CN105824811A - Big data analysis method and device - Google Patents

Abstract

The embodiment of the present invention discloses a big data analysis method, including: based on the input first set of data and second set of data, obtaining at least two feature information satisfying preset conditions; the first set of data and the second set of data The data are all data in the first communication network; the first set of data satisfies a first preset rule; the second set of data satisfies a second preset rule; according to the at least two feature information, the first One set of data and the second set of data are analyzed to determine the first type of rules and the second type of rules; according to the first type of rules and the second type of rules, it is determined in the input third set of data Target data of a preset rule; the third group of data is data in other communication networks except the first communication network. The embodiment of the invention also discloses a big data analysis device.

Description

A big data analysis method and device thereof

technical field

The invention relates to communication technology, in particular to a big data analysis method and a device thereof.

Background technique

With the commercialization of the 4th generation mobile communication technology (4G, the4Generation mobile communication technology), the competition among major operators has become increasingly fierce; the return of high-value users on different networks and the penetration of 4G terminals play an important role in the development of mobile operators; therefore The identification of high-value users of different networks is very important.

At present, there are methods in the industry to analyze and model user behavior to determine user attributes. However, the existing methods generally focus on counting the number of users on different networks, not on the identification of users on different networks. identification of the terminal type.

Contents of the invention

In order to solve the existing technical problems, the embodiment of the present invention provides a big data analysis method and its device, which can determine the target data satisfying the preset rules in the data of different networks according to the data rules of the local network.

The technical solution of the embodiment of the present invention is achieved in this way: the embodiment of the present invention provides a big data analysis method, the method comprising:

Based on the input first set of data and second set of data, at least two feature information satisfying preset conditions are acquired; the first set of data and the second set of data are both data in the first communication network; the first set of The set of data satisfies a first preset rule; the second set of data satisfies a second preset rule;

Analyzing the first set of data and the second set of data according to the at least two feature information to determine the first type of rules and the second type of rules;

According to the first type of rule and the second type of rule, determine the target data that satisfies the first preset rule in the input third set of data; the third set of data is other than the first communication network data in other communication networks.

In the above solution, the first set of data and the second set of data are analyzed according to the at least two feature information, and the first type of rules and the second type of rules are determined, including:

Using a logistic regression algorithm to analyze the first set of data and the second set of data according to the at least two feature information to determine the first type of rules;

A decision tree algorithm is used to analyze the first set of data and the second set of data according to the at least two feature information to determine a second type of rule.

In the above solution, the decision tree algorithm is used to analyze the first set of data and the second set of data according to the at least two feature information, and determine the second type of rules, including:

Using a decision tree algorithm to analyze the first set of data and the second set of data according to the at least two feature information, and determine N rules; the N is a positive integer greater than or equal to 2;

Among the N rules, a second-type rule that satisfies the third preset rule is determined.

In the above solution, according to the first type of rules and the second type of rules, determining the target data that satisfies the first preset rule from the input third set of data includes:

Analyzing the input third set of data according to the first type of rules and the second type of rules respectively, to obtain the first suspected target data and the second suspected target data;

Target data satisfying the first preset rule is determined based on the first suspected target data and the second suspected target data.

In the above solution, the second type of rules includes: the first type of sub-rules; the first type of sub-rules satisfy the first preset rule;

Correspondingly, according to the first type of rules and the second type of rules, the input third set of data is analyzed to obtain the first suspected target data and the second suspected target data, including:

Analyzing the input third set of data according to the first type of rules to obtain the first suspected target data;

According to the first type of sub-rules, the input third set of data is analyzed to obtain the second suspected target data.

In the above solution, the second type of rule further includes: a second type of sub-rule; the second type of sub-rule satisfies a second preset rule; the method further includes:

Analyzing the first suspected target data and the second suspected target data according to the second type of sub-rules to obtain suspected non-target data;

Correspondingly, the determining the target data based on the first suspected target data and the second suspected target data includes:

The target data is determined based on the first suspected target data, the second suspected target data and the suspected non-target data.

The embodiment of the present invention also provides a big data analysis device, the device includes:

An acquisition unit, configured to acquire at least two feature information satisfying preset conditions based on the input first set of data and second set of data; the first set of data and the second set of data are both data in the first communication network ; The first set of data satisfies a first preset rule; the second set of data satisfies a second preset rule;

An analysis unit, configured to analyze the first set of data and the second set of data according to the at least two characteristic information, and determine the first type of rules and the second type of rules;

A determining unit, configured to determine, from the input third set of data, target data satisfying the first preset rule according to the first type of rule and the second type of rule; the third set of data is Data in a communication network other than the first communication network.

