respectively, and then when the call time corresponding to a certain call and the software opening time in the software usage data meet the first preset condition, the call is determined to be a valid call.

[0105] In actual applications, takeout orders are generally placed by users on the target software, and then the delivery man receives the order and delivers the takeout. Therefore, the software opening time when the user generally opens the target software should be earlier than the outgoing call time of the takeout call. Based on this, when the outgoing call time corresponding to a certain call is later than the software opening time, it is considered that the outgoing call time corresponding to the call and the software opening time in the software usage data meet the first preset condition. At this time, the call is determined to be a valid call.

[0106] S304: When the number of valid outgoing calls meets the second preset condition, determine that the pre-selected number is a take-out number. Determine that the pre-selected number is a take-out number.

[0107] In some embodiments, for food delivery software, when the pre-selected number is a food delivery number, the corresponding valid call out times will usually be larger. Therefore, a threshold for the number can be set, that is, when the valid call out times corresponding to the pre-selected number is greater than a certain threshold, the pre-selected number is determined to be a food delivery number.

[0108] Specifically, the threshold can be set to 500 orders per month, that is, when the number of orders received by the pre-selected number is greater than or equal to 500 orders in a month without holidays, the pre-selected number is determined to be a takeaway number; the threshold can be the average number of orders received per day in a month, specifically 18 orders, that is, when the number of orders received by the pre-selected number is greater than or equal to 18 orders in a month without holidays, the pre-selected number is determined to be a takeaway number. It should be understood that this is only an example, and the specific duration and the corresponding number of orders can be adjusted according to the situation.

[0109] In real life, deliverymen usually make more calls during meal times. Therefore, a target time period may be set. When the number of valid calls during the target time period for several consecutive days is greater than a certain threshold, the pre-selected number may be determined to be a delivery number.

[0110] For example, during lunch time, which is generally from 11:00 to 13:00, when the number of orders received by the deliveryman is greater than or equal to 10 during this time period, the pre-selected number is determined to be the delivery phone number.

[0111] Based on this, the disclosed embodiment not only starts from the perspective of the number to be identified, obtains the call data corresponding to the number to be identified and the Internet access data for the target software, but also starts from the user side, combines the corresponding software usage data when the user uses the target software, and analyzes the data of different dimensions, so that the number to be identified can be analyzed more accurately to determine whether the number to be identified is a takeaway number. Specifically, firstly, a pre-trained target classification model is used to directly analyze the data corresponding to the number to be identified, and then a preliminary judgment is made on the number to be identified. When the target classification model determines that the number to be identified is a takeaway number, the software usage data corresponding to the user when using the target software is combined to further analyze the number to be identified, so as to determine whether the number to be identified is a takeaway number. It can be seen that the disclosed embodiment can cross-check the Internet access data and call data corresponding to the number to be identified with the software usage data corresponding to the user, thereby improving the accuracy of takeaway number recognition.

[0112] In some embodiments, when the software usage data includes the software usage time, for any user, when the software usage time corresponding to the user is greater than or equal to the preset software usage time, the user's behavior of using the target software is determined to be a valid behavior; when the user's behavior of using the target software is a valid behavior, the number of valid outgoing calls is filtered out from the call data corresponding to the pre-selected number.

[0113] In actual applications, users may accidentally touch the target software, that is, they may accidentally open the target software. In the case of an accidental touch, the user will directly close the target software after it is opened and will not use it. Since the target software is generally closed directly after an accidental touch, a preset software usage time can be set to determine whether the user's use of the target software is effective.

[0114] For example, the preset software usage time can be half a minute, that is, only when the user opens the target software and the software usage time of the target software is longer than half a minute, will the user's use of the target software be determined as valid behavior, and when the user's use of the target software is determined to be valid behavior, the number of valid outgoing calls is further screened from the call data corresponding to the pre-selected number. Based on this, analysis can be performed from the user side to further improve the recognition accuracy of the pre-selected number.

[0115] In some embodiments, the time difference between each outgoing call and the software opening time can be determined respectively according to the outgoing call time and the software opening time corresponding to each outgoing call in the call data; for any outgoing call, when the time difference corresponding to the outgoing call is greater than the first preset threshold and less than or equal to the second preset threshold, it is determined that the outgoing call time corresponding to the effective outgoing call and the software opening time in the software usage data meet the first preset condition.

