User behavior analysis method, device, electronic device and storage medium

Abstract

The embodiment of the present invention relates to the field of data processing technology, and discloses a user behavior analysis method, device, electronic device and storage medium, the method comprising: obtaining a user behavior sequence, the user behavior sequence comprising multiple user behaviors and the occurrence time of each user behavior; determining the number of first user behaviors in the user behavior sequence and the total number of at least one second user behavior according to the preset occurrence order of multiple user behaviors; if the number of first user behaviors is greater than or equal to the total number of at least one second user behavior, dividing the user behavior sequence into windows according to the third user behavior to obtain a first window sequence, the third user behavior is any one of the at least one second user behavior; according to each user behavior in the first window sequence, determining the conversion information corresponding to the user behavior sequence. Applying the technical solution of the present invention can improve the analysis efficiency of the user behavior sequence and save computer resources.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of data processing technology, and specifically to a user behavior analysis method, device, electronic device and storage medium. Background Art

[0002] With the development of Internet technology, many businesses have shifted from offline to online processing, which provides great convenience for users and also allows users' behavioral data to be preserved, allowing service providers to analyze users' personal preferences based on their behavioral data, thereby continuously optimizing business content and providing users with better services.

[0003] Among them, funnel analysis is a method for analyzing user conversion and loss in each step of a multi-step process. This method can help business providers quickly grasp the conversion status of products in each step over a period of time, so that they can locate the links and causes of user loss. Since this method has the advantages of quickly locating problems, multi-dimensional segmentation analysis, and monitoring funnel conversion trends, it has been widely used in user behavior analysis of websites and applications. However, as business content continues to enrich, business processes and conversion steps will continue to increase. How to improve the efficiency of analyzing user behavior using funnel analysis has become the current research focus. Summary of the invention

[0004] In view of the above problems, embodiments of the present invention provide a user behavior analysis method, device, electronic device and storage medium, which are used to solve the problem of low efficiency of user behavior analysis in the prior art.

[0005] According to one aspect of an embodiment of the present invention, a user behavior analysis method is provided, the method comprising: obtaining a user behavior sequence, the user behavior sequence comprising multiple user behaviors and the occurrence time of each of the user behaviors; determining the number of first user behaviors in the user behavior sequence and the total number of at least one second user behavior according to a preset occurrence order of the multiple user behaviors; wherein the second user behavior is any user behavior whose preset occurrence order is later than the preset occurrence order of the first user behavior; if the number of the first user behaviors is greater than or equal to the total number of at least one second user behavior, then dividing the user behavior sequence into windows according to a third user behavior to obtain a first window sequence, the third user behavior being any user behavior of the at least one second user behavior, the first window sequence comprising the first user behavior; determining conversion information corresponding to the user behavior sequence according to each user behavior in the first window sequence, the conversion information comprising at least one of a conversion depth and a conversion duration.

[0006] In an optional manner, the multiple user behaviors in the user behavior sequence are arranged from early to late according to the occurrence time of each user behavior, and the user behavior sequence is divided into windows according to the third user behavior to obtain a first window sequence, including: arranging the multiple user behaviors in the user behavior sequence from late to early according to the occurrence time of each user behavior to obtain a reverse sequence of the user behavior sequence; in the reverse sequence, taking each of the third user behaviors as the first user behavior in the first window sequence, dividing the reverse sequence according to a preset window length to obtain at least one first window sequence.

[0007] In an optional manner, the multiple user behaviors in the user behavior sequence are arranged from early to late according to the occurrence time of each user behavior, and the user behavior sequence is divided into windows according to the third user behavior to obtain a first window sequence, including: taking each of the third user behaviors in the user behavior sequence as the last user behavior in the first window sequence, dividing the user behavior sequence according to a preset window length, and obtaining at least one first window sequence.

[0008] In an optional manner, determining conversion information corresponding to the user behavior sequence based on each user behavior in the first window sequence includes: arranging the occurrence time of each user behavior in the first window sequence to obtain at least one time series; determining, based on the at least one time series, a target user behavior sequence whose occurrence time conforms to the preset occurrence order, wherein any two adjacent user behaviors in the target user behavior sequence are any two adjacent user behaviors in the preset occurrence order, and the target user behavior sequence includes the first user behavior; determining conversion information corresponding to the user behavior sequence based on the number of user behaviors included in the target user behavior sequence and / or the occurrence time of each user behavior.

[0009] In an optional manner, the conversion information corresponding to the user behavior sequence is determined based on the number of user behaviors included in the target user behavior sequence and / or the occurrence time of each user behavior, including: if the number of user behaviors included in the target user behavior sequence is equal to a preset conversion depth, then the conversion depth corresponding to the user behavior sequence is determined as the preset conversion depth, and the difference between the latest occurrence time and the earliest occurrence time in the target user behavior sequence is determined as the conversion duration corresponding to the user behavior sequence.

[0010] In an optional manner, determining conversion information corresponding to the user behavior sequence based on the number of user behaviors included in the target user behavior sequence and / or the occurrence time of each user behavior includes: if the number of user behaviors included in the target user behavior sequence is less than a preset conversion depth, dividing the user behavior sequence into windows based on a fourth user behavior to obtain a second window sequence, the fourth user behavior being a user behavior among the multiple user behaviors whose preset occurrence order is earlier than the preset occurrence order of the third user behavior, and the second window sequence including the first user behavior; and determining conversion information corresponding to the user behavior sequence based on each user behavior in the second window sequence.

[0011] In an optional manner, the method further includes: if the number of the first user behaviors is less than the total number of the at least one second user behavior, dividing the user behavior sequence into windows according to the first user behaviors to obtain a third window sequence; and determining conversion information corresponding to the user behavior sequence according to each user behavior in the third window sequence.

[0012] According to another aspect of an embodiment of the present invention, a user behavior analysis device is provided, the device comprising: an acquisition module, used to acquire a user behavior sequence, the user behavior sequence comprising multiple user behaviors and the occurrence time of each of the user behaviors; a first determination module, used to determine the number of first user behaviors in the user behavior sequence and the total number of at least one second user behavior according to a preset occurrence order of the multiple user behaviors; wherein the first user behavior is the user behavior with the earliest preset occurrence order among the multiple user behaviors, and the second user behavior is any user behavior whose preset occurrence order is later than the preset occurrence order of the first user behavior; a division module, used to divide the user behavior sequence into windows according to a third user behavior to obtain a first window sequence if the number of the first user behaviors is greater than or equal to the total number of at least one second user behavior, the third user behavior being any user behavior among the at least one second user behavior, the first window sequence comprising the first user behavior; a second determination module, used to determine conversion information corresponding to the user behavior sequence according to each user behavior in the first window sequence, the conversion information comprising at least one of conversion depth and conversion duration.

[0013] In an optional manner, the multiple user behaviors in the user behavior sequence are arranged from early to late according to the occurrence time of each user behavior, and the division module is used to arrange the multiple user behaviors in the user behavior sequence from late to early according to the occurrence time of each user behavior, so as to obtain an inverse sequence of the user behavior sequence, in which each of the third user behaviors is used as the first user behavior in the first window sequence, and the inverse sequence is divided according to a preset window length to obtain at least one first window sequence.

[0014] In an optional manner, the multiple user behaviors in the user behavior sequence are arranged from early to late according to the occurrence time of each user behavior, and the division module is used to divide the user behavior sequence according to a preset window length, with each of the third user behaviors in the user behavior sequence as the last user behavior in the first window sequence, to obtain at least one first window sequence.

[0015] In an optional manner, the second determination module is used to arrange the occurrence time of each user behavior in the first window sequence to obtain at least one time series, and determine a target user behavior sequence whose occurrence time conforms to the preset occurrence order based on the at least one time series, wherein any two adjacent user behaviors in the target user behavior sequence are any two adjacent user behaviors in the preset occurrence order, and the target user behavior sequence includes the first user behavior, and the conversion information corresponding to the user behavior sequence is determined based on the number of user behaviors included in the target user behavior sequence and / or the occurrence time of each user behavior.