In the above scheme, the analysis unit includes:

The first analysis subunit is configured to use a logistic regression algorithm to analyze the first set of data and the second set of data according to the at least two feature information to determine a first type of rule;

The second analysis subunit is configured to use a decision tree algorithm to analyze the first set of data and the second set of data according to the at least two feature information to determine a second type of rule.

In the above solution, the second analysis subunit is further configured to use a decision tree algorithm to analyze the first set of data and the second set of data according to the at least two feature information, and determine N Rules; said N is a positive integer greater than or equal to 2;

It is also used to determine, among the N rules, a second type of rule that satisfies the third preset rule.

In the above scheme, the determination unit includes:

The first determination subunit is configured to analyze the input third set of data according to the first type of rules and the second type of rules respectively, to obtain the first suspected target data and the second suspected target data;

The second determining subunit is configured to determine target data satisfying the first preset rule based on the first suspected target data and the second suspected target data.

In the above solution, the second type of rules includes: the first type of sub-rules; the first type of sub-rules satisfy the first preset rule; correspondingly,

The first determining subunit is further configured to analyze the input third set of data according to the first type of rules to obtain the first suspected target data;

It is also used to analyze the input third set of data according to the first type of sub-rules to obtain the second suspected target data.

In the above solution, the second type of rule further includes: a second type of sub-rule; the second type of sub-rule satisfies the second preset rule;

The first determination subunit is further configured to analyze the first suspected target data and the second suspected target data according to the second type of sub-rules to obtain suspected non-target data;

Correspondingly, the second determination subunit is further configured to determine the target data based on the first suspected target data, the second suspected target data and the suspected non-target data.

The big data analysis method and its device provided by the embodiments of the present invention can determine at least two feature information in the first set of data and the second set of data in the first communication network, and use two different algorithms, based on the At least two pieces of feature information determine the first type of rules and the second type of rules for different algorithms, so that, through the first type of rules and the second type of rules, other communication networks except the first communication network Analyze the third group of data in the third group of data to determine the target data that satisfies the preset rules in the third group of data. The purpose of the target data for preset rules.

Description of drawings

Fig. 1 is a schematic diagram of the implementation flow of the big data analysis method of the embodiment of the present invention;

FIG. 2 is a schematic structural diagram of a big data analysis device according to an embodiment of the present invention;

Fig. 3 is the specific structural schematic diagram of the analyzing unit of the embodiment of the present invention;

FIG. 4 is a schematic structural diagram of a determination unit according to an embodiment of the present invention;

FIG. 5 is a schematic flowchart of a specific implementation of a big data analysis method according to an embodiment of the present invention.

detailed description

In order to understand the characteristics and technical content of the present invention in more detail, the implementation of the present invention will be described in detail below in conjunction with the accompanying drawings. The attached drawings are only for reference and description, and are not intended to limit the present invention.

Embodiment one

Fig. 1 is a schematic diagram of the implementation process of the big data analysis method of the embodiment of the present invention; as shown in Fig. 1, the method includes:

Step 101: Based on the input first set of data and second set of data, obtain at least two feature information satisfying preset conditions; the first set of data and the second set of data are both data in the first communication network; The first set of data satisfies a first preset rule; the second set of data satisfies a second preset rule;

In this embodiment, the first preset rule may be a rule that the communication device type of the user corresponding to the data in the first communication network belongs to the first type; the second preset rule may be a rule in the first communication network The communication device type of the user corresponding to the data does not belong to the first type of rule; thus, in the first communication network, the communication device types corresponding to the first group of data are all of the first type; the second group The type of communication equipment corresponding to the data is not the first type; since the characteristic rules of the data corresponding to different types of communication equipment are different, by analyzing the respective characteristic rules of the first set of data and the second set of data, it can be determined that M pieces of feature information that meet the preset conditions; analyze the data based on the M pieces of feature information, and estimate the characteristics such as the communication device type corresponding to the data; based on the above process, the embodiment of the present invention can be based on the first communication network The feature information in is determined from a large amount of data in different networks to determine that the communication device type belongs to the first type of data, which lays the foundation for big data analysis; here, the M is a positive integer greater than or equal to 2.

In this embodiment, the feature information is specifically a key variable index that meets the preset conditions, and different algorithms are used to analyze the big data in the first communication network through the key variable index, that is, the first group of data and the second group of data The second set of data is analyzed, thus laying the foundation for determining the rules in the big data of the first communication network.