[0116] Specifically, since there may be multiple outgoing calls corresponding to the number to be identified, when the user's behavior of using the target software is valid, the outgoing call times corresponding to the multiple outgoing calls of the pre-selected number can be matched one by one with the software opening time corresponding to the user. The specific matching process can be, first, obtaining the time difference between the multiple outgoing calls and the software opening time respectively, when the time difference is greater than 0 and less than or equal to 1.2h, at this time, it can be determined that the outgoing call time corresponding to the valid outgoing call and the software opening time in the software usage data meet the first preset condition, and the outgoing call corresponding to the time difference is determined to be a valid outgoing call.

[0117] In some embodiments, when there are many users, the call times corresponding to multiple calls of the pre-selected number can be matched one by one with the software opening time corresponding to different users to determine the number of valid calls corresponding to the pre-selected number, and then the pre-selected number can be determined to be a takeout number when the ratio of the number to the total number of calls is greater than a preset threshold. Based on this, the number of valid calls corresponding to the pre-selected number can be determined in combination with the software usage data corresponding to the user, and the pre-selected number can be further analyzed based on the valid number of calls, thereby improving the recognition accuracy of the pre-selected number.

[0118] In some embodiments, for any outgoing call, when the software opening time is earlier than the outgoing call time corresponding to the outgoing call, it is determined that the outgoing call time corresponding to the effective outgoing call and the software opening time in the software usage data meet the first preset condition.

[0119] In actual applications, a takeaway order is usually placed by a user in the target software, and then a takeaway delivery person receives the order and delivers the takeaway. Therefore, the software opening time when the user opens the target software should be earlier than the outgoing call time of the takeaway call. Based on this, when the outgoing call time corresponding to a certain outgoing call is later than the software opening time, the outgoing call can be determined to be a valid outgoing call.

[0120] Figure 4 FIG. 1 is a flow chart of another method for identifying user behavior provided by an exemplary embodiment of the present disclosure. Figure 4 As shown, specifically including:

[0121] S401, obtaining the user's call data and software usage data when using the target software, and determining the incoming takeout number and the corresponding incoming call data from the call data, wherein the takeout number is determined by the takeout number identification method provided by the embodiment of the present disclosure.

[0122] In some embodiments, the user's call data and software usage data when using the target software can be directly obtained from the operator side. There will be multiple incoming calls in the call data corresponding to the user. Therefore, it is also possible to determine the incoming calls to the takeaway number among the multiple incoming calls in combination with the takeaway number identification method provided in the embodiment of the present disclosure.

[0123] S402, matching the incoming call data with the software usage data, and if the incoming call data and the software usage data are matched successfully, determining that the user's behavior of using the target software is a valid behavior.

[0124] In some embodiments, when the target software is a food delivery software, when the user uses the food delivery software, corresponding software usage data will be generated, and when there is incoming call data matching the software usage data in the user's corresponding communication data, the user's use of the food delivery software is considered to be a valid behavior, that is, it can be considered that the user has generated a valid food delivery order by using the food delivery software. It should be understood that the software opening time corresponding to the valid order is earlier than the incoming call time of the food delivery number.

[0125] Specifically, the incoming call data may be the incoming call time, and the software usage data may be the software opening time. Based on this, since the incoming calls corresponding to the takeaway number have been screened out in S401, it is only necessary to match the software opening time corresponding to the user with the incoming call time corresponding to the takeaway number. When the software opening time corresponding to the user successfully matches the incoming call time corresponding to the takeaway number, the user's behavior of using the target software is considered to be a valid behavior. It can be seen that the embodiment of the present disclosure can first determine the takeaway number, and then match the incoming call data of the takeaway number with the software usage data of the user using the target software, thereby more accurately analyzing the user's behavior of using the target software.

[0126] In some embodiments, the software usage data includes software usage time, and the method further includes: when the user's software usage time is greater than or equal to the preset software usage time, determining that the user's behavior of opening the target software is a valid behavior; when the user's behavior of opening the target software is a valid behavior and the incoming call time matches the software usage data, determining that the user's behavior of using the target software is a valid behavior.

[0127] In actual applications, users may accidentally touch the target software, that is, they may accidentally open the target software. In the case of an accidental touch, the user will directly close the target software after it is opened and will not use it. Since the target software is generally closed directly after an accidental touch occurs, a preset software usage time can be set to determine whether the user's behavior of using the target software is effective.