[0016] In an optional manner, the second determination module is also used to determine the conversion depth corresponding to the user behavior sequence as the preset conversion depth if the number of user behaviors included in the target user behavior sequence is equal to the preset conversion depth, and determine the difference between the latest occurrence time and the earliest occurrence time in the target user behavior sequence as the conversion duration corresponding to the user behavior sequence.

[0017] In an optional manner, the second determination module is also used to divide the user behavior sequence into windows according to a fourth user behavior to obtain a second window sequence if the number of user behaviors included in the target user behavior sequence is less than a preset conversion depth, the fourth user behavior being a user behavior among the multiple user behaviors whose preset occurrence order is earlier than the preset occurrence order of the third user behavior, the second window sequence including the first user behavior, and determining conversion information corresponding to the user behavior sequence according to each user behavior in the second window sequence.

[0018] In an optional manner, the division module is also used to divide the user behavior sequence into windows according to the first user behavior to obtain a third window sequence if the number of the first user behaviors is less than the total number of the at least one second user behavior, and the second determination module is also used to determine the conversion information corresponding to the user behavior sequence according to each user behavior in the third window sequence.

[0019] According to another aspect of an embodiment of the present invention, an electronic device is provided, comprising: a processor; a memory for storing at least one executable instruction; the executable instruction enables the processor to perform the operation of the user behavior analysis method as described in any one of the above items.

[0020] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is provided, wherein at least one executable instruction is stored in the storage medium, and when the executable instruction is executed on an electronic device, the electronic device executes the operation of the user behavior analysis method as described in any one of the above items.

[0021] An embodiment of the present invention can determine the number of first user behaviors and the total number of at least one second user behavior in a user behavior sequence according to a preset occurrence order of multiple user behaviors. When the number of first user behaviors is greater than or equal to the total number of at least one second user behavior, the user behavior sequence is divided into windows according to the user behavior with the latest preset occurrence order among the multiple user behaviors, that is, the third user behavior, to obtain a first window sequence, where the first window sequence includes the first user behavior, and then according to each user behavior in the first window sequence, the conversion depth and conversion duration corresponding to the user behavior sequence are determined.

[0022] In this way, the window sequence can be divided according to the proportion of the first user behavior and the second user behavior in the user behavior sequence, and the third user behavior with a smaller number in the user behavior sequence can be used to reduce the amount of data for user behavior analysis, improve computing efficiency, and save computer resources. In addition, since in some cases the proportion of the first user behavior is very high, the application of this solution does not require window division for each first user behavior, so there will be no problem of data skew, which can improve computing stability.

[0023] The above description is only an overview of the technical solution of the embodiment of the present invention. In order to more clearly understand the technical means of the embodiment of the present invention, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the embodiment of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The accompanying drawings are only used to illustrate the embodiments and are not to be considered as limiting the present invention. In addition, the same reference symbols are used to represent the same components throughout the accompanying drawings. In the accompanying drawings:

[0025] Figure 1 A flow chart of a user behavior analysis method provided by the present invention is shown;

[0026] Figure 2 A schematic diagram of a user behavior sequence provided by the present invention is shown;

[0027] Figure 3 A flow chart of another user behavior analysis method provided by the present invention is shown;

[0028] Figure 4 A schematic diagram showing a reverse sequence of a user behavior sequence provided by the present invention;

[0029] Figure 5 A schematic diagram of window division provided by the present invention is shown;

[0030] Figure 6 A schematic diagram showing another window division provided by the present invention is shown;

[0031] Figure 7 A flowchart of another user behavior analysis method provided by the present invention is shown;

[0032] Figure 8 A schematic diagram showing another window division provided by the present invention is shown;

[0033] Fig. 9 A flowchart of another user behavior analysis method provided by the present invention is shown;

[0034] Fig.10 A schematic diagram of a two-dimensional array provided by the present invention is shown;

[0035] Fig.11 A flowchart of another user behavior analysis method provided by the present invention is shown;

[0036] Fig.12 A flowchart of another user behavior analysis method provided by the present invention is shown;

[0037] Fig.13 A schematic diagram showing another window division provided by the present invention is shown;

[0038] Fig.14 A flowchart of another user behavior analysis method provided by the present invention is shown;

[0039] Fig.15 A flowchart of another user behavior analysis method provided by the present invention is shown;

[0040] Fig.16 A schematic diagram of the structure of a user behavior analysis device provided by the present invention is shown;

[0041] Fig.17 A structural schematic diagram of an electronic device provided by the present invention is shown. DETAILED DESCRIPTION

[0042] Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0043] In order to determine the user conversion status in the business process, the user behavior sequence can be divided into windows according to the conversion order of each user behavior, starting from the first user behavior corresponding to the conversion order, to obtain multiple window sequences, and then find the user behavior that meets the conversion order in each window sequence to obtain the user's conversion information. However, since the steps for users to participate in the business usually start from the first user behavior, this method will greatly increase the number of window sequences obtained by dividing the user behavior sequence, and it will also take a very long time to calculate the conversion information of each window sequence to obtain the conversion information of the entire user behavior sequence, so the efficiency of user behavior analysis is low.

[0044] In view of one or more of the above problems, Figure 1 The flowchart of a user behavior analysis method provided by an embodiment of the present invention is shown. The method can be executed by an electronic device, and can obtain a user behavior sequence, and based on the analysis of the user behavior sequence, determine the user conversion information, so that the service provider can customize the design according to the user's personal preferences, making the user's operation path more convenient. The electronic device can be a background server or service cluster of the service provider. Figure 1 As shown, the method may include the following steps:

[0045] Step 110: Obtain user behavior sequence.

[0046] The user behavior sequence may include multiple user behaviors and the time when each user behavior occurs. The types and quantities of user behaviors may vary according to different business scenarios. For example, in e-commerce, user behaviors may include clicking on item advertisements, browsing details, placing orders, and so on. In vehicle sales, user behaviors may include clicking on advertisements, retaining user information, test driving, and placing orders, and so on.

[0047] In this embodiment, the user behavior sequence may be sequence data recording a user's behavior sequence. When the generated user behavior sequence includes behavior sequences of multiple users, the user behavior sequence of each user may be screened out from the user behavior sequence using the user identifier as an index.

[0048] User behavior sequences can be generated by electronic devices, such as servers or server clusters, by monitoring user operations on terminal devices, such as ad clicks, browsing details, joining memberships, placing orders, etc. At the same time, due to different business types, in some business scenarios, offline behavior information of users input by salespersons can also be obtained, and the behavior information can be combined with the behavior information of the same user obtained online to generate a user behavior sequence.

[0049] Taking the above vehicle sales as an example, the user's default operation process includes four processes: ad click, user information retention, user test drive and user order. Among them, the two processes of ad click and user information retention can be online behaviors of users through mobile phones, while the two processes of user test drive and user order are offline behaviors of users at the vehicle sales point. In this case, the data entered online and the data entered offline by the salesperson can together constitute the user's user behavior sequence.

[0050] Step 120: Determine the number of first user behaviors in the user behavior sequence and the total number of at least one second user behavior according to a preset occurrence order of the plurality of user behaviors.

[0051] The first user behavior is the user behavior with the earliest preset occurrence order among the multiple user behaviors, and the second user behavior is any user behavior with a preset occurrence order later than the preset occurrence order of the first user behavior.

[0052] The preset occurrence sequence is the sequence of user behaviors that are preset based on the user conversion sequence. Generally speaking, as the sequence of user behaviors goes from front to back, the user conversion degree corresponding to the user behaviors continues to increase, the user conversion depth continues to increase, and the user stickiness continues to increase. Taking the above-mentioned vehicle sales as an example, the preset occurrence sequence of multiple user behaviors can be ad clicks, user information retention, user test drives, and user orders, and the user conversion degree continues to increase.

[0053] The first user behavior is the user behavior with the earliest preset occurrence order among multiple user behaviors. For example, in the above-mentioned vehicle sales business scenario, the first user behavior may be the first user behavior "advertisement click" in the preset occurrence order; the second user behavior is any user behavior with a preset occurrence order later than the preset occurrence order of the first user behavior, such as the last user behavior "user placing an order" in the above-mentioned preset occurrence order, or the second to last user behavior "user test drive".

