Big data-based user behavior analysis method and system

Abstract

The present invention relates to the technical field of big data analysis. Disclosed are a big data-based user behavior analysis method and system. The present invention comprises: collecting user behavior data from data terminals with user tags and without user tags, performing classification processing on the basis of the user tags, obtaining a standard data list and a hybrid database, and performing classification on the hybrid database on the basis of the standard data list to obtain an extended data list and an additional data list, so as to ensure data consistency and standardization; obtaining target behavior data, and, on the basis of the standard data list, the extended data list and the additional data list, performing feature analysis, so as to extract list position information of the target behavior data; and performing feature analysis processing on the basis of the list position information of the target behavior data, so as to generate behavior analysis tags. The present invention helps to gain deeper insights into user behavior patterns and predict behavior trends, solving the problem in the prior art that confusion is likely to occur during user behavior analysis.

Description

Big data-based user behavior analysis method and system TECHNICAL FIELD

[0001] The present application relates to the technical field of big data analysis, and particularly relates to a big data-based user behavior analysis method and system. BACKGROUND

[0002] With the rapid development of the Internet and mobile Internet, user behavior data has become one of the important assets of enterprises, and plays a key role in understanding user behavior, improving user experience, and precise marketing.

[0003] In the prior art, big data analysis technology is usually used to extract and compare data features of a large amount of terminal data to find the change rule of different data features, and then the rule is used to predict the behavior of a specific user.

[0004] However, when analyzing a large amount of data, there are many similar data features that are easily confused, which leads to deviation in the analysis of user behavior.

[0005] SUMMARY

[0006] The present application aims to provide a big data-based user behavior analysis method and system, which aims to solve the problem of confusion in the analysis of user behavior in the prior art.

[0007] The present application is implemented as follows: in a first aspect, the present application provides a big data-based user behavior analysis method, comprising:

[0008] Based on a plurality of data terminals with user labels, user behavior data is continuously collected, and the collected first user behavior data is processed according to the corresponding user labels to obtain a standard data list;

[0009] Based on a plurality of data terminals without user labels, user behavior data is continuously collected, and the collected second user behavior data is integrated to obtain a mixed database;

[0010] According to the standard data list, the mixed database is classified and processed to convert the mixed database into an extended data list corresponding to the standard data list and an additional data list not corresponding to the standard data list;

[0011] Obtain target behavior data, and perform feature analysis on the target behavior data according to the standard data list, the extended data list and the additional data list to obtain list position information of the target behavior data;

[0012] The target behavior data is analyzed based on the list position information of the target behavior data to obtain a behavior analysis label of the target behavior data.

[0013] In a second aspect, the application provides a user behavior analysis system based on big data, which is used to implement the user behavior analysis method based on big data in any one of the first aspect.

[0014] A list induction module is configured to continuously collect user behavior data based on a plurality of data terminals with user labels, and to induce the collected first user behavior data according to the corresponding user labels to obtain a standard data list.

[0015] A data integration module is configured to continuously collect user behavior data based on a plurality of data terminals without user labels, and to integrate the collected second user behavior data to obtain a mixed database.

[0016] A data analysis module is configured to classify the mixed database according to the standard data list to convert the mixed database into an extended data list corresponding to the standard data list and an additional data list not corresponding to the standard data list.

[0017] A list analysis module is configured to obtain target behavior data, and to analyze the target behavior data according to the standard data list, the extended data list and the additional data list to obtain list position information of the target behavior data.

[0018] A behavior analysis module is configured to analyze the target behavior data based on the list position information of the target behavior data to obtain a behavior analysis label of the target behavior data.

[0019] The application provides a user behavior analysis method based on big data, which has the following beneficial effects:

[0020] The application collects user behavior data from data terminals with user labels and data terminals without user labels, classifies the user behavior data according to the user labels to obtain a standard data list and a mixed database, classifies the mixed database according to the standard data list to form an extended data list and an additional data list, ensures data consistency and standardization, obtains target behavior data, analyzes the target behavior data according to the standard data list, the extended data list and the additional data list to extract list position information of the target behavior data, analyzes the target behavior data based on the list position information of the target behavior data to generate a behavior analysis label, which helps to deeply understand user behavior patterns and predict behavior trends, and solves the problem that user behavior analysis is easily confused in the prior art. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1 is a schematic diagram of the steps of a user behavior analysis method based on big data provided in an embodiment of the present invention;

[0022] Figure 2 is a schematic diagram of the structure of a user behavior analysis system based on big data provided in an embodiment of the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0024] The implementation of the present invention will be described in detail below with reference to specific embodiments.