In this embodiment, the preset conditions include, but are not limited to: a condition that the number of users is greater than or equal to the first one, a condition that the communication device type of the communication object is the first type, and the like.

Step 102: Analyze the first set of data and the second set of data according to the at least two characteristic information, and determine the first type of rules and the second type of rules;

In this embodiment, according to at least two characteristic information determined in the first communication network, different algorithms are used to analyze the first group of data and the second group of data, and then determine the information based on the first communication network. The first type of rules and the second type of rules.

In practical applications, different algorithms are usually used for data analysis of big data, so as to improve the accuracy of the analysis results; Two sets of data were analyzed.

In the above solution, the first set of data and the second set of data are analyzed according to the at least two feature information, and the first type of rules and the second type of rules are determined, including:

Using a logistic regression algorithm to analyze the first set of data and the second set of data according to the at least two feature information to determine the first type of rules;

A decision tree algorithm is used to analyze the first set of data and the second set of data according to the at least two feature information to determine a second type of rule.

In the above solution, the decision tree algorithm is used to analyze the first set of data and the second set of data according to the at least two feature information, and determine the second type of rules, including:

Using a decision tree algorithm to analyze the first set of data and the second set of data according to the at least two feature information, and determine N rules; the N is a positive integer greater than or equal to 2;

Among the N rules, a second-type rule that satisfies the third preset rule is determined.

In this embodiment, since the number of feature information determined in step 101 is different, the number of rules determined by the decision tree algorithm is different, that is, N is different; therefore, the value of N is limited by the feature information the number of .

In this embodiment, the second type of rules is a general term, which is a general term for all rules satisfying the third preset rule among the N rules, and therefore, does not refer to a specific rule.

Step 103: According to the first type of rule and the second type of rule, determine the target data that satisfies the first preset rule from the input third set of data; the third set of data is Data in communication networks other than communication networks.

In this embodiment, by using the first-type rules and the second-type rules determined in the first communication network, it can be determined among a large amount of data in other communication networks other than the first communication network that satisfies the first type of rules. The target data of a preset rule, that is, in the data of other communication networks, it is determined that the type of the user’s communication device belongs to the first type of target data, so that based on the data rules in this network, it is determined in the data of other networks that satisfy The purpose of the target data for preset rules.

In the above solution, according to the first type of rules and the second type of rules, determining the target data that satisfies the first preset rule from the input third set of data includes:

Analyzing the input third set of data according to the first type of rules and the second type of rules respectively, to obtain the first suspected target data and the second suspected target data;

Target data satisfying the first preset rule is determined based on the first suspected target data and the second suspected target data.

In this embodiment, the first suspected target data is data corresponding to the first type of rule, that is, through the first type of rule, it is determined in a communication network other than the first communication network that satisfies the first type of rule. Suspected target data of preset rules; the second suspected target data is data corresponding to the second type of rule, that is, determined in other communication networks except the first communication network through the second type of rule Suspected target data satisfying the first preset rule.

In the above solution, the second type of rules includes: the first type of sub-rules; the first type of sub-rules satisfy the first preset rule;

Correspondingly, according to the first type of rules and the second type of rules, the input third set of data is analyzed to obtain the first suspected target data and the second suspected target data, including:

Analyzing the input third set of data according to the first type of rules to obtain the first suspected target data;

According to the first type of sub-rules, the input third set of data is analyzed to obtain the second suspected target data.

In this embodiment, since the second type of rule is a rule determined by using a decision tree algorithm, the second suspected target data that meets the first preset rule can be determined through the second type of rule, and the second suspected target data that satisfies the second preset rule can be determined through the second type of rule. Set the suspected non-target data of the rule; that is, the second type of rule includes: the first type of sub-rule and the second type of sub-rule; through the first type of sub-rule, it is possible to determine the first Two suspected target data; through the second type of sub-rules, the suspected non-target data that meets the second preset rule can be determined; therefore, this embodiment also needs to be removed from the first suspected target data and the second suspected target data Suspected non-target data to determine the final target data.

In this embodiment, the first type of sub-rule is a rule that satisfies the first preset rule; the second type of sub-rule is a rule that does not satisfy the first preset rule; that is, it is a rule that satisfies the second A rule of a preset rule; when the second type of sub-rule is a rule that does not satisfy the first preset rule, the suspected non-target data is a type of interference data; therefore, the suspected non-target data can also be called noise data.