[0054] After obtaining the user behavior sequence of the user, the number of first user behaviors in the user behavior sequence and the total number of at least one second user behavior may be counted according to a preset occurrence order of the plurality of user behaviors.

[0055] Take the business scenario of vehicle sales as an example. Figure 2 A schematic diagram of a user behavior sequence provided by an embodiment of the present invention is shown, such as Figure 2 As shown in the figure, there are 18 user behaviors in chronological order, namely E1, E1, E2, ... E1, t1 to t 18 They represent the time nodes at which the corresponding user behaviors occur. It can be seen that the user behavior sequence includes four user behaviors, namely E1, E2, E3 and E4. Among them, E1 can represent "ad click", E2 can represent "user information retention", E3 can represent "user test drive", and E4 can represent "user order".

[0056] In this user behavior sequence, the number of first user behaviors E1 is equal to 6, and the second user behavior can be any one of E2, E3 and E4. Assuming that the second user behavior is E4, the total number of second user behaviors is 2.

[0057] In an optional manner, the second user behavior may be a user behavior in the second half of the preset occurrence sequence. Specifically, the user behavior in the second half may be determined according to the total number of user behaviors corresponding to the preset occurrence sequence. For example, for the user behaviors whose preset occurrence sequence is E1, E2, E3, and E4, there is an even number of user behaviors, and the second user behavior may be the second half thereof: any one of E3 and E4.

[0058] When there is an odd number of user behaviors corresponding to the preset occurrence order, the second half may include the remainder part. For example, for user behaviors whose preset occurrence order is X1, X2, X3, X4 and X5, the second user behavior may be any one of X3, X4 and X5.

[0059] In an optional manner, after obtaining the user behavior sequence, in order to determine the number and position of each user behavior, etc., the number and proportion of each type of user behavior, and the offset of one or more types of user behaviors may also be counted.

[0060] For example, in Figure 2 In the user behavior sequence shown, the total number of user behaviors is 18, of which there are 6 E1s, 6 E2s, 4 E3s, and 2 E4s, accounting for 33.3%, 33.3%, 22.2%, and 11.1% respectively; the ratio of the first user behavior, that is, the number of the first user behavior E1 in the preset occurrence sequence of multiple user behaviors, and the second half, that is, the total number of E3 and E4 is 1, and the offset positions of the user behaviors distributed in the second half are E4: (12, 17), E3: (7, 10, 11, 14).

[0061] By conducting statistical analysis, the basic distribution of the sequence can be quickly determined, which facilitates the subsequent positioning and analysis of various user behaviors.

[0062] Step 130: If the number of the first user behaviors is greater than or equal to the total number of at least one second user behavior, the user behavior sequence is divided into windows according to the third user behavior to obtain a first window sequence.

[0063] The third user behavior is any one of the at least one second user behavior, and the first window sequence includes the first user behavior. For example, the third user behavior can be any one of the multiple second user behaviors E3 and E4. When the at least one second user behavior includes only one user behavior, the third user behavior is the same as the second user behavior.

[0064] When the number of first user behaviors is greater than or equal to the total number of one or more second user behaviors, that is, the number of first user behaviors / the total number of one or more second user behaviors ≥ 1, it means that in the entire user behavior sequence, the number of first user behaviors is the largest, that is, most of the user's operations are the first user behaviors arranged in the preset occurrence order. At this time, the user behavior sequence can be divided into windows according to the third user behavior, that is, any user behavior whose preset occurrence order is later than the preset occurrence order of the first user behavior, to obtain the first window sequence.

[0065] Since the number of first user behaviors is greater than the total number of second user behaviors, and the third user behavior is any one of at least one second user behavior, the number of third user behaviors in the user behavior sequence must be smaller than the number of first user behaviors, and the number of window sequences divided according to the third user behavior will also be greatly reduced, thereby reducing the amount of data for subsequent calculation of user conversion conditions and improving the analysis efficiency of user behavior sequences.

[0066] Step 140: Determine conversion information corresponding to the user behavior sequence according to each user behavior in the first window sequence.

[0067] The conversion information may include at least one of conversion depth and conversion duration.

[0068] Exemplarily, based on each user behavior in the first window sequence, the type and time of the user behavior included in the first window sequence can be determined, thereby determining the conversion status of the user at each user behavior stage, and determining the conversion information corresponding to the user behavior sequence of the user. In addition, the user behavior in the first window sequence may include some user behaviors among multiple user behaviors, or may include all user behaviors among multiple user behaviors, which is not limited in the embodiments of the present application.

[0069] For example, taking the user behavior in the first window sequence including some user behaviors among multiple user behaviors, and the first window sequence is {E1: t1; E2: t2} as an example, according to the type and time sequence of the user behavior, the user conversion depth corresponding to the first window sequence is the number of user behaviors 2, and the conversion time is t2-t1.

[0070] In an optional manner, after determining the conversion information corresponding to the user behavior sequence, the user conversion rate can be further calculated, that is, the conversion depth of all users can be counted based on the calculated conversion depth of each user, and then the user conversion rate = conversion depth of all users / number of users can be calculated.

[0071] For example, assuming there are 10 users, among which: 5 users have a conversion depth of 2, 3 users have a conversion depth of 3, and 2 users have a conversion depth of 4. Then, according to the preset order of occurrence of multiple user behaviors, the user conversion rates of the first user behavior and the second user behavior are both 100%, the user conversion rate of the third user behavior is 50%, and the user conversion rate of the fourth user behavior is 20%.

[0072] According to the user behavior analysis method provided by the present embodiment, the number of first user behaviors and the total number of at least one second user behavior in the user behavior sequence can be determined according to the preset occurrence order of multiple user behaviors, and the second user behavior is any user behavior whose preset occurrence order is later than the preset occurrence order of the first user behavior. When the number of first user behaviors is greater than or equal to the total number of at least one second user behavior, the user behavior sequence is divided into windows according to any one user behavior of the at least one second user behavior, that is, the third user behavior, to obtain a first window sequence, which includes the first user behavior, and then according to each user behavior in the first window sequence, the conversion information corresponding to the user behavior sequence, that is, the conversion depth and the conversion duration, is determined.

[0073] In actual business processes, the number of first user behaviors in a user behavior sequence is generally the largest. Taking the above-mentioned vehicle sales business scenario as an example, the number of the first user behavior "ad clicks" is the largest, while the number of other user behaviors, such as "user information retention", is smaller than the number of the first user behavior.

[0074] Therefore, in this way, the window sequence can be divided according to the proportion of the first user behavior and the second user behavior in the user behavior sequence, and the third user behavior with a smaller number in the user behavior sequence can be used to reduce the amount of data for user behavior analysis, improve computing efficiency, and save computer resources. In addition, since in most cases the proportion of the first user behavior is very high, the application of this solution does not require window division for each first user behavior, so the number of window sequences obtained by division will not be large, and it will not cause the system resources to be concentrated on the window sequence obtained by division according to the first user behavior. Therefore, there will be no problem of data skew, and the computing stability can be improved.

[0075] In an optional manner, multiple user behaviors in the user behavior sequence can be arranged from early to late according to the occurrence time of each user behavior. On this basis, Figure 3 FIG. 4 shows a flow chart of another user behavior analysis method provided by this embodiment. Figure 3 As shown, the user behavior analysis method may include the following steps 310 to 350:

[0076] Step 310: Obtain user behavior sequence.

[0077] The user behavior sequence may include multiple user behaviors and the occurrence time of each user behavior, and the multiple user behaviors in the user behavior sequence are arranged in order from early to late according to the occurrence time of the user behaviors.

[0078] Step 320: Determine the number of first user behaviors and the total number of at least one second user behavior in the user behavior sequence according to a preset occurrence order of the plurality of user behaviors.

[0079] The number of first user behaviors and the total number of at least one second user behavior in the acquired user behavior sequence are counted according to a preset occurrence order of the plurality of user behaviors.

[0080] Step 330: If the number of the first user behaviors is greater than or equal to the total number of at least one second user behavior, arrange the multiple user behaviors in the user behavior sequence from latest to earliest according to the occurrence time of each user behavior to obtain the reverse sequence of the user behavior sequence.

