WO2016054908A1

WO2016054908A1 - Internet of things big data platform-based intelligent user profiling method and apparatus

Info

Publication number: WO2016054908A1
Application number: PCT/CN2015/077320
Authority: WO
Inventors: 杨桂荣
Original assignee: 中兴通讯股份有限公司
Priority date: 2014-10-10
Filing date: 2015-04-23
Publication date: 2016-04-14
Also published as: CN105574730A

Abstract

Disclosed are an Internet of Things big data platform-based intelligent user profiling method and apparatus. The method comprises: collecting user data from an Internet of Things information center; processing the collected user data to generate a bill data file; and performing data mining on the bill data file by using a big data platform, and depicting a user profile. By means of the present invention, data is acquired according to an RFID tag and a sensor network by an Internet of Things information center, data is mined by using a big data platform system, and a user profile is depicted according to indexes such as attributes and goods behaviors of a user, so that the accuracy and efficiency of data analysis on user profiling are improved, and sufficient resources are further provided for performing consumer group analysis, function planning and adjustment and marketing strategy making on the user.

Description

Intelligent user portrait method and device based on Internet of Things big data platform

Technical field

The invention relates to the field of data mining technology, in particular to a smart user portrait method and device based on the Internet of Things big data platform.

Background technique

User portraits are an effective tool for sketching target users, contacting users' appeals and design directions. User portrait is a kind of carrier that can combine qualitative and quantitative methods. Through quantitative preliminary research, we can get a more accurate understanding of the user group, and it can be good for users in the later establishment of user roles. Prioritize the rankings to highlight core, large-scale users.

At present, the data source of user portraits is relatively single and the classification is not clear, resulting in low efficiency and accuracy of data analysis and mining.

Summary of the invention

The embodiment of the invention provides a smart user portrait method and device based on the Internet of Things big data platform, aiming at at least improving the accuracy and efficiency of data analysis of the user portrait, and further providing sufficient analysis for the consumer group analysis and marketing strategy formulation. Resources.

In order to achieve the above object, an embodiment of the present invention provides a smart user portrait method based on the Internet of Things big data platform, including:

Collect user data from the IoT Information Center;

Processing the collected user data to generate a bill data file;

The big data platform is used to perform data mining on the bill data file to describe the user portrait.

The step of processing the collected user data to generate the bill data file includes:

Establishing a bill conversion table for the collected user data;

The user data is processed based on the bill conversion table to form a bill, and the bill is set to describe the user's use behavior of the item, and the main dimensions include time, place, user, and item type;

The bill is generated into a bill data file.

The data mining is performed on the bill data file by using a big data platform, and the steps of describing the user image include:

Store the bill data file to the big data platform;

On the big data platform, using the first MR model, all the bill data files of a single type of bill of a single user are merged into one user-related data block, and the individual analysis of the data mining is performed to obtain individual analysis parameters;

Using the second MR model, calculating global parameters of the bill data file;

Using the third MR model, final data mining is performed according to the individual analysis parameters and global parameters, the user portrait is described according to the mining result, and the user tag is output to the database.

The first MR model is used to combine all the bill data files of a single user's bills into one user-related data block, and perform individual analysis of the data mining, and the steps of obtaining individual analysis parameters include:

Using the first MR model, all the bill data files of a single type of bill of a single user are combined in multiple dimensions to generate a user-related data block;

Saving the data block as a binary new record to the database;

Based on the merged data block, user attributes are extracted, and the behavior characteristics of the items are counted, and various types of tags related to the individual are generated, and individual analysis parameters are obtained.

The method also includes:

According to the user's portrait, the user is analyzed for needs and strategic planning.

The user data is obtained by the Internet of Things information center based on the RFID tag and the sensor network.

The embodiment of the invention also provides a smart user portrait device based on the Internet of Things big data platform, comprising:

The acquisition module is configured to collect user data from the Internet of Things information center;

The processing module is configured to process the collected user data to generate a bill data file;

The mining module is configured to perform data mining on the bill data file by using a big data platform to describe a user image.

The processing module is further configured to: establish a bill conversion table for the collected user data; process the user data according to the bill conversion table to form a bill, and the bill is set to describe the user's use behavior of the item, The main dimensions include usage time, location, user, and item type; the bill is generated into the bill data file.

The mining module is further configured to store the bill data file to the big data platform; on the big data platform, use the first MR model to merge all the bill data files of a single type of bill of the single user into one User-related data blocks, and individual analysis of data mining, to obtain individual analysis parameters; using the second MR model to calculate global parameters of the bill data file; using the third MR model, according to the individual analysis parameters and global parameters, Perform final data mining, describe user images based on mining results, and output user tags to the database.