[0128] For example, the preset software usage time can be half a minute, that is, only after the user opens the target software, when the software usage time of the target software is greater than half a minute, will the user's behavior of opening the target software be determined to be a valid behavior, and when it is determined that a user's behavior of opening the target software is a valid behavior, the incoming call data and the software usage data are further analyzed, that is, when the user's behavior of opening the target software is a valid behavior and the incoming call time matches the software usage data, the user's behavior of using the target software is determined to be a valid behavior.

[0129] In actual applications, even if the user opens the target software for more than the preset software usage time, it may be invalid use, that is, the user may accidentally touch the control of the target software, causing the target software to open and then go to handle other things, causing the target software to remain open. At this time, although the software usage time of the target software is greater than the preset software usage time, the user did not perform other operations, resulting in invalid use. Therefore, it is necessary to further analyze the incoming call data and software usage data to more accurately analyze the user's behavior in using the target software.

[0130] In some embodiments, the software usage data includes multiple software opening times, and the incoming call data includes incoming call times. The method also includes: in the case where there are multiple incoming takeout numbers, obtaining the incoming call times corresponding to the multiple takeout numbers from the call data; respectively determining the time differences between the multiple incoming call times and the multiple software opening times, and determining the target incoming call time from the multiple incoming call times based on the time differences; determining the number of target incoming call times in the call data, and when the number is greater than or equal to a preset threshold, determining the user as a heavy user of the target software.

[0131] In actual applications, for food delivery software, a user may generate multiple food delivery orders at different times, that is, corresponding to multiple food delivery numbers, and the deliveryman may call the wrong number due to the large number of outgoing calls. Therefore, after determining the deliveryman's number in the user's call data, it is necessary to further analyze the incoming call time and the software opening time to determine the target incoming call corresponding to the user. It should be understood that the target incoming call is the incoming call received by the user after the deliveryman initiates a call to the corresponding user for the food delivery order.

[0132] Specifically, the time difference between multiple incoming call times and multiple software opening times can be determined respectively, and when the time difference is greater than the third preset threshold and less than or equal to the fourth preset threshold, the incoming call time is determined to be the target incoming call time. Since the incoming call corresponding to each target incoming call time is an input of a takeaway number, each target incoming call time corresponds to a valid takeaway order. At this time, the number of target incoming call times in the call data can be directly obtained, and when the number is greater than or equal to the preset threshold, the user is determined to be a heavy user of the takeaway software.

[0133] The above mainly introduces the solution provided by the embodiment of the present disclosure from the perspective of the server. It is understandable that, in order to implement the above functions, the server includes hardware structures and / or software modules corresponding to the execution of each function. Those skilled in the art should easily realize that, in combination with the units and algorithm steps of each example described in the embodiments disclosed herein, the present disclosure can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present disclosure.

[0134] The embodiments of the present disclosure may divide the server into functional units according to the above method examples. For example, each functional module may be divided corresponding to each function, or two or more functions may be integrated into one management module. The above integrated modules may be implemented in the form of hardware or in the form of software functional modules. It should be noted that the division of modules in the embodiments of the present disclosure is schematic and is only a logical functional division. There may be other division methods in actual implementation.

[0135] In the case of dividing each functional module according to each function, an exemplary embodiment of the present disclosure provides a takeout number recognition device, which may be a server or a chip applied to a server. Figure 5 The structure diagram of a takeaway number recognition device provided by an exemplary embodiment of the present disclosure is as follows. Figure 5 As shown, the takeaway number recognition device 500 includes:

[0136] The acquisition module 501 is used to acquire the call data corresponding to the number to be identified and the Internet access data for the target software, and acquire the software usage data when multiple users use the target software;

[0137] The model processing module 502 is used to input the call data and the Internet access data into a pre-trained target classification model to obtain a classification result corresponding to the number to be identified, and determine that the number to be identified is a pre-selected number when the classification result is a take-out number, wherein the target classification model is obtained by training a preset classification model through training samples, and the training samples include sample call data and sample Internet access data corresponding to different sample numbers;

[0138] A screening module 503 is used to screen out the number of valid outgoing calls from the call data corresponding to the preselected number, wherein the valid outgoing calls are outgoing calls under a preset scenario, and the outgoing call time corresponding to the valid outgoing calls and the software opening time in the software usage data meet a first preset condition;

[0139] The determination module 504 is configured to determine that the preselected number is a take-out number when the number of valid outgoing calls meets a second preset condition.