[0081] If the number of first user behaviors is greater than or equal to the total number of at least one second user behavior, it means that the number of first user behaviors in the preset occurrence sequence of multiple user behaviors completed by the user is large, and as time goes by, the user's conversion depth continues to deepen. At this time, the user behavior sequence can be arranged from late to early according to the occurrence time of each user behavior to obtain the reverse sequence of the user behavior sequence. For example, for Figure 2 The user behavior sequence shown in the figure is arranged from late to early according to the occurrence time of each user behavior, and the following can be obtained: Figure 4 Reverse sequence shown.

[0082] Step 340: taking each third user behavior in the reverse sequence as the first user behavior in the first window sequence, dividing the reverse sequence according to a preset window length to obtain at least one first window sequence.

[0083] The preset window length refers to the time window length for dividing the window sequence, which can be set according to business scenarios, etc. Taking vehicle sales as an example, the preset window length can be set to 1 hour, 1 week, 1 month, 3 months, etc. According to business needs, the preset window length can be the length of time required to complete a user conversion, wherein a user conversion can be set according to business needs and the distribution of user behavior. For example, when user A completes the three stages of user behavior of "ad click", "user information retention" and "user test drive" within 7 days, it can be determined that a user conversion is completed, and the preset window length is 7 days.

[0084] In the reverse sequence, each third user behavior can be taken as the first user behavior in the first window sequence, and all user behaviors after this user behavior can be determined according to the preset window length to obtain a first window sequence. The reverse sequence is divided in the direction from the starting point to the end point to obtain all first window sequences obtained by dividing the user behavior sequence.

[0085] Since the third user behavior is any one of the one or more second user behaviors, the third user behavior can be the user behavior with the latest preset occurrence order among the multiple user behaviors, or can be any user behavior with a preset occurrence order earlier than the latest user behavior among all the second user behaviors. For example, the third user behavior can be as follows Figure 2 E4 in the user behavior sequence shown can also be any one of E3 or E2.

[0086] Figure 5 A schematic diagram of window division provided by an embodiment of the present invention is shown, such as Figure 5 As shown, the user behavior sequence is Figure 4 Assuming that the third user behavior is the user behavior E4 with the latest preset occurrence order in the reverse sequence, according to the position of E4 and the preset window length 7, the window sequences are selected from the front to the back, and each window sequence contains the first user behavior E1.

[0087] Thus, the position of each window sequence can be located through E1 and E4 in the reverse sequence, and all the first window sequences in the reverse sequence can be obtained, that is, Figure 5 The window sequence shown in 1: {E4:t 17 ; E2:t 16 ; E2:t 15 ; E3:t 14 ; E1:t 13 ; E4:t 12 ; E3:t 11}, Window sequence 2: {E4:t 12 ; E3:t 11 ; E3:t 10 ;E2:t9;E1:t8;E3:t7;E1:t6}.

[0088] Figure 6 FIG. 4 shows another schematic diagram of window division provided by an embodiment of the present invention. Figure 6 As shown, the user behavior sequence is also Figure 4 Assume that the third user behavior is a user behavior E2 that occurs earlier than the latest user behavior in the preset order among all second user behaviors. According to the position of E2 and the preset window length 7, select window sequences from the front to the back, and each window sequence contains the first user behavior E1.

[0089] Therefore, the position of each window sequence can be located through E1 and E2 in the reverse sequence, and all the first window sequences in the reverse sequence can be obtained, that is, Figure 6 The window sequence shown in 3: {E2:t 16 ; E2:t 15 ; E3:t14 ; E1:t 13 ; E4:t 12 ; E3:t 11 ; E3:t 10} and window sequence 4: {E2:t 15 ; E3:t 14 ; E1:t 13 ; E4:t 12 ; E3:t 11 ; E3:t 10 ; E2:t9}, window sequence 5: {E2:t9; E1:t8; E3:t7; E1:t6; E2:t5; E1:t4; E2:t3}.

[0090] Step 350: Determine conversion information corresponding to the user behavior sequence according to each user behavior in the first window sequence.

[0091] Exemplarily, after all first window sequences are obtained, the types of user behaviors contained in each first window sequence and the time sequence of each type of user behavior can be determined according to each user behavior in each first window sequence. According to the time sequence, it can be determined whether each type of user behavior occurs in sequence over time and there is no missing user behavior in the middle, so as to obtain the conversion depth corresponding to the user behavior sequence and calculate the conversion duration.

[0092] Through the above steps 310 to 350, when the number of first user behaviors is greater than or equal to the total number of at least one second user behavior, the reverse sequence of the user behavior sequence can be divided into windows to generate one or more first window sequences. Since the number of first user behaviors is greater than or equal to the total number of one or more other user behaviors, in the reverse sequence obtained, the window sequence at the front is most likely to have the user behavior with the highest conversion depth, so the maximum conversion depth of the entire user behavior sequence can be quickly traversed in the divided window sequence. Therefore, the analysis efficiency of the user behavior sequence is higher.

[0093] It should be noted that the specific implementation of the above steps 310 to 320 can refer to the specific implementation of steps 110 to 120 in the above embodiment, which will not be repeated here.

[0094] Figure 7 A flowchart of another user behavior analysis provided by an embodiment of the present invention is shown. Figure 7 As shown, the following steps may be included:

[0095] Step 710: Obtain user behavior sequence.

[0096] Step 720: Determine the number of first user behaviors and the total number of at least one second user behavior in the user behavior sequence according to a preset occurrence order of the plurality of user behaviors.

[0097] Step 730: If the number of first user behaviors is greater than or equal to the total number of at least one second user behavior, in the user behavior sequence, each third user behavior is taken as the last user behavior in the first window sequence, and the user behavior sequence is divided according to a preset window length to obtain at least one first window sequence.

[0098] If the number of first user behaviors is greater than or equal to the total number of at least one second user behavior, then in the user behavior sequence, each third user behavior can be used as the last user behavior in the first window sequence, and all user behaviors from the last user behavior forward can be determined according to the preset window length to obtain a first window sequence. All first window sequences can be obtained by sequentially dividing the user behavior sequence from the starting point to the end point.

[0099] Figure 8 A schematic diagram of another window division provided by this embodiment is shown, such as Figure 8 As shown, the user behavior sequence is as follows Figure 2 When performing window division, each third user behavior, such as E4, can be taken as the last user behavior in a first window sequence, and all user behaviors in the first window sequence are determined from back to front according to the preset window length to obtain all first window sequences, that is, window sequence 6: {E1:t6; E3:t7; E1:t8; E2:t9; E3:t 10 ; E3:t 11 ; E4:t 12}, Window sequence 7: {E3:t 11 ; E4:t 12 ; E1:t 13 ; E3:t 14 ; E2:t 15 ; E2:t 16 ; E4:t 17}.

[0100] Step 740: Determine conversion information corresponding to the user behavior sequence according to each user behavior in the first window sequence.

[0101] For each first window sequence, the user behaviors that occur sequentially in the window sequence in time order and without missing user behaviors in the middle can be determined according to the types and time order of the user behaviors contained therein, so as to determine the conversion information of each first window sequence. For example, for the above window sequence 6, the sequence that occurs sequentially in time order is: {E1: t8; E2: t9; E3: t10 ; E4:t 12}, so the conversion depth of this sequence is 4, and the conversion time is t 12 -t8.

[0102] After obtaining the conversion information of each first window sequence, that is, the conversion depth and conversion duration, the conversion depths and conversion durations of the first window sequences may be further compared to determine the conversion depth and conversion duration of the user behavior sequence.

[0103] Exemplarily, the user conversion depths obtained from each first window sequence may be compared, the maximum conversion depth therein may be used as the conversion depth of the user behavior sequence, and the average conversion duration of all first window sequences may be used as the conversion duration of the user behavior sequence.

[0104] Through the above steps 710 to 740, it is possible to obtain each first window sequence without reversing the user behavior sequence, and then calculate the conversion information of the user behavior sequence, which can simplify the processing flow of the user behavior sequence and improve the analysis efficiency.

[0105] It should be noted that the specific implementation of the above steps 710 to 720 can refer to the specific implementation of steps 110 to 120 in the above embodiment, which will not be repeated here.