[0025] Referring to Figures 1 and 2, a preferred embodiment of the present invention is provided.

[0026] In a first aspect, the present invention provides a user behavior analysis method based on big data, comprising:

[0027] S1: Based on several data terminals with user tags, user behavior data is continuously collected, and the collected first user behavior data is summarized and processed according to the corresponding user tags to obtain a standard data list.

[0028] S2: Based on several data terminals without user tags, continuously collect user behavior data, and integrate the collected second user behavior data to obtain a hybrid database;

[0029] S3: Perform data classification processing on the hybrid database according to the standard data list to convert the hybrid database into an extended data list corresponding to the standard data list and an additional data list not corresponding to the standard data list;

[0030] S4: Obtain target behavior data, and perform feature analysis processing on the target behavior data according to the standard data list, the extended data list and the additional data list to obtain the list position information of the target behavior data;

[0031] S5: Based on the list position information of the target behavior data, perform data feature analysis and processing on the target behavior data to obtain the behavior analysis tags of the target behavior data.

[0032] Specifically, in step S1 of the embodiment provided by the present application, user behavior data is continuously collected from data terminals with user tags, i.e. electronic intelligent terminals with interactive functions such as mobile phones, computers, etc., and the user behavior data, i.e. operation data of the user on the data terminal, includes various operations of the user on the data terminal and the time, type, etc. corresponding to various operations.

[0033] More specifically, the user tag described above is a tag for describing the image of the holder of the data terminal, and the user tag can be generated by the user filling in the tag generation program sent to the data terminal. The user behavior data collected by the data terminal is classified and analyzed through the user tag, so that the data characteristics of the user behavior data corresponding to each type of user tag can be obtained for subsequent comparative analysis.

[0034] More specifically, the data induction processing: according to the user tag, the collected user behavior data is induced and processed, and the data after the induction processing is arranged into a standard data list, which includes different user tags and corresponding behavior data.

[0035] It can be understood that by using the user tag for data processing, the behavior habits and preferences of each user can be better understood, thereby realizing personalized data processing and analysis, and the user behavior data is induced and arranged into a standard data list, which is helpful to unify the data format and structure, and facilitates subsequent data analysis and application.

[0036] Specifically, in step S2 of the embodiment provided by the present application, user behavior data is continuously collected from data terminals without user tags, and the collected user behavior data is integrated and processed and merged into a database.

[0037] More specifically, the data terminal without a user tag can be a data terminal that has sent a tag generation program but the user has not responded. The user of these data terminals has potential user tags, i.e. user tags that would be generated if the user filled in the tag generation program. These user tags may correspond to the user tags in the standard data list, or there may be no correspondence.

[0038] It can be understood that by integrating and processing data terminals without user tags, data from different terminals can be integrated into a unified database, facilitating subsequent data management and analysis.

[0039] Specifically, in step S3 of the embodiment provided by the present application, the data in the standard data list and the mixed database are matched, the data matching the tags in the standard data list is classified as an extended data list, and the data that fails to match the tags in the standard data list is classified as an additional data list.

[0040] It can be understood that by classifying the data in the mixed database according to the labels in the standard data list, the classification and standardization of the data are realized, the data is easier to manage and analyze, the data matched to the standard data list is integrated into the extended data list, and the data can be further analyzed and applied to better understand user behavior and trends, and the data that fails to match the standard data list is classified into the additional data list, which helps to find abnormal data or supplement information that may be missing in the standard data list, and provides a more comprehensive basis for data analysis.

[0041] Specifically, in step S4 of the embodiments provided by the present application, target behavior data is obtained from the data source, which is newly collected or existing data, and the target behavior data is matched with the standard data list, the extended data list and the additional data list, which can be achieved by the corresponding relationship of the data fields or other associated information.

[0042] More specifically, the extracted feature information is analyzed and processed, which can use statistical analysis, machine learning algorithm and other methods to obtain the list position information of the target behavior data, that is, to identify the association information of the target behavior data with the standard data list, the extended data list and the additional data list, which can be considered as the expression of the target behavior data based on the standard data list, the extended data list and the additional data list.