In the above solution, the second type of rule further includes: a second type of sub-rule; the second type of sub-rule satisfies a second preset rule; the method further includes:

Analyzing the first suspected target data and the second suspected target data according to the second type of sub-rules to obtain suspected non-target data;

Correspondingly, the determining the target data based on the first suspected target data and the second suspected target data includes:

The target data is determined based on the first suspected target data, the second suspected target data and the suspected non-target data.

In order to implement the above method, an embodiment of the present invention also provides a big data analysis device, as shown in Figure 2, the device includes:

An acquisition unit 21, configured to acquire at least two feature information satisfying preset conditions based on the input first set of data and second set of data; both the first set of data and the second set of data are in the first communication network data; the first set of data satisfies a first preset rule; the second set of data satisfies a second preset rule;

An analyzing unit 22, configured to analyze the first set of data and the second set of data according to the at least two characteristic information, and determine the first type of rules and the second type of rules;

The determining unit 23 is configured to determine, from the input third set of data, target data satisfying the first preset rule according to the first type of rule and the second type of rule; the third set of data is Data in other communication networks other than the first communication network.

In the above solution, as shown in Figure 3, the analysis unit 22 includes:

The first analysis subunit 221 is configured to use a logistic regression algorithm to analyze the first set of data and the second set of data according to the at least two feature information to determine a first type of rule;

The second analysis subunit 222 is configured to use a decision tree algorithm to analyze the first set of data and the second set of data according to the at least two feature information to determine a second type of rule.

In the above solution, the second analysis subunit 222 is also configured to use a decision tree algorithm to analyze the first set of data and the second set of data according to the at least two feature information, and determine N rules; said N is a positive integer greater than or equal to 2;

It is also used to determine, among the N rules, a second type of rule that satisfies the third preset rule.

In the above solution, as shown in FIG. 4, the determining unit 23 includes:

The first determination subunit 231 is configured to analyze the input third set of data according to the first type of rules and the second type of rules respectively, to obtain the first suspected target data and the second suspected target data;

The second determining subunit 232 is configured to determine target data satisfying the first preset rule based on the first suspected target data and the second suspected target data.

In the above solution, the second type of rules includes: the first type of sub-rules; the first type of sub-rules satisfy the first preset rule; correspondingly,

The first determination subunit 231 is further configured to analyze the input third set of data according to the first type of rules to obtain the first suspected target data;

It is also used to analyze the input third set of data according to the first type of sub-rules to obtain the second suspected target data.

In the above solution, the second type of rule further includes: a second type of sub-rule; the second type of sub-rule satisfies the second preset rule;

The first determination subunit 231 is further configured to analyze the first suspected target data and the second suspected target data according to the second type of sub-rules to obtain suspected non-target data;

Correspondingly, the second determination subunit 232 is further configured to determine the target data based on the first suspected target data, the second suspected target data and the suspected non-target data.

The acquisition unit 21, the analysis unit 22 and the determination unit 23 all can run on a computer, and can be located on a central processing unit (CPU), or a microprocessor (MPU), or a digital signal processor (DSP), or Programmable Gate Array (FPGA) implementation.

Embodiment two

The first software, such as IMESSAGE software, refers to the built-in software for sending short messages between users of the first type of terminal. This software can make short messages directly send from the GPRS end, saving the cost of short messages for users who use the first type of terminal; therefore, using the first type of terminal The first type of terminal users of a software may greatly reduce the usage of short messages, forming a short message black hole phenomenon. This embodiment is based on the above short message black hole phenomenon, and determines the terminal type as the first type of users in the different network.

This embodiment mainly uses the communication data of the existing sub-systems to analyze the communication behavior of the first type of terminal users using the first software on this network and the characteristics of the people in their communication circle, and identify that the different network has the above communication behavior, and The data of the people in the communication circle conforming to the above characteristics, that is, users, can finally determine the terminal type as the first type of users in the different network, so as to help the return work and marketing strategy of the high-value customers of the different network of the operator.

Specifically, this embodiment is mainly based on the user communication circle model, and by analyzing the habitual characteristics of the customer's voice communication circle and SMS communication circle using the first software in the first type of terminal on this network, among a large number of users in different networks, it is analyzed that The user group of the first type of end users, and then analyze the probability of whether a certain user of the different network is the first type of end users, so as to provide operators with data information with reference value.