[0106] In an optional manner, when all first window sequences are obtained by division, the conversion information of the user behavior sequence can be calculated according to the sequence distribution of all first window sequences obtained by division. Fig. 9 A flowchart of another user behavior analysis provided by an embodiment of the present invention is shown. Fig. 9 As shown, the following steps may be included:

[0107] Step 910: Obtain user behavior sequence.

[0108] Step 920: Determine the number of first user behaviors and the total number of at least one second user behavior in the user behavior sequence according to a preset occurrence order of the plurality of user behaviors.

[0109] Step 930: If the number of the first user behaviors is greater than or equal to the total number of at least one second user behavior, the user behavior sequence is divided into windows according to the third user behavior to obtain a first window sequence.

[0110] When the number of first user behaviors is greater than or equal to the total number of all second user behaviors, each third user behavior can be located in the user behavior sequence, and then each window sequence containing the third user behavior and the first user behavior can be searched in the user behavior sequence to obtain all first window sequences.

[0111] Step 940: Arrange the occurrence time of each user behavior in the first window sequence to obtain at least one time series.

[0112] For example, in Figure 5 In the user behavior sequence shown, the first window sequence includes two window sequences, namely window sequence 1 and window sequence 2. According to the occurrence time of each user behavior in the two window sequences, window sequence 1 and window sequence 2 are traversed from E1 to E4 respectively, and time series 1 can be obtained: {E3:t 11 ; E4:t 12 ; E1:t 13 ; E3:t 14 ; E2:t 15 ; E2:t 16 ; E4:t 17}, and time series 2: {E1:t6; E3:t7; E1:t9; E2:t9; E3:t 10 ; E3:t 11 ; E4:t 12}.

[0113] Step 950: Determine a target user behavior sequence whose occurrence time conforms to a preset occurrence order according to at least one time sequence.

[0114] Among them, any two adjacent user behaviors in the target user behavior sequence are any two adjacent user behaviors in the preset occurrence order, and the target user behavior sequence includes the first user behavior. In other words, the two adjacent user behaviors in the target user behavior sequence are continuous, for example, the two adjacent user behaviors are E1 and E2, or E2 and E3. At the same time, since the conversion depth is calculated starting from the first user behavior, the target user behavior sequence can include the first user behavior, such as E1.

[0115] For example, when the occurrence time of each user behavior that meets the preset occurrence order is determined to be increasing according to the time series, it means that the occurrence time of the time series meets the preset occurrence order. Then, the target user behavior sequence can be determined according to each user behavior in the time series and its occurrence time. For example, for the above time series 1, the target user behavior sequence that meets the preset occurrence order is {E1:t 13 ; E2:t 15}, for the above time series 2, the target user behavior sequence that conforms to the preset occurrence order is {E1: t6; E2: t9; E3: t 10 ; E4:t 12}.

[0116] When the number of occurrences of the same user behavior determined according to the time series and whose occurrence time conforms to the preset occurrence order is greater than 1, the earliest time of the user behavior may be used as the basis.

[0117] Step 960: Determine conversion information corresponding to the user behavior sequence according to the number of each user behavior included in the target user behavior sequence and / or the occurrence time of each user behavior.

[0118] For example, the number of user behaviors included in the target user behavior sequence can be used as the conversion depth corresponding to the target user behavior sequence, and the time difference between the last user behavior and the first user behavior in the target user behavior sequence is the conversion duration corresponding to the target user behavior sequence. 13 ; E2:t 15}, the conversion depth is 2, and the conversion time is t 15 -t 13 , target user behavior sequence {E1:t6; E2:t9; E3:t 10 ; E4:t 12}, the conversion depth is 4, and the conversion time is t 12 -t6.

[0119] After obtaining the conversion depth and conversion duration of each target user behavior sequence, the conversion depth and conversion duration of the entire user behavior sequence can be further calculated based on the conversion depth and conversion duration of each target user behavior sequence.

[0120] In an optional manner, in steps 940 to 950, the first window sequence can be traversed to convert each first window sequence into a two-dimensional array, and the longest time sequence whose occurrence time conforms to the preset occurrence order is determined based on the two-dimensional array to determine the conversion depth and conversion duration of the user behavior sequence. In a two-dimensional array, a user behavior only appears once.

[0121] For example, in Figure 5 In the user behavior sequence shown in the figure, the user behaviors in window sequence 1 are: E4>E2>E2>E3>E1>E4>E3, and the user behaviors in window sequence 2 are: E4>E3>E3>E2>E1>E3>E2. According to the order of occurrence time, window sequence 1 and window sequence 2 are traversed from E1 to E4 respectively, and window sequence 1 and window sequence 2 are transformed into the following Fig.10 The two-dimensional array shown in FIG. 1 shows a two-dimensional array corresponding to each window sequence, and determines the longest time sequence whose occurrence time conforms to the preset occurrence order. For example, in the two-dimensional array 1 corresponding to the window sequence 1, the longest time sequence corresponding to E1 to E4 is: {E1:t 13 ; E2:t16 ; E3: t 14 ; E4: t 17}, in the two-dimensional array 2 corresponding to the window sequence 2, the corresponding longest time series from E1 to E4 is: {E1: t9; E2: t9; E3: t 11 ; E4: t 12}.

[0122] In the process of generating the above two-dimensional array, when the window sequence includes multiple time nodes of the same user behavior, the column position of the time node of the same user behavior can be determined according to the sequence order. For example, for the above window sequence 1, according to the sequence order, the time node when E2 first appears is t 16 , the second time point is t 15 , so t 16 is the first column, t 15 is the second column. The longest time series is the sequence corresponding to the first column of the two-dimensional array.

[0123] For the two-dimensional array corresponding to window sequence 1, the user behaviors that occur sequentially over time are E1 and E2, so the conversion depth of window sequence 1 is 2. For the two-dimensional array corresponding to window sequence 2, the user behaviors that occur sequentially over time are E1, E2, E3, and E4, so the conversion depth of window sequence 2 is 4. The conversion depth and conversion duration of the user behavior sequence are the conversion depth and conversion duration of window sequence 2.

[0124] Through the above steps 910 to 960, the user behavior and its occurrence time in each first window sequence can be analyzed to obtain the conversion depth and conversion duration of the user behavior sequence, thereby completing the conversion analysis of the user behavior data.

[0125] Fig.11 A flowchart of another user behavior analysis provided by an embodiment of the present invention is shown. Fig.11 As shown, the following steps may be included:

[0126] Step 1110: Obtain user behavior sequence.

[0127] Step 1120: Determine the number of first user behaviors in the user behavior sequence and the total number of at least one second user behavior according to a preset occurrence order of the plurality of user behaviors.

[0128] Step 1130: If the number of the first user behaviors is greater than or equal to the total number of at least one second user behavior, the user behavior sequence is divided into windows according to the third user behavior to obtain a first window sequence.

[0129] Step 1140: Arrange the occurrence time of each user behavior in the first window sequence to obtain at least one time series.

[0130] Step 1150: Determine a target user behavior sequence whose occurrence time conforms to a preset occurrence order according to at least one time sequence.

[0131] Any two adjacent user behaviors in the target user behavior sequence are any two adjacent user behaviors in a preset occurrence order, and the target user behavior sequence includes the first user behavior.

[0132] Step 1160: If the number of user behaviors included in the target user behavior sequence is equal to the preset conversion depth, the conversion depth corresponding to the user behavior sequence is determined as the preset conversion depth, and the difference between the latest occurrence time and the earliest occurrence time in the target user behavior sequence is determined as the conversion duration corresponding to the user behavior sequence.

[0133] The preset conversion depth may be the maximum conversion depth set according to the preset occurrence order of multiple user behaviors. Figure 2 The user behavior sequence in vehicle sales shown has a preset occurrence order of E1, E2, E3 and E4, and its preset conversion depth may be 4.

[0134] After obtaining the conversion depth and conversion duration of the target user behavior sequence, if the number of user behaviors included in a target user behavior sequence is equal to the preset conversion depth, the conversion depth corresponding to the entire user behavior sequence is equal to the above preset conversion depth, and the conversion duration can be the difference between the latest occurrence time and the earliest occurrence time in this target user behavior sequence.