[0043] It can be understood that by matching with the standard data list, the extended data list and the additional data list, the association analysis of the target behavior data is realized, the relationship and characteristics between the data are found, the data in the extended data list can enrich the feature information of the target behavior data, and the accuracy and comprehensiveness of data analysis are improved, and through the feature analysis processing, the list position information of the target behavior data is obtained, which helps to understand the user behavior in depth.

[0044] Specifically, in step S5 of the embodiments provided by the present application, the target behavior data and the list position information related thereto are prepared, the relevant features are extracted from the target behavior data according to each list and the corresponding list position information to obtain the behavior analysis label.

[0045] More specifically, the behavior analysis label is used to describe the future trend of user behavior and the user image feedback by user behavior: in each list, the behavior characteristics of various types of users are reflected, so according to the list position information, the information feedback by the target behavior data can be positioned, and through the prediction analysis of the positioned data, the future behavior trend of the user and the image feedback of the user are obtained, thereby producing the behavior analysis label.

[0046] More specifically, the behavior analysis label is used to feedback the information implied by the behavior from the target behavior data, so other aspects of analysis can also be carried out.

[0047] It can be understood that by analyzing the characteristics of the target behavior data, the behavior patterns and trends of the user can be identified, providing a basis for understanding the user behavior.

[0048] The present application provides a user behavior analysis method based on big data, which has the following beneficial effects:

[0049] The present application collects user behavior data from data terminals with and without user labels, classifies and processes the data according to the user labels to obtain a standard data list and a mixed database, classifies the mixed database according to the standard data list to form an extended data list and an additional data list, ensures data consistency and standardization, obtains target behavior data, performs feature analysis based on the list position information of the target behavior data, generates behavior analysis labels, helps to understand user behavior patterns and predict behavior trends, and solves the problem of confusion in user behavior analysis in the prior art.

[0050] Preferably, the step of collecting user behavior data based on a plurality of data terminals with user labels and processing the collected first user behavior data according to the corresponding user labels to obtain a standard data list comprises:

[0051] S11: Marking the data terminals with user labels as standard terminals, and generating classification numbers of the standard terminals according to the user labels;

[0052] S12: Continuously collecting user behavior data from each standard terminal, and combining the user behavior data of each standard terminal to obtain a terminal data group;

[0053] S13: Constructing a classification coordinate axis and setting a plurality of classification scales on the classification coordinate axis according to the classification numbers of each standard terminal; wherein each classification scale corresponds to each classification number respectively;

[0054] S14: Constructing a data filling axis based on each classification scale of the classification coordinate axis, and filling the terminal data group with the classification number corresponding to the classification scale into the data filling axis;

[0055] S15: performing data feature extraction processing on each group of terminal data groups in each of the data filling axes on the classification coordinate axis to obtain standard data features of the classification number corresponding to each of the data filling axes;

[0056] S16: performing correlation analysis processing between each of the standard data features of the classification number to obtain similarity parameters between each of the standard data features of the classification number, and adjusting the positions of each of the classification scales on the classification coordinate axis and each of the data filling axes according to the similarity parameters between each of the standard data features to complete the construction of the standard data list.

[0057] Specifically, data terminals with user labels are marked as standard terminals, and a classification number is generated for each terminal. The user behavior data of the standard terminals is continuously collected, and the user behavior data of each terminal is combined to form a data group of each terminal.

[0058] More specifically, a classification coordinate axis and a data filling axis are constructed, classification scales are set on the classification coordinate axis, and each scale corresponds to a classification number. Then, a data filling axis is constructed for each classification scale, and the terminal data group of the corresponding classification number is filled into the data filling axis.

[0059] More specifically, data features are extracted from the terminal data groups in each data filling axis to obtain standard data features of the classification number.

[0060] More specifically, the standard data features are correlated with each other to determine similarity parameters between the features. Then, according to these similarity parameters, the positions of the classification scales on the classification coordinate axis and the data filling axes are adjusted to complete the construction of the standard data list.

[0061] It can be understood that through the collection and processing of user behavior data, the standardization and classification of data are realized, making the data more easily managed and analyzed. Extracting features from the user behavior data of each terminal helps to understand user behavior patterns and trends. Through similarity analysis and position adjustment of standard data features, the organization structure of data is optimized, and the readability and analysis efficiency of data are improved.