[0140] In an optional manner, the software usage data includes software usage time, and the takeout number identification device 500 also includes an acquisition module 505, which is used to determine, for any user, that the user's behavior of using the target software is a valid behavior when the software usage time corresponding to the user is greater than or equal to the preset software usage time; and when the user's behavior of using the target software is a valid behavior, filter out the number of valid outgoing calls from the call data corresponding to the preselected number.

[0141] In an optional manner, the acquisition module 505 is also used to determine the time difference between each outgoing call and the software opening time according to the outgoing call time corresponding to each outgoing call in the call data and the software opening time; for any outgoing call, when the time difference corresponding to the outgoing call is greater than a first preset threshold and less than or equal to a second preset threshold, it is determined that the outgoing call time corresponding to the valid outgoing call and the software opening time in the software usage data meet the first preset condition.

[0142] In an optional manner, the model processing module 502 is also used to determine, for any outgoing call, whether the call time corresponding to the valid call and the software opening time in the software usage data meet a first preset condition when the software opening time is earlier than the call time corresponding to the outgoing call.

[0143] In the case of dividing each functional module according to each function, an exemplary embodiment of the present disclosure provides a user behavior recognition device, which may be a server or a chip applied to a server. Figure 6 The following is a schematic diagram of a user behavior recognition device provided by an exemplary embodiment of the present disclosure. Figure 6 As shown, the user behavior recognition device 600 includes:

[0144] The acquisition module 601 is used to acquire the user's call data and the software usage data when using the target software, and determine the incoming takeout number and the corresponding incoming call data from the call data, wherein the takeout number is determined by the takeout number identification method of the embodiment of the present disclosure;

[0145] The determination module 602 is used to match the incoming call data with the software usage data, and if the incoming call data and the software usage data are matched successfully, determine that the user's behavior of using the target software is a valid behavior.

[0146] In an optional manner, the software usage data includes software usage duration, and the user behavior identification device 600 also includes a judgment module 603, wherein the judgment module 603 is used to determine that the user's behavior of opening the target software is a valid behavior when the user's software usage duration is greater than or equal to a preset software usage duration; and to determine that the user's behavior of opening the target software is a valid behavior and the incoming call data matches the software usage data.

[0147] In an optional manner, the software usage data includes multiple software opening times, and the incoming call data includes incoming call times. The judgment module 603 is also used to obtain the incoming call times corresponding to multiple takeout numbers from the call data when there are multiple incoming takeout numbers; respectively determine the time differences between the multiple incoming call times and the multiple software opening times, and determine the target incoming call time from the multiple incoming call times based on the time differences; determine the number of target incoming call times in the call data, and determine that the user is a heavy user of the target software when the number is greater than or equal to a preset threshold.

[0148] In an optional manner, the judgment module 603 is further configured to determine that the incoming call time is the target incoming call time when the time difference is greater than a third preset threshold and less than or equal to a fourth preset threshold.

[0149] An embodiment of the present disclosure also provides an electronic device, comprising: at least one processor; a memory for storing instructions executable by at least one processor; wherein the at least one processor is used to execute instructions to implement the steps of the above method disclosed in the embodiment of the present disclosure.

[0150] Figure 7 The structure diagram of an electronic device provided by an exemplary embodiment of the present disclosure is shown in FIG. Figure 7As shown, the electronic device 700 includes at least one processor 701 and a memory 702 coupled to the processor 701. The processor 701 can execute corresponding steps in the above method disclosed in the embodiment of the present disclosure.

[0151] The processor 701 may also be referred to as a central processing unit (CPU), which may be an integrated circuit chip having signal processing capabilities. Each step in the method disclosed in the embodiment of the present disclosure may be completed by an integrated logic circuit of hardware in the processor 701 or by instructions in the form of software. The processor 701 may be a general-purpose processor, a digital signal processor (DSP), an ASIC, a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. A general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in the embodiment of the present disclosure may be directly embodied as being executed by a hardware decoding processor, or may be executed by a combination of hardware and software modules in a decoding processor. The software module may be located in a memory 702, such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, or other mature storage media in the art. The processor 701 reads the information in the memory 702 and completes the steps of the method in combination with its hardware.

[0152] In addition, when various operations / processes according to the present disclosure are implemented by software and / or firmware, they can be transmitted from a storage medium or a network to a computer system having a dedicated hardware structure, for example, Figure 8 The computer system 800 shown installs the programs constituting the software, and when the various programs are installed, the computer system can execute various functions, including the functions described above. Figure 8 A schematic diagram of the structure of a computer system provided by an exemplary embodiment of the present disclosure.