[0135] Through the above steps 1110 to 1160, it is possible to determine whether the conversion depth and conversion duration of the user behavior sequence are the conversion depth and conversion duration of the target user behavior sequence based on the quantitative relationship between the number of user behaviors included in the target user behavior sequence and the preset conversion depth. After the determination, it is no longer necessary to perform window division and calculation of the remaining user behavior sequences, thereby improving the analysis efficiency of the user behavior sequence.

[0136] It should be noted that the specific implementation of the above steps 1110 to 1150 can refer to the specific implementation of steps 910 to 950 in the above embodiment, which will not be repeated here.

[0137] Considering that in some cases, the user's conversion depth will not reach the maximum depth, the window sequence divided according to the third user behavior cannot calculate the user's maximum conversion depth. In this case, in order to calculate the conversion depth and conversion time of the user behavior sequence, Fig.12A flowchart of another user behavior analysis provided by an embodiment of the present invention is shown. Fig.12 As shown, the following steps may be included:

[0138] Step 1210: Obtain user behavior sequence.

[0139] Step 1220: Determine the number of first user behaviors and the total number of at least one second user behavior in the user behavior sequence according to a preset occurrence order of the plurality of user behaviors.

[0140] Step 1230: If the number of the first user behaviors is greater than or equal to the total number of at least one second user behavior, the user behavior sequence is divided into windows according to the third user behavior to obtain a first window sequence.

[0141] Step 1240: Arrange the occurrence time of each user behavior in the first window sequence to obtain at least one time series.

[0142] Step 1250: Determine a target user behavior sequence whose occurrence time conforms to a preset occurrence order based on at least one time sequence.

[0143] Any two adjacent user behaviors in the target user behavior sequence are any two adjacent user behaviors in a preset occurrence order, and the target user behavior sequence includes the first user behavior.

[0144] Step 1260: If the number of user behaviors included in the target user behavior sequence is less than a preset conversion depth, the user behavior sequence is divided into windows according to the fourth user behavior to obtain a second window sequence.

[0145] The fourth user behavior is a user behavior whose preset occurrence order is earlier than the preset occurrence order of the third user behavior among the multiple user behaviors, and the second window sequence includes the first user behavior. For example, in the above vehicle sales, if the third user behavior is E4, the fourth user behavior can be the second to last user behavior E3 therein.

[0146] If the number of user behaviors included in all target user behavior sequences is less than the preset conversion depth, it means that the user conversion depth corresponding to these target user behavior sequences has not reached the maximum conversion depth. At this time, the user behavior sequence can be divided into windows according to the user behavior that occurs earlier than the third user behavior in a preset order among the multiple user behaviors, that is, the fourth user behavior, to obtain one or more second window sequences.

[0147] by Figure 4 Take the reverse sequence shown as an example, Fig.13 A schematic diagram of another window division sequence provided by an embodiment of the present invention is shown. Fig.13As shown, in the reverse sequence of the user behavior sequence, according to the user behavior E3 that is preset to occur earlier than the third user behavior in multiple user behaviors, as the first user behavior in each second window sequence, each second window sequence can be divided according to the preset window length, and the second window sequence includes both the fourth user behavior E3 and the first user behavior E1, to obtain all the second window sequences, such as window sequences 8 to 11 shown in the figure.

[0148] Step 1270: Determine conversion information corresponding to the user behavior sequence according to each user behavior in the second window sequence.

[0149] After the second window sequence is obtained, conversion information corresponding to the user behavior sequence, that is, conversion depth and conversion duration, may be determined according to each user behavior in each second window sequence.

[0150] Exemplarily, the second window sequence may be processed according to the method of steps 940 to 960 to determine the conversion depth and conversion duration of the user behavior sequence. Specifically, the occurrence time of each user behavior in the second window sequence may be arranged to obtain one or more time series, and then from these time series, a target user behavior sequence whose occurrence time conforms to a preset occurrence order may be determined, and the conversion depth and conversion duration corresponding to the user behavior sequence may be determined according to the number of each user behavior and / or the occurrence time of each user behavior included in the target user behavior sequence.

[0151] In an optional manner, when the conversion depth of the user behavior sequence calculated according to the second window sequence is less than the preset conversion depth-1, the user behavior sequence can be further divided into windows according to the fifth user behavior to obtain the second window sequence, and then the conversion depth and conversion duration of the new user behavior sequence are calculated. The fifth user behavior is a user behavior in a plurality of user behaviors that is adjacent to the fourth user behavior in a preset occurrence order and is earlier than the fourth user behavior. For example, when the fourth user behavior is E3, the fifth user behavior is E2.

[0152] Through the above steps 1210 to 1270, when the number of user behaviors in the target user behavior sequence is less than the preset conversion depth, the second window sequence can be further divided to calculate the conversion depth of the new user behavior sequence. Therefore, after completing one window division, the conversion depth can be calculated based on the currently divided window sequence. If the conversion depth is the maximum conversion depth that can be achieved by the window division, the calculation is stopped. Therefore, this method realizes the iterative division and calculation of the user behavior sequence, and does not need to divide all the window sequences at one time, thereby greatly reducing the calculation amount and efficiency of the conversion analysis.

[0153] It should be noted that the specific implementation of the above steps 1210 to 1250 can refer to the specific implementation of steps 910 to 950 in the above embodiment, which will not be repeated here.

[0154] In addition, due to the Figure 5 In the user behavior sequence shown, the maximum conversion depth is 4, and the window sequence obtained based on the third user behavior division has concluded that the conversion depth of the user behavior sequence is 4, so the calculation can be stopped after the conversion depth of the user behavior sequence is obtained according to the window sequence obtained by the third user behavior division. The above method of dividing the window according to the fourth user behavior and calculating the conversion depth and conversion time of the user behavior sequence is only for example description and should not limit the scope of this embodiment.

[0155] In an optional way, the number of first user behaviors may be less than the total number of all second user behaviors. For example, in the case of car sales, most users may directly perform the user behaviors of "user information retention" and "user test drive" over a period of time, but rarely perform the user behavior of "ad click". In this case, in order to calculate the conversion depth and conversion duration of the user behavior sequence, Fig.14 A flowchart of another user behavior analysis provided by an embodiment of the present invention is shown. Specifically, Fig.14 As shown, the following methods may be included:

[0156] Step 1410: Obtain user behavior sequence.

[0157] Step 1420: Determine the number of first user behaviors and the total number of at least one second user behavior in the user behavior sequence according to a preset occurrence order of the plurality of user behaviors.

[0158] Step 1430: If the number of the first user behaviors is less than the total number of at least one second user behavior, the user behavior sequence is divided into windows according to the first user behaviors to obtain a third window sequence.

[0159] If the number of first user behaviors is less than the total number of all second user behaviors, that is, the number of first user behaviors / the total number of one or more second user behaviors < 1, it means that in the entire user behavior sequence, the number of first user behaviors is the least, that is, very few of the user's operations are the first user behaviors arranged in the preset occurrence order. At this time, the user behavior sequence can be divided into windows according to the first user behavior, that is, the user behavior with the earliest preset occurrence order, to obtain a third window sequence.

[0160] Since the number of first user behaviors is less than the total number of second user behaviors, the number of window sequences divided according to the first user behaviors will also be greatly reduced, thereby reducing the amount of data for subsequent calculation of user conversion conditions and improving the analysis efficiency of user behavior sequences.

[0161] Step 1440: Determine conversion information corresponding to the user behavior sequence according to each user behavior in the third window sequence.

[0162] Exemplarily, the third window sequence may be processed according to the method of steps 940 to 960 to determine the conversion information of the user behavior sequence. Specifically, the occurrence time of each user behavior in the third window sequence may be arranged to obtain one or more time series, and then from these time series, a target user behavior sequence whose occurrence time conforms to a preset occurrence order may be determined, and according to the number of each user behavior included in the target user behavior sequence and the occurrence time of each user behavior, the conversion depth and conversion duration corresponding to the user behavior sequence may be determined.

[0163] Through the above method, when the number of the first user behavior is less than the total number of at least one second user behavior, the first user behavior can divide the user behavior sequence into windows to generate one or more third window sequences. Since the number of the first user behavior is less than the total number of one or more other user behaviors, the number of the third window sequences obtained is small, so the conversion depth and conversion time of the user behavior sequence can be quickly calculated, that is, the analysis efficiency of the user behavior sequence is higher.