[0062] Preferably, the step of continuously collecting user behavior data based on a plurality of data terminals without user labels and integrating the collected second user behavior data to obtain a mixed database comprises:

[0063] S21: marking the data terminals without user labels as unknown terminals, and continuously collecting user behavior data of each unknown terminal, and combining the user behavior data of each unknown terminal to obtain each unknown data group;

[0064] S22: constructing a mixed database, and setting each unknown data group in the mixed database.

[0065] Specifically, the data terminals without user labels are marked as unknown terminals, and the user behavior data of these unknown terminals is continuously collected to form each unknown data group, a mixed database is created, and each unknown data group is integrated into the mixed database.

[0066] It can be understood that by constructing the mixed database, the integration and unified management of the data terminals without user labels are realized, which facilitates subsequent data analysis and application.

[0067] Preferably, the step of performing data classification processing on the mixed database according to the standard data list to convert the mixed database into an extended data list corresponding to the standard data list and an additional data list not corresponding to the standard data list comprises:

[0068] S31: comparing each unknown data group in the mixed database with the standard data list for data features to obtain deviation feature distribution of each unknown data group and the standard data list; wherein the deviation feature distribution is used to describe the deviation degree between the standard data features of the classification numbers corresponding to each data filling axis of the unknown data group and the standard data list;

[0069] S32: dividing each unknown data group into an extended data group and an additional data group according to the deviation feature distribution of each unknown data group and the standard data list; wherein the deviation degree between the data features of the extended data group and the standard data features of the classification numbers corresponding to one data filling axis is within a preset standard, and the deviation degree between the data features of the additional data group and the standard data features of the classification numbers corresponding to each data filling axis is not within the preset standard;

[0070] S33: constructing each extended data according to the format of the standard data list to obtain the extended data list;

[0071] S34: On the basis of the standard data list, according to the deviation characteristic distribution of each of the additional data groups and the positional relationship between each of the data filling axes on the classification coordinate axis in the standard data list, the data filling axes are adjusted, and each of the additional data groups is filled into the adjusted data filling axes to construct the additional data list.

[0072] Specifically, each unknown data group in the mixed database is compared with the standard data list to obtain the deviation characteristic distribution of each unknown data group and the standard data list. This distribution describes the deviation degree between each data filling axis of the unknown data group and the standard data list.

[0073] More specifically, according to the deviation characteristic distribution, the unknown data groups are divided into extended data groups and additional data groups. The data characteristics of the extended data groups deviate from the standard data characteristics within a preset standard, while the data characteristics of the additional data groups deviate from the standard data characteristics beyond the preset standard.

[0074] More specifically, according to the format of the standard data list, an extended data list is constructed, and the extended data groups are filled into the extended data list, that is, a classification coordinate axis is also constructed, a plurality of classification scales are set on the classification coordinate axis, and a data filling axis is set based on each classification scale, and then data filling processing is sequentially performed.

[0075] More specifically, on the basis of the standard data list, according to the deviation characteristic distribution of the additional data groups and the positional relationship between the data filling axes, the data filling axes are adjusted, and the additional data groups are filled into the adjusted data filling axes to construct the additional data list; that is, the additional data groups deviate from the standard data groups of each data filling axis of the standard data list, so the data filling axis corresponding to the additional data groups and the data filling axis corresponding to the standard data groups should have a corresponding positional difference.

[0076] It can be understood that through the comparison processing and the deviation characteristic distribution, the data classification of the mixed database is realized, and the mixed database is converted into the extended data list and the additional data list. The data in the extended data list conforms to the standard data characteristics, and the additional data list supplements the data characteristics not in the standard data list, so that the data is more complete and accurate.

[0077] Preferably, the step of obtaining target behavior data and performing feature analysis processing on the target behavior data according to the standard data list, the extended data list, and the additional data list to obtain list position information of the target behavior data includes:

[0078] S41: Obtain target behavior data, and perform data feature extraction processing on the target behavior data to obtain target behavior features corresponding to the target behavior data;

[0079] S42: Perform analysis processing on the target behavior features according to the standard data list to obtain a standard deviation distribution of the target behavior features and the standard data list; wherein the standard deviation distribution is used to describe a deviation degree between the target behavior features and the standard data features of the classification number corresponding to each data filling axis of the standard data list;

[0080] S43: Perform analysis processing on the target behavior features according to the extended data list to obtain an extended deviation distribution of the target behavior features and the extended data list; wherein the extended deviation distribution is used to describe a deviation degree between the target behavior features and various data features of the extended data list;

[0081] S44: Perform analysis processing on the target behavior features according to the additional data list to obtain an additional deviation distribution of the target behavior features and the additional data list; wherein the additional deviation distribution is used to describe a deviation degree between the target behavior features and various data features of the additional data list;

[0082] S45: The standard deviation distribution, the extended deviation distribution, and the additional deviation distribution are collectively used as list position information of the target behavior data.