[0153] Computer system 800 is intended to represent various forms of digital electronic computer devices, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic devices can also represent various forms of mobile devices, such as personal digital processing, cellular numbers, smart numbers, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0154] like Figure 8 As shown, the computer system 800 includes a computing unit 801, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 802 or a computer program loaded from a storage unit 808 into a random access memory (RAM) 803. In the RAM 803, various programs and data required for the operation of the computer system 800 can also be stored. The computing unit 801, the ROM 802, and the RAM 803 are connected to each other via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0155] A plurality of components in the computer system 800 are connected to the I / O interface 805, including: an input unit 806, an output unit 807, a storage unit 808, and a communication unit 809. The input unit 806 can be any type of device capable of inputting information to the computer system 800, and the input unit 806 can receive input digital or character information, and generate key signal inputs related to user settings and / or function control of the electronic device. The output unit 807 can be any type of device capable of presenting information, and can include but is not limited to a display, a speaker, a video / audio output terminal, a vibrator, and / or a printer. The storage unit 808 can include but is not limited to a disk, an optical disk. The communication unit 809 allows the computer system 800 to exchange information / data with other devices over a network such as the Internet, and can include but is not limited to a modem, a network card, an infrared communication device, a wireless communication transceiver, and / or a chipset, for example, a Bluetooth™ device, a WiFi device, a WiMax device, a cellular communication device, and / or the like.

[0156] The computing unit 801 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 801 performs the various methods and processes described above. For example, in some embodiments, the above methods disclosed in the embodiments of the present disclosure may be implemented as a computer software program, which is tangibly included in a machine-readable medium, such as a storage unit 808. In some embodiments, part or all of the computer program may be loaded and / or installed on an electronic device via ROM 802 and / or a communication unit 809. In some embodiments, the computing unit 801 may be configured to perform the above methods disclosed in the embodiments of the present disclosure by any other appropriate means (e.g., by means of firmware).

[0157] The embodiment of the present disclosure also provides a computer-readable storage medium, wherein when the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the above method disclosed in the embodiment of the present disclosure.

[0158] The computer-readable storage medium in the disclosed embodiments may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. The computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. More specifically, the computer-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0159] The computer-readable medium may be included in the electronic device, or may exist independently without being incorporated into the electronic device.

[0160] Fig. 9 Schematic diagram of a computer program product provided by an exemplary embodiment of the present disclosure. Fig. 9 As shown, the computer program product 900 includes a computer program 901, wherein the computer program 901 implements the above method disclosed in the embodiment of the present disclosure when executed by a processor.

[0161] In embodiments of the present disclosure, computer program codes for performing the operations of the present disclosure may be written in one or more programming languages ​​or combinations thereof, including but not limited to object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" language or similar programming languages. The program code may be executed entirely on a user's computer, partially on a user's computer, as an independent software package, partially on a user's computer, partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer.

[0162] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0163] The modules, components or units involved in the embodiments described in the present disclosure may be implemented by software or hardware, wherein the names of the modules, components or units do not, in some cases, limit the modules, components or units themselves.

[0164] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chip (SOCs), complex programmable logic devices (CPLDs), and the like.

[0165] The above descriptions are only some embodiments of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present disclosure is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, a technical solution formed by replacing the above features with the technical features with similar functions disclosed in the present disclosure (but not limited to).

[0166] Although some specific embodiments of the present disclosure have been described in detail by way of example, it should be understood by those skilled in the art that the above examples are for illustration only and are not intended to limit the scope of the present disclosure. It should be understood by those skilled in the art that the above embodiments may be modified without departing from the scope and spirit of the present disclosure. The scope of the present disclosure is defined by the appended claims.

Claims

1. A takeaway number recognition method, characterized in that: include: Acquire call data corresponding to the number to be identified and Internet access data for the target software, and acquire software usage data when the user uses the target software; Input the call data and the Internet access data into a pre-trained target classification model to obtain a classification result corresponding to the number to be identified, and determine that the number to be identified is a pre-selected number when the classification result is a take-out number, wherein the target classification model is obtained by training a preset classification model through training samples, and the training samples include sample call data and sample Internet access data corresponding to different sample numbers; Filtering the number of valid outgoing calls from the call data corresponding to the preselected number, wherein the valid outgoing calls are outgoing calls under a preset scenario, and the outgoing call time corresponding to the valid outgoing calls and the software opening time in the software usage data meet a first preset condition; When the number of valid outgoing calls meets a second preset condition, the preselected number is determined to be a take-out number.