The mining module is further configured to use a first MR model to combine all the bill data files of a type of bill of a single user in multiple dimensions to generate a user-related data block; The new binary records are saved to the database; based on the merged data blocks, the user attributes are extracted, the behavior characteristics of the items are counted, various types of tags related to the individual are generated, and individual analysis parameters are obtained.

The device also includes:

The planning module is set to perform demand analysis and strategic planning for the user according to the user portrait.

An intelligent user portrait method and device based on the Internet of Things big data platform proposed by the embodiment of the present invention, the Internet of Things information center acquires data according to the RFID tag and the sensor network, and the data is mined by the big data platform system according to the attribute of the user. The object behavior and other indicators are used to describe the user's portrait, thereby improving the accuracy and efficiency of the data analysis of the user's portrait, and providing sufficient resources for the user to analyze the consumer group, adjust the function plan, and formulate the marketing strategy.

DRAWINGS

1 is a schematic flow chart of an embodiment of a smart user portrait method based on the Internet of Things big data platform of the present invention;

2 is a schematic diagram of a data processing flow of a big data platform in an embodiment of the present invention;

3 is a schematic flow chart of another embodiment of a smart user portrait method based on the Internet of Things big data platform of the present invention;

4 is a schematic structural diagram of an embodiment of an intelligent user portrait device based on the Internet of Things big data platform of the present invention;

FIG. 5 is a schematic structural diagram of another embodiment of an intelligent user portrait device based on the Internet of Things big data platform of the present invention.

In order to make the technical solutions of the present invention clearer and clearer, the following will be further described in detail with reference to the accompanying drawings.

detailed description

The solution of the embodiment of the present invention is mainly: the Internet of Things information center acquires data according to the RFID tag and the sensor network, and the data is mined by the big data platform system, and the user portrait is described according to the attributes of the user, the behavior of the item, and the like, thereby improving The accuracy and efficiency of data analysis of user images provide users with sufficient resources for consumer group analysis, function planning adjustment and marketing strategy formulation.

As shown in FIG. 1 , an embodiment of the present invention provides a smart user portrait method based on the Internet of Things big data platform, including:

Step S101, collecting user data from the Internet of Things information center;

The basis of the user's portrait is to have a preliminary understanding of the massive users through the data. Generally, the user data extraction analysis and the questionnaire survey are used to determine the dimensional indicators of the statistical analysis according to the product target. Among them, the dimension of analysis can be comprehensively analyzed according to population attributes and product behavior attributes. Population attributes generally include: region, age, gender, culture, occupation, income, living habits, consumption habits, etc.; product behavior generally includes: product categories, frequency of activity, product preferences, product drivers, usage habits, product consumption, etc.

The solution of the embodiment mainly completes the method of converting the use of the item by the user into a description of the user's portrait, including collecting the information of the Internet of Things, storing the information of the Internet of Things, mining the data, and describing the user's portrait, so that the generation of the user's portrait can be based on the collection. Support and mining of big data from the Internet of Things, combined with quantitative and qualitative methods to create user portraits.

Specifically, first, original user data is extracted from the Internet of Things information center, and the original user data is acquired by the Internet of Things information center according to the RFID tag and the sensor network.

Based on cloud computing platforms and intelligent networks, the Internet of Things can capture and store data based on RFID tags and sensor networks.

Step S102, processing the collected user data to generate a bill data file;

Forming a bill conversion table for the collected user data; processing the user data based on the bill conversion table to form a bill, the bill being set to describe the user's use behavior of the item, and the main dimensions include the use time, location, User, item type, etc.

After that, the bill data file is generated according to the bill.

Multiple bill data files are generated for different types of bills every day.

Step S103: Perform data mining on the bill data file by using a big data platform to describe a user portrait.

The data file generated after processing is stored in the big data platform, and the data is mined by the big data platform system to obtain the distribution vector of different users on each subject, thereby describing according to the attributes of the user, the behavior of the item, the payment behavior and the like. User portraits make it easy to segment users, find user needs, segment target markets, adjust function planning, and marketing strategies.

Specifically, as shown in FIG. 2, on the big data platform, the big data platform Hadoop is used to merge the bill data files, and the MapReduce model is cached and built in the Reduce instance, the data is mined, and the user tags are output to the HBASE database.

Among them, the individual user image mining process is an analysis of the different types of CDRs associated with the user, and considers the distribution of the overall parameters of each type of CDR regardless of the individual.

Because of the different types of bills every day, multiple bill data files are generated. These CDR data files need to be merged according to the MapReduce model using the Hadoop system. In the merged file, all the original CDR data files of one user type CDR are merged into one data block, and this data block is used as a binary new file. The record is saved to the HBASE database.

The specific data mining process is as follows:

First, on the big data platform, using the first MR model, all the bill data files of a single type of bill of a single user are merged into one user-related data block, and the individual analysis of the data mining is performed to obtain individual analysis parameters. Corresponding to the binary type record in the MR model, in this way, the data of a single user is guaranteed to be on one data node.