[0083] Specifically, first, data features are extracted from target behavior data to obtain target behavior features, analysis processing is performed using a standard data list to obtain a standard deviation distribution of the target behavior features and the standard data list, and a deviation degree of the target behavior features and each data filling axis of the standard data list is described.

[0084] More specifically, analysis processing is performed using an extended data list to obtain an extended deviation distribution of the target behavior features and the extended data list, and a deviation degree of the target behavior features and various data features of the extended data list is described.

[0085] More specifically, analysis processing is performed using an additional data list to obtain an additional deviation distribution of the target behavior features and the additional data list, and a deviation degree of the target behavior features and various data features of the additional data list is described.

[0086] More specifically, the standard deviation distribution, the extended deviation distribution, and the additional deviation distribution are used as list position information of the target behavior data.

[0087] It can be understood that the comprehensive analysis: by analyzing and processing the deviation distribution of the target behavior characteristics and different data lists, the relationship between the target behavior data and the standard data, the extended data and the additional data is comprehensively considered, the accuracy and comprehensiveness of data analysis are improved, the characteristics of the target behavior data are compared and analyzed with the characteristics of different data lists, the differences and similarities between them are revealed, which is helpful for further data matching and correlation analysis.

[0088] Preferably, the step of analyzing and processing the data characteristics of the target behavior data based on the list position information of the target behavior data to obtain the behavior analysis label of the target behavior data comprises:

[0089] S51: According to the standard deviation distribution of the list position information of the target behavior data, if the analysis result of the standard deviation distribution shows that the deviation degree of the target behavior characteristics of the target behavior data and the standard data characteristics corresponding to the classification number of one data filling axis of the standard data list meets the preset standard, the classification number and the standard data characteristics of the classification number are taken as the predicted characteristics of the target behavior data;

[0090] S52: According to the extended deviation distribution of the list position information of the target behavior data, if the analysis result of the extended deviation distribution shows that the deviation degree of the target behavior characteristics of the target behavior data and one data characteristic of one data filling axis of the extended data list meets the preset standard, the data characteristic is taken as the predicted characteristic of the target behavior data;

[0091] S53: According to the additional deviation distribution of the list position information of the target behavior data, if the analysis result of the additional deviation distribution shows that the deviation degree of the target behavior characteristics of the target behavior data and one data characteristic of one data filling axis of the additional data list meets the preset standard, the data characteristic is taken as the predicted characteristic of the target behavior data;

[0092] S54: According to the predicted characteristics of the target behavior data, the behavior prediction processing of the target behavior data is carried out to obtain the behavior prediction information of the target behavior data;

[0093] S55: According to the predicted characteristics of the target behavior data, the user image tracing processing of the target behavior data is carried out to obtain the user image data of the target behavior data;

[0094] S56: The behavior prediction information and the user image data are taken as the behavior analysis label of the target behavior data together.

[0095] Specifically, according to the standard deviation distribution of the list position information of the target behavior data, if the deviation of the target behavior feature of the target behavior data and the standard data feature of a certain classification number of the standard data list meets the preset standard, the classification number and the corresponding standard data feature are taken as the prediction feature.

[0096] More specifically, according to the extended deviation distribution of the list position information of the target behavior data, if the deviation of the target behavior feature of the target behavior data and a certain data feature of the extended data list meets the preset standard, the data feature is taken as the prediction feature.

[0097] More specifically, according to the additional deviation distribution of the list position information of the target behavior data, if the deviation of the target behavior feature of the target behavior data and a certain data feature of the additional data list meets the preset standard, the data feature is taken as the prediction feature.

[0098] More specifically, according to the prediction feature, the target behavior data is processed for behavior prediction, to obtain behavior prediction information, and the target behavior data is processed for user image tracing according to the prediction feature, to obtain user image data, and the behavior prediction information and the user image data are taken as the behavior analysis label of the target behavior data.

[0099] It can be understood that by analyzing and processing the deviation distribution and the prediction feature, the behavior prediction of the target behavior data is realized, the future behavior trend is provided with reference, the user image tracing of the target behavior data is performed according to the prediction feature, which helps to understand the interests and preferences of the user, and the behavior prediction information and the user image data are combined to generate the behavior analysis label.