2. The method according to claim 1, characterized in that The software usage data includes the software usage time, and the method further includes: For any user, if the software usage time corresponding to the user is greater than or equal to the preset software usage time, determining that the user's behavior of using the target software is a valid behavior; In the case that the user's behavior of using the target software is valid, the number of valid outgoing calls is screened out from the call data corresponding to the preselected number.

3. The method according to claim 1, characterized in that The method further comprises: According to the outgoing call time corresponding to each outgoing call in the call data and the software opening time, respectively determine the time difference between each outgoing call and the software opening time; For any outgoing call, when the time difference corresponding to the outgoing call is greater than a first preset threshold and less than or equal to a second preset threshold, it is determined that the outgoing call time corresponding to the valid outgoing call and the software opening time in the software usage data meet the first preset condition.

4. The method according to claim 1, characterized in that: The method further comprises: For any outgoing call, when the software opening time is earlier than the outgoing call time corresponding to the outgoing call, it is determined that the outgoing call time corresponding to the valid outgoing call and the software opening time in the software usage data meet a first preset condition.

5. A user behavior identification method, characterized in that: include: Acquire the user's call data and software usage data when using the target software, and determine the incoming takeout number and the corresponding incoming call data from the call data, wherein the takeout number is determined by the takeout number identification method according to any one of claims 1 to 4; The incoming call data and the software usage data are matched, and when the incoming call data and the software usage data are successfully matched, it is determined that the user's behavior of using the target software is a valid behavior.

6. The method according to claim 5, characterized in that The software usage data includes the software usage time, and the method further includes: When the software usage time of the user is greater than or equal to the preset software usage time, determining that the user's behavior of opening the target software is a valid behavior; When the user's behavior of opening the target software is a valid behavior and the incoming call data matches the software usage data, it is determined that the user's behavior of using the target software is a valid behavior.

7. The method according to claim 5, characterized in that The software usage data includes multiple software opening times, the incoming call data includes incoming call times, and the method includes: In the case where there are multiple incoming takeout numbers, obtaining the incoming call times corresponding to the multiple takeout numbers from the call data; respectively determining the time differences between the plurality of incoming call times and the plurality of software opening times, and determining a target incoming call time from the plurality of incoming call times according to the time differences; The number of target incoming call times in the call data is determined, and when the number is greater than or equal to a preset threshold, the user is determined to be a heavy user of the target software.

8. The method according to claim 7, characterized in that Determining a plurality of target incoming call times from the plurality of incoming call times according to the time difference comprises: When the time difference is greater than the third preset threshold and less than or equal to the fourth preset threshold, the incoming call time is determined as the target incoming call time.

9. A takeaway number recognition device, characterized in that: include: An acquisition module, used to acquire call data corresponding to the number to be identified and Internet access data for the target software, and to acquire software usage data when multiple users use the target software; a model processing module, configured to input the call data and the online data into a pre-trained target classification model, obtain a classification result corresponding to the number to be identified, and determine that the number to be identified is a pre-selected number when the classification result is a take-out number, wherein the target classification model is obtained by training a preset classification model through training samples, and the training samples include sample call data and sample online data corresponding to different sample numbers; A screening module, screening the number of valid outgoing calls from the call data corresponding to the preselected number, wherein the valid outgoing calls are calls made under a preset scenario, and the outgoing call time corresponding to the valid outgoing calls and the software opening time in the software usage data meet a first preset condition; The determination module is used to determine that the pre-selected number is a take-out number when the number of valid outgoing calls meets a second preset condition.

10. A user behavior recognition device, characterized in that: include: an acquisition module, used to acquire the user's call data and software usage data when using the target software, and determine the incoming takeout number and the corresponding incoming call data from the call data, wherein the takeout number is determined by the takeout number identification method according to any one of claims 1 to 4; The determination module is used to match the incoming call data with the software usage data, and when the incoming call data and the software usage data are matched successfully, determine that the user's behavior of using the target software is a valid behavior.

11. A computer device comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the method according to any one of claims 1 to 8.

12. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: When the computer program / instructions are executed by a processor, the method according to any one of claims 1 to 8 is implemented.

13. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the method according to any one of claims 1 to 8 is implemented.

Images