[0164] It should be noted that the specific implementation of steps 1410 to 1420 can refer to the specific implementation of steps 110 to 120 in the above embodiment, which will not be repeated here.

[0165] Fig.15 A flowchart of another user behavior analysis method provided by this embodiment is shown. Fig.15 As shown, the method comprises the following steps:

[0166] Step 1510: Obtain user behavior sequence.

[0167] Step 1510: Determine the number of first user behaviors and the total number of at least one second user behavior in the user behavior sequence according to a preset occurrence order of the plurality of user behaviors.

[0168] Step 1530: Determine whether the number of the first user behavior is less than the total number of at least one second user behavior. If not, execute steps 1540 to 1560; if yes, execute steps 1570 to 1590.

[0169] Step 1540: Arrange multiple user behaviors in the user behavior sequence from late to early according to the occurrence time of each user behavior to obtain a reverse sequence of the user behavior sequence, take each third user behavior in the reverse sequence as the first user behavior in the first window sequence, divide the reverse sequence according to a preset window length, and obtain at least one first window sequence.

[0170] The third user behavior may be a user behavior with the latest preset occurrence order among the multiple user behaviors.

[0171] When the number of the first user behaviors is greater than or equal to the total number of the at least one second user behavior, the user behavior sequence may be arranged from latest to earliest according to the occurrence time of each user behavior to obtain a reverse sequence of the user behavior sequence.

[0172] Then, in the reverse sequence, each third user behavior is taken as the first user behavior in the first window sequence, and all user behaviors after this user behavior are determined according to the preset window length to obtain a first window sequence, which is divided in the direction from the starting point to the end point of the reverse sequence to obtain all first window sequences obtained by dividing the user behavior sequence.

[0173] Step 1550: Convert each first window sequence into a two-dimensional array.

[0174] The first window sequence is traversed to convert each first window sequence into a two-dimensional array. In a two-dimensional array, a user behavior appears only once.

[0175] Step 1560: In the two-dimensional array corresponding to each first window sequence, determine the longest time sequence whose occurrence time conforms to the preset occurrence order, and determine the conversion depth and conversion duration of the user behavior sequence according to the longest time sequence.

[0176] The longest time series is the time series that contains the most types of user behaviors. Based on the longest time series, it can be determined whether each user behavior occurs in sequence over time, so that the conversion depth and conversion duration of the user behavior sequence can be obtained. The conversion depth is the maximum conversion depth of the user behavior sequence.

[0177] Step 1570: In the user behavior sequence, each first user behavior is used as the first user behavior in the first window sequence, and the user behavior sequence is divided according to a preset window length to obtain at least one third window sequence.

[0178] When the number of first user behaviors is less than the total number of all second user behaviors, the user behavior sequence can be divided into windows according to the first user behaviors, that is, each first user behavior in the user behavior sequence is the first user behavior in the first window sequence, and the user behavior sequence is divided according to the preset window length to obtain one or more third window sequences. The third window sequences obtained by such division are the least, and the conversion depth and conversion time of the user behavior sequence can also be quickly obtained.

[0179] Step 1580: Convert each third window sequence into a two-dimensional array.

[0180] The first window sequence is traversed to convert each first window sequence into a two-dimensional array. In a two-dimensional array, a user behavior appears only once.

[0181] Step 1590: In the two-dimensional array corresponding to each third window sequence, determine the longest time sequence whose occurrence time conforms to the preset occurrence order, and determine the conversion depth and conversion duration of the user behavior sequence according to the longest time sequence.

[0182] Exemplarily, the occurrence time of each user behavior in the two-dimensional array corresponding to the third window sequence may be arranged in time order to obtain a longest time sequence whose occurrence time conforms to a preset occurrence order.

[0183] Then, the conversion depth and conversion duration corresponding to the user behavior sequence may be determined according to the number of each user behavior included in the longest time series and the occurrence time of each user behavior.

[0184] According to the user behavior analysis method in this embodiment, different window division strategies can be determined by comparing the number of first user behaviors and second user behaviors in the user behavior sequence to reduce the number of window sequences obtained by division, thereby quickly calculating the conversion depth and conversion duration of the user behavior sequence based on a smaller number of window sequences, thereby improving the efficiency of user conversion analysis and saving computing resources.

[0185] It should be noted that the specific implementation of steps 1510 to 1520 can refer to the specific implementation of steps 110 to 120 in the above embodiment, which will not be repeated here.

[0186] Fig.16 FIG. 1 is a schematic diagram showing the structure of a user behavior analysis device provided by an embodiment of the present invention. Fig.16As shown, the user behavior analysis device 1600 may include: an acquisition module 1610, which is used to acquire a user behavior sequence, the user behavior sequence includes multiple user behaviors and the occurrence time of each user behavior; a first determination module 1620, which is used to determine the number of first user behaviors in the user behavior sequence and the total number of at least one second user behavior according to the preset occurrence order of the multiple user behaviors; wherein the first user behavior is the user behavior with the earliest preset occurrence order among the multiple user behaviors, and the second user behavior is any user behavior whose preset occurrence order is later than the preset occurrence order of the first user behavior; a division module 1630, which is used to divide the user behavior sequence into windows according to the third user behavior to obtain a first window sequence if the number of the first user behavior is greater than or equal to the total number of the at least one second user behavior, the third user behavior is any user behavior among the above-mentioned at least one second user behavior, and the first window sequence includes the first user behavior; a second determination module 1640, which is used to determine the conversion information corresponding to the user behavior sequence according to each user behavior in the first window sequence, and the conversion information includes conversion depth and / or conversion duration.

[0187] In an optional manner, multiple user behaviors in a user behavior sequence are arranged from early to late according to the occurrence time of each user behavior. The division module 1630 can be used to arrange multiple user behaviors in the user behavior sequence from late to early according to the occurrence time of each user behavior to obtain a reverse sequence of the user behavior sequence. In the reverse sequence, each third user behavior is used as the first user behavior in the first window sequence, and the reverse sequence is divided according to a preset window length to obtain at least one first window sequence.

[0188] In an optional manner, multiple user behaviors in a user behavior sequence are arranged from early to late according to the time of occurrence of each user behavior. The division module 1630 can be used to divide the user behavior sequence according to a preset window length, with each third user behavior in the user behavior sequence being the last user behavior in the first window sequence, to obtain at least one first window sequence.

[0189] In an optional manner, the second determination module 1640 can be used to arrange the occurrence time of each user behavior in the first window sequence to obtain at least one time series, and determine a target user behavior sequence whose occurrence time conforms to a preset occurrence order based on the at least one time series, wherein any two adjacent user behaviors in the target user behavior sequence are any two adjacent user behaviors in the preset occurrence order, and the target user behavior sequence includes the first user behavior, and the conversion information corresponding to the user behavior sequence is determined based on the number of user behaviors included in the target user behavior sequence and / or the occurrence time of each user behavior.

[0190] In an optional manner, the second determination module 1640 can also be used to determine the conversion depth corresponding to the user behavior sequence as the preset conversion depth if the number of user behaviors included in the target user behavior sequence is equal to the preset conversion depth, and determine the difference between the latest occurrence time and the earliest occurrence time in the target user behavior sequence as the conversion duration corresponding to the user behavior sequence.

[0191] In an optional manner, the second determination module 1640 can also be used to divide the user behavior sequence into windows according to the fourth user behavior to obtain a second window sequence if the number of user behaviors included in the target user behavior sequence is less than the preset conversion depth. The fourth user behavior is a user behavior in a plurality of user behaviors whose preset occurrence order is earlier than the preset occurrence order of the third user behavior. The second window sequence includes the first user behavior. According to each user behavior in the second window sequence, the conversion information corresponding to the user behavior sequence is determined.

[0192] In an optional manner, the division module 1630 can also be used to perform window division on the user behavior sequence according to the first user behavior to obtain a third window sequence if the number of the first user behaviors is less than the total number of at least one second user behavior. The second determination module 1640 can also be used to determine the conversion information corresponding to the user behavior sequence according to each user behavior in the third window sequence.