[0100] Referring to FIG. 2, in a second aspect, the present application provides a user behavior analysis system based on big data, which is used to realize the user behavior analysis method based on big data in any one of the first aspect, and comprises:

[0101] The list induction module is used to continuously collect the user behavior data based on a plurality of data terminals with user labels, and to process the collected user behavior data according to the corresponding user labels to obtain a standard data list.

[0102] The data integration module is used to continuously collect the user behavior data based on a plurality of data terminals without user labels, and to process the collected user behavior data to obtain a mixed database.

[0103] a data analysis module, configured to perform data classification processing on the mixed database according to the standard data list, so as to convert the mixed database into an extended data list corresponding to the standard data list and an additional data list not corresponding to the standard data list;

[0104] a list analysis module, configured to acquire target behavior data, and perform feature analysis processing on the target behavior data according to the standard data list, the extended data list and the additional data list, to obtain list position information of the target behavior data;

[0105] a behavior analysis module, configured to perform data feature analysis processing on the target behavior data based on the list position information of the target behavior data, to obtain a behavior analysis label of the target behavior data.

[0106] In the embodiment, the specific implementation of each module in the system embodiment is described above in the method embodiment, and will not be described here.

[0107] The above merely describes the preferred embodiments of the present application and is not intended to limit the present application. Any modification, equivalent replacement and improvement within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A big data based user behavior analysis method, characterized in that, The application comprises the following steps: Based on a number of data terminals with user tags, the user behavior data is continuously collected, and the first user behavior data collected is inductively processed according to the corresponding user tags to obtain a standard data list; Based on a number of data terminals without user tags, the user behavior data is continuously collected, and the second user behavior data collected is integrated to obtain a mixed database; According to the standard data list, the mixed database is classified and processed to convert the mixed database into an extended data list corresponding to the standard data list and an additional data list not corresponding to the standard data list; Obtain target behavior data, and perform feature analysis processing on the target behavior data according to the standard data list, the extended data list and the additional data list to obtain list position information of the target behavior data; Based on the list position information of the target behavior data, the data features of the target behavior data are analyzed and processed to obtain the behavior analysis tag of the target behavior data.

2. The big data based user behavior analysis method of claim 1, wherein, The step of continuously collecting user behavior data based on a number of data terminals with user tags and inductively processing the first user behavior data collected according to the corresponding user tags to obtain a standard data list comprises the following steps: Mark the data terminal with the user tag as a standard terminal, and generate a classification number of the standard terminal according to the user tag; Continuously collect user behavior data of each standard terminal, combine the first user behavior data collected by each standard terminal to obtain a terminal data group; Build a classification coordinate axis, and set a plurality of classification scales on the classification coordinate axis according to the classification number of each standard terminal; wherein each classification scale corresponds to each classification number respectively; Based on each classification scale of the classification coordinate axis, a data filling axis is constructed respectively, and the terminal data group with the classification number corresponding to the classification scale is filled into the data filling axis; Extract the data features of each group of terminal data groups in each data filling axis on the classification coordinate axis to obtain the standard data features of the classification number corresponding to each data filling axis; Perform correlation analysis processing on the standard data features of each classification number to obtain the similarity parameters between the standard data features of each classification number, adjust the positions of each classification scale on the classification coordinate axis and each data filling axis according to the similarity parameters between each standard data feature, and take the classification coordinate axis and each adjusted data filling axis as the standard data list. The step of continuously collecting user behavior data based on a number of data terminals without user tags and integrating the second user behavior data collected to obtain a mixed database comprises the following steps:

3. The big data based user behavior analysis method of claim 2, wherein, ​ Mark the data terminal without user label as unknown terminal, and continuously collect user behavior data of each unknown terminal, combine the second user behavior data collected by each unknown terminal to obtain each unknown data group; Construct a mixed database, and set each unknown data group in the mixed database.