In the merge, the first MR model is used to combine all the bill data files of a single type of bill of a single user in multiple dimensions, and the top item such as frequency or time is used as a label to describe the user. It mainly includes extracting target user attributes, counting product behavior characteristics, generating various types of labels related to individual users, and obtaining Individual analysis parameters, including content preferences, frequent use of items, etc., in order to segment users, find core users, segment target markets, and adjust functional planning and marketing strategies.

A user-related data block is generated by merging, and the data block is saved to the database as a binary new record.

Then, based on the combined bill data file, the user portrait is described according to the attributes of the user, the behavior of the item, and the payment behavior. The second MR model is used to calculate global parameters of the bill data file. The output is likely to be in a relational database depending on the access efficiency.

Finally, using the third MR model, the final data mining is performed based on the individual analysis parameters and global parameters obtained from the previous two MR models, the user images are described according to the mining results, and the user tags are output to the HBASE database. At this point, the description of the user's portrait is completed.

In the embodiment, the Internet of Things information center acquires data according to the RFID tag and the sensor network, and the data is mined by the big data platform system, and the user portrait is described according to the attributes of the user and the behavior of the item, thereby improving the user image. The accuracy and efficiency of data analysis provide users with sufficient resources for consumer group analysis, functional planning adjustment, and marketing strategy development.

As shown in FIG. 3, another embodiment of the present invention provides a smart user portrait method based on the Internet of Things big data platform. Based on the foregoing embodiment, the method further includes:

Step S104, performing demand analysis and strategy planning on the user according to the user portrait.

Compared with the foregoing embodiment, the embodiment further includes a solution for performing requirement analysis and policy planning on the user.

Specifically, after obtaining the created user portrait, a relatively accurate understanding of the user group is obtained, and the user can be subdivided based on the user portrait, and the user role can be well established in the later generation of the user role. User prioritization is sorted to highlight core, large-scale users. Therefore, it provides resources and basis for finding user needs, subdividing target markets, adjusting function planning and marketing strategies.

As shown in FIG. 4, an embodiment of the present invention provides an intelligent user portrait device based on the Internet of Things big data platform, comprising: an acquisition module 201, a processing module 202, and a mining module 203, wherein:

The collecting module 201 is configured to collect user data from the Internet of Things information center; the collecting module 201 is disposed between the access layer and the network layer. The sensing layer of the collection module 201 is mainly composed of an RFID and a sensor network, and the access layer mainly completes network access of various devices, such as a 3G/4G, Mesh network.

The processing module 202 is configured to process the collected user data to generate a bill data file.

The mining module 203 is configured to perform data mining on the bill data file by using a big data platform to describe a user image.

Specifically, the basis of the user's portrait is to have a preliminary understanding of the mass of users through the data, generally using user data extraction analysis and questionnaire research to determine the dimensional indicators of statistical analysis according to product goals. Among them, the dimension of analysis can be comprehensively analyzed according to population attributes and product behavior attributes. Population attributes generally include: region, age, gender, culture, occupation, income, living habits, consumption habits, etc.; product behavior generally includes: product categories, frequency of activity, product preferences, product drivers, usage habits, product consumption, etc.

After that, the collected user data is processed to generate a bill data file.

Specifically, the method includes: establishing a bill conversion table for the collected user data; processing the user data according to the bill conversion table to form a bill, and the bill is set to describe the user's use behavior of the item, and the main dimension includes the use time. , location, user, item type, etc. After that, the bill data file is generated according to the bill.

Multiple bill data files are generated for different types of bills every day.

Then, the CDR data file generated after processing is stored to the big data platform, and the data is mined by the big data platform system, and the distribution vectors of different users on each subject are obtained, thereby according to the attributes of the user, the behavior of the goods, the behavior of payment, and the like. To describe user portraits, to facilitate segmentation of users, to find user needs, to segment target markets, to adjust functional planning and marketing strategies.

Because of the different types of bills every day, multiple bill data files are generated. These CDR data files need to be merged according to the MepReduce model using the Hadoop system. In the merged file, all the original CDR data files of one user type CDR are merged into one data block, and this data block is used as a binary new file. The record is saved to the HBASE database.

The specific data mining process is as follows:

In the merge, the first MR model is used to combine all the bill data files of a single type of bill of a single user in multiple dimensions, and the top item such as frequency or time is used as a label to describe the user. It mainly includes extracting target user attributes, counting product behavior characteristics, generating various types of labels related to individual users, and obtaining individual analysis parameters, including content preferences and frequent use of items, so as to subdivide users, find core users, and subdivide targets. Market, adjustment function planning and marketing strategy.