[0193] Fig.17 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention is shown. The specific embodiment of the present invention does not limit the specific implementation of the electronic device.

[0194] like Fig.17 As shown, the electronic device may include: a processor (processor) 1702 , a communication interface (Communications Interface) 1704 , a memory (memory) 1706 , and a communication bus 1708 .

[0195] The processor 1702, the communication interface 1704, and the memory 1706 communicate with each other via the communication bus 1708. The communication interface 1704 is used to communicate with other devices such as a client or other server network elements. The processor 1702 is used to execute the program 1710, which can specifically execute the relevant steps in the above-mentioned embodiment of the method for analyzing user behavior.

[0196] Specifically, program 1710 may include program code including computer executable instructions.

[0197] The processor 1702 may be a central processing unit (CPU), or an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention. The one or more processors included in the electronic device may be processors of the same type, such as one or more CPUs; or processors of different types, such as one or more CPUs and one or more ASICs.

[0198] The memory 1706 is used to store the program 1710. The memory 1706 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0199] Program 1710 can be specifically called by processor 1702 to enable the electronic device to execute the operation steps of the above-mentioned user behavior analysis method.

[0200] An embodiment of the present invention provides a computer-readable storage medium, which stores at least one executable instruction. When the executable instruction is executed on an electronic device, the electronic device executes the user behavior analysis method in any of the above method embodiments.

[0201] The executable instructions can be specifically used to enable the electronic device to execute the operating steps of the above-mentioned user behavior analysis method.

[0202] The algorithms or displays provided herein are not inherently related to any particular computer, virtual system or other device. In addition, the embodiments of the present invention are not directed to any particular programming language.

[0203] In the description provided herein, a large number of specific details are described. However, it is understood that embodiments of the present invention can be practiced without these specific details. Similarly, in order to simplify the present invention and help understand one or more of the various inventive aspects, in the above description of exemplary embodiments of the present invention, the various features of the embodiments of the present invention are sometimes grouped together into a single embodiment, figure, or description thereof. Wherein, the claims that follow the specific embodiment are hereby expressly incorporated into the specific embodiment, wherein each claim itself is a separate embodiment of the present invention.

[0204] Those skilled in the art will appreciate that the modules in the devices in the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and further may be divided into a plurality of submodules or subunits or subcomponents. Except that at least some of such features and / or processes or units are mutually exclusive.

[0205] It should be noted that the above embodiments illustrate the present invention rather than limit it, and that those skilled in the art may devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference symbol between brackets shall not be construed as a limitation on the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "one" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present invention may be implemented by means of hardware comprising a number of different elements and by means of a suitably programmed computer. In a unit claim enumerating a number of devices, several of these devices may be embodied by the same hardware item. The use of the words first, second, and third, etc. does not indicate any order. These words may be interpreted as names. The steps in the above embodiments, unless otherwise specified, should not be understood as limitations on the order of execution.

Claims

1. A user behavior analysis method, characterized in that: The method comprises: Acquire a user behavior sequence, wherein the user behavior sequence includes multiple user behaviors and the occurrence time of each of the user behaviors; Determine the number of first user behaviors in the user behavior sequence and the total number of at least one second user behavior according to the preset occurrence order of the multiple user behaviors; wherein the first user behavior is the user behavior with the earliest preset occurrence order among the multiple user behaviors, and the second user behavior is any user behavior with a preset occurrence order later than the preset occurrence order of the first user behavior; If the number of the first user behaviors is greater than or equal to the total number of at least one of the second user behaviors, dividing the user behavior sequence into windows according to a third user behavior to obtain a first window sequence, wherein the third user behavior is any one of the at least one second user behavior, and the first window sequence includes the first user behavior; According to each user behavior in the first window sequence, conversion information corresponding to the user behavior sequence is determined, and the conversion information includes at least one of a conversion depth and a conversion duration.

2. The method according to claim 1, characterized in that The multiple user behaviors in the user behavior sequence are arranged from early to late according to the occurrence time of each user behavior, and the user behavior sequence is divided into windows according to the third user behavior to obtain a first window sequence, including: Arrange the multiple user behaviors in the user behavior sequence from latest to earliest according to the occurrence time of each user behavior to obtain a reverse sequence of the user behavior sequence; In the reverse sequence, each of the third user behaviors is used as the first user behavior in the first window sequence, and the reverse sequence is divided according to a preset window length to obtain at least one first window sequence.

3. The method according to claim 1, characterized in that The multiple user behaviors in the user behavior sequence are arranged from early to late according to the occurrence time of each user behavior, and the user behavior sequence is divided into windows according to the third user behavior to obtain a first window sequence, including: In the user behavior sequence, each of the third user behaviors is taken as the last user behavior in the first window sequence, and the user behavior sequence is divided according to a preset window length to obtain at least one first window sequence.

4. The method according to claim 2 or 3, characterized in that: The determining, according to each user behavior in the first window sequence, conversion information corresponding to the user behavior sequence includes: Arranging the occurrence time of each user behavior in the first window sequence to obtain at least one time series; Determine, according to the at least one time sequence, a target user behavior sequence whose occurrence time conforms to the preset occurrence sequence; wherein any two adjacent user behaviors in the target user behavior sequence are any two adjacent user behaviors in the preset occurrence sequence, and the target user behavior sequence includes the first user behavior; According to the number of each user behavior and / or the occurrence time of each user behavior included in the target user behavior sequence, conversion information corresponding to the user behavior sequence is determined.

5. The method according to claim 4, characterized in that Determining conversion information corresponding to the user behavior sequence according to the number of user behaviors included in the target user behavior sequence and / or the occurrence time of each user behavior includes: If the number of user behaviors included in the target user behavior sequence is equal to the preset conversion depth, the conversion depth corresponding to the user behavior sequence is determined as the preset conversion depth, and the difference between the latest occurrence time and the earliest occurrence time in the target user behavior sequence is determined as the conversion duration corresponding to the user behavior sequence.

6. The method according to claim 4, characterized in that Determining conversion information corresponding to the user behavior sequence according to the number of user behaviors included in the target user behavior sequence and / or the occurrence time of each user behavior includes: If the number of user behaviors included in the target user behavior sequence is less than a preset conversion depth, dividing the user behavior sequence into windows according to a fourth user behavior to obtain a second window sequence, wherein the fourth user behavior is a user behavior whose preset occurrence order is earlier than the preset occurrence order of the third user behavior among the multiple user behaviors, and the second window sequence includes the first user behavior; According to each user behavior in the second window sequence, conversion information corresponding to the user behavior sequence is determined.

7. The method according to claim 1, characterized in that The method further comprises: If the number of the first user behaviors is less than the total number of the at least one second user behavior, dividing the user behavior sequence into windows according to the first user behaviors to obtain a third window sequence; According to each user behavior in the third window sequence, conversion information corresponding to the user behavior sequence is determined.

8. A user behavior analysis device, characterized in that: The device comprises: An acquisition module, used to acquire a user behavior sequence, wherein the user behavior sequence includes multiple user behaviors and the occurrence time of each of the user behaviors; A first determination module, configured to determine the number of first user behaviors in the user behavior sequence and the total number of at least one second user behavior according to a preset occurrence order of the multiple user behaviors; wherein the first user behavior is the user behavior with the earliest preset occurrence order among the multiple user behaviors, and the second user behavior is any user behavior with a preset occurrence order later than the preset occurrence order of the first user behavior; a partitioning module, configured to, if the number of the first user behaviors is greater than or equal to the total number of at least one of the second user behaviors, partition the user behavior sequence into windows according to a third user behavior to obtain a first window sequence, wherein the third user behavior is any one of the at least one second user behavior, and the first window sequence includes the first user behavior; The second determination module is used to determine conversion information corresponding to the user behavior sequence according to each user behavior in the first window sequence, and the conversion information includes at least one of a conversion depth and a conversion duration.

9. An electronic device, characterized in that: include: processor; A memory, for storing at least one executable instruction; The executable instructions enable the processor to perform the operations of the user behavior analysis method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The storage medium stores at least one executable instruction. When the executable instruction is executed on the electronic device, the electronic device executes the operation of the user behavior analysis method according to any one of claims 1 to 7.

Images