4. The big data based user behavior analysis method of claim 3, wherein, The step of performing data classification processing on the mixed database according to the standard data list to convert the mixed database into an extended data list corresponding to the standard data list and an additional data list not corresponding to the standard data list comprises: Compare each unknown data group in the mixed database with the standard data list in data feature to obtain the deviation feature distribution of each unknown data group and the standard data list; wherein the deviation feature distribution is used to describe the deviation degree between the standard data feature of the classification number corresponding to each data filling axis of the unknown data group and the standard data list; According to the deviation feature distribution of each unknown data group and the standard data list, divide each unknown data group into an extended data group and an additional data group; wherein the deviation degree of the data feature of the extended data group and the standard data feature of the classification number corresponding to one data filling axis is within the preset standard, and the deviation degree of the data feature of the additional data group and the standard data feature of the classification number corresponding to each data filling axis is not within the preset standard; According to the format of the standard data list, construct the extended data list from each extended data group; On the basis of the standard data list, according to the deviation feature distribution of each additional data group and the positional relationship between each data filling axis on the classification coordinate axis of the standard data list, adjust each data filling axis, and fill each additional data group to the adjusted data filling axis to construct the additional data list. The step of obtaining target behavior data and performing feature analysis processing on the target behavior data according to the standard data list, the extended data list and the additional data list to obtain the list position information of the target behavior data comprises:

5. The big data based user behavior analysis method of claim 4, wherein, Obtain target behavior data and extract the data feature of the target behavior data to obtain the target behavior feature corresponding to the target behavior data; According to the standard data list, analyze the target behavior feature to obtain the standard deviation distribution of the target behavior feature and the standard data list; wherein the standard deviation distribution is used to describe the deviation degree between the standard data feature of the classification number corresponding to each data filling axis of the target behavior feature and the standard data list; ​ According to the extended data list, the target behavior feature is analyzed and processed to obtain an extended deviation distribution of the target behavior feature and the extended data list; wherein the extended deviation distribution is used to describe the deviation degree between the target behavior feature and various data features of the extended data list; According to the additional data list, the target behavior feature is analyzed and processed to obtain an additional deviation distribution of the target behavior feature and the additional data list; wherein the additional deviation distribution is used to describe the deviation degree between the target behavior feature and various data features of the additional data list; The standard deviation distribution, the extended deviation distribution and the additional deviation distribution are collectively used as the list position information of the target behavior data. Based on the list position information of the target behavior data, the target behavior data is analyzed and processed in terms of data features to obtain a behavior analysis label of the target behavior data, and the step includes:

6. The big data based user behavior analysis method of claim 5, wherein, According to the standard deviation distribution of the list position information of the target behavior data, if the analysis result of the standard deviation distribution shows that the deviation degree of the target behavior feature of the target behavior data and the standard data feature corresponding to the classification number of one data filling axis of the standard data list meets a preset standard, the classification number and the standard data feature of the classification number are used as the predicted feature of the target behavior data; According to the extended deviation distribution of the list position information of the target behavior data, if the analysis result of the extended deviation distribution shows that the deviation degree of the target behavior feature of the target behavior data and one data feature of one data filling axis of the extended data list meets a preset standard, the data feature is used as the predicted feature of the target behavior data; According to the additional deviation distribution of the list position information of the target behavior data, if the analysis result of the additional deviation distribution shows that the deviation degree of the target behavior feature of the target behavior data and one data feature of one data filling axis of the additional data list meets a preset standard, the data feature is used as the predicted feature of the target behavior data; According to the predicted feature of the target behavior data, the target behavior data is processed for behavior prediction to obtain behavior prediction information of the target behavior data; According to the predicted feature of the target behavior data, the target behavior data is processed for user image tracing to obtain user image data of the target behavior data; The behavior prediction information and the user image data are collectively used as the behavior analysis label of the target behavior data. A user behavior analysis method based on big data according to any one of claims 1-6, comprising: 7.A big data based user behavior analysis system, characterized in that, a list induction module, configured to continuously collect user behavior data based on a plurality of data terminals with user labels, and to process the collected first user behavior data according to the corresponding user labels to obtain a standard data list; ​ The data integration module is used for continuously collecting user behavior data based on a plurality of data terminals without user labels, and integrating the collected second user behavior data to obtain a mixed database; The data analysis module is used for performing data classification processing on the mixed database according to the standard data list, so as to convert the mixed database into an extended data list corresponding to the standard data list and an additional data list not corresponding to the standard data list; The list analysis module is used for obtaining target behavior data, and performing feature analysis processing on the target behavior data according to the standard data list, the extended data list and the additional data list to obtain list position information of the target behavior data; The behavior analysis module is used for performing data feature analysis processing on the target behavior data based on the list position information of the target behavior data to obtain a behavior analysis label of the target behavior data.

Images