As shown in FIG. 5, another embodiment of the present invention provides a smart user portrait device based on the Internet of Things big data platform. Based on the foregoing embodiment, the method further includes:

The planning module 204 is configured to perform demand analysis and strategic planning on the user according to the user portrait.

The embodiment of the present invention is based on the intelligent user portrait method and device of the Internet of Things big data platform. The Internet of Things information center acquires data according to the RFID tag and the sensor network, and the data is mined by the big data platform system according to the attributes of the user, the behavior of the item, etc. The indicators describe the user's portrait, which improves the accuracy and efficiency of the data analysis of the user's portrait, and provides sufficient resources for the consumer to analyze the consumer group, adjust the function plan and formulate the marketing strategy.

The above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the invention, and the equivalent structure or process changes made by the specification and the drawings of the present invention may be directly or indirectly applied to other related technical fields. The same is included in the scope of patent protection of the present invention.

Industrial applicability

As described above, the smart user portrait method and apparatus based on the Internet of Things big data platform provided by the embodiments of the present invention have the following beneficial effects: the Internet of Things information center acquires data according to the RFID tag and the sensor network, and the big data platform system The data is mined, and the user's portrait is described according to the attributes of the user, the behavior of the item, etc., thereby improving the accuracy and efficiency of data analysis of the user's portrait, and further providing sufficient analysis for the consumer group analysis, function planning adjustment, and marketing strategy formulation. Resources.

Claims

A smart user portrait method based on the Internet of Things big data platform, comprising:

Collect user data from the IoT Information Center;

Processing the collected user data to generate a bill data file;

The big data platform is used to perform data mining on the bill data file to describe the user portrait.
The method of claim 1, wherein the processing of the collected user data to generate a bill data file comprises:

Establishing a bill conversion table for the collected user data;

The user data is processed based on the bill conversion table to form a bill, and the bill is set to describe the user's use behavior of the item, and the main dimensions include time, place, user, and item type;

The bill is generated into a bill data file.
The method of claim 1, wherein the step of data mining the CDR data file by using a big data platform, the step of describing a user image comprises:

Store the bill data file to the big data platform;

On the big data platform, using the first MR model, all the bill data files of a single type of bill of a single user are merged into one user-related data block, and the individual analysis of the data mining is performed to obtain individual analysis parameters;

Using the second MR model, calculating global parameters of the bill data file;

Using the third MR model, final data mining is performed according to the individual analysis parameters and global parameters, the user portrait is described according to the mining result, and the user tag is output to the database.
The method according to claim 3, wherein the first MR model is used to merge all the bill data files of a single type of bill of a single user into one user-related data block, and perform individual analysis of data mining. The steps to obtain individual analysis parameters include:

Using the first MR model, all the bill data files of a single type of bill of a single user are combined in multiple dimensions to generate a user-related data block;

Saving the data block as a binary new record to the database;

Based on the merged data block, user attributes are extracted, and the behavior characteristics of the items are counted, and various types of tags related to the individual are generated, and individual analysis parameters are obtained.
The method of any one of claims 1 to 4, further comprising:

According to the user's portrait, the user is analyzed for needs and strategic planning.
The method of any of claims 1-4, wherein the user data is obtained by the IoT information center based on an RFID tag and a sensor network.
A smart user portrait device based on the Internet of Things big data platform, comprising:

The acquisition module is configured to collect user data from the Internet of Things information center;

The processing module is configured to process the collected user data to generate a bill data file;

The mining module is configured to perform data mining on the bill data file by using a big data platform to describe a user image.
The apparatus according to claim 7, wherein

The processing module is further configured to: establish a bill conversion table for the collected user data; process the user data according to the bill conversion table to form a bill, and the bill is set to describe the user's use behavior of the item, The main dimensions include usage time, location, user, and item type; the bill is generated into the bill data file.
The apparatus according to claim 7, wherein

The mining module is further configured to store the bill data file to the big data platform; on the big data platform, use the first MR model to merge all the bill data files of a single type of bill of the single user into one User-related data blocks, and individual analysis of data mining, to obtain individual analysis parameters; using the second MR model to calculate global parameters of the bill data file; using the third MR model, according to the individual analysis parameters and global parameters, Perform final data mining, describe user images based on mining results, and output user tags to the database.
The apparatus according to claim 9, wherein

The mining module is further configured to use a first MR model to combine all the bill data files of a type of bill of a single user in multiple dimensions to generate a user-related data block; The new binary records are saved to the database; based on the merged data blocks, the user attributes are extracted, the behavior characteristics of the items are counted, various types of tags related to the individual are generated, and individual analysis parameters are obtained.
The apparatus according to any one of claims 7 to 10, further comprising:

The planning module is set to perform demand analysis and strategic planning for the user according to the user portrait.