Fig. 5 is a schematic flow chart of a specific implementation of the big data analysis method according to the embodiment of the present invention; before performing big data analysis, it is necessary to determine the first set of data and the second set of data; specifically, it is determined in the first communication network that the The first group of data with the first data amount, and the second group of data with the first data amount; wherein, the user equipment type corresponding to each data in the first group of data is the first type; the second group of data corresponds to The type of user equipment is not the first type; as shown in Figure 5, the method includes:

Step 501: In the first set of data and the second set of data, combine the feature rules of the user's social circle corresponding to the first set of data and the second set of data, the feature rules of the voice and text messages in the social circle, and whether the contact object uses M pieces of feature information are selected from the feature rules of the first type of terminal; wherein, M is a positive integer greater than or equal to 2;

Here, the characteristic information is also referred to as a key variable index.

Step 502: Using a logistic regression algorithm to analyze the first set of data and the second set of data according to the M pieces of feature information, and simulate a first type of rule that satisfies the first preset rule;

Here, the first type of rule may be a logistic regression formula; the first preset rule is a rule that the user terminal type is the first type.

In this embodiment, the analysis of the first set of data and the second set of data to simulate a first type of rule that satisfies the first preset rule includes:

Based on the M pieces of feature information, a logistic regression algorithm is used to analyze the first set of data and the second set of data, and simulate a first type of rule that satisfies the first preset rule.

Step 503: Determine the third group of data, and calculate the probability of each data in the third group of data according to the first type of rules, so as to determine the first suspected target data; the third group of data is related to the Data corresponding to users in other communication networks that are communicated by users in the first communication network;

Here, the calculation of the probability of each data in the third group of data according to the first type of rules to determine the first suspected target data further includes:

calculating the probability of each data in the third group of data according to the first type of rules;

According to the data service requirements and the preset number of users corresponding to the logistic regression level of the logistic regression algorithm, among the probabilities corresponding to each data in the third group of data, determine the data whose probability is greater than or equal to the preset threshold, and set the probability greater than or equal to the preset threshold. The data equal to the preset threshold is taken as the first suspected target data.

Step 504: Using the C5 decision tree algorithm to analyze the first set of data and the second set of data according to the M feature information, and determine m1 rules A and m2 rules B;

Step 505: According to the number of users and confidence levels corresponding to rule A and rule B, filter rule A and rule B to determine the first type of sub-rule in rule A, and determine the second type of sub-rule in rule B class subrules;

Here, the first type of sub-rule satisfies the first preset rule; the second type of sub-rule satisfies the second preset rule; the m1 and m2 are positive integers greater than or equal to 1.

Specifically, when the number of users of the first set of data and the number of users of the second set of data are both 10W, a rule with a confidence degree greater than 85% and a number of users greater than 2W is selected from rule A, and is determined as the first type of sub-rule; In B, the rules with confidence greater than 90% and the number of users greater than 1.8W are selected and determined as the second type of sub-rules;

In this embodiment, both the first type of sub-rules and the second type of sub-rules belong to the second type of rules.

Step 506: Analyze the third group of data according to the first type of sub-rules, and determine the second suspected target data;

Step 507: Determine the intersection data of the first suspected target data and the second suspected target data as the third suspected target data;

Step 508: Eliminate the data conforming to the second type of sub-rules in the third suspected target data, and use the remaining third suspected target data as target data.

In the embodiment of the present invention, it is possible to determine key variable indicators, that is, feature information, from the first set of data and the second set of data in the first communication network; Analyze with the second set of data, determine the first type of rules corresponding to the logistic regression algorithm, and the second type of rules corresponding to the decision tree algorithm; wherein, the second type of rules include the first type of rules rules and second-type sub-rules; then, analyze the third group of data in the different network according to the first-type rules and the first-type sub-rules respectively, and determine the first suspected target data and the second suspected target data; Since the first type of rule satisfies the first preset rule; the first type of sub-rule also satisfies the first preset rule; and the second type of sub-rule satisfies the second preset rule, therefore, the After the third suspected target data is determined by the intersection of the first suspected target data and the second suspected target data, the data satisfying the second type of sub-rule is eliminated from the third suspected target data, that is, in the third suspected target data Suspected non-target data is removed from the data to finally obtain target data. The target data is the target data that satisfies the first preset rule determined from the data of other networks according to the data rules of the local network.

Those skilled in the art should understand that the embodiments of the present invention may be provided as methods, systems, or computer program products. Accordingly, the present invention can take the form of a hardware embodiment, a software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each procedure and/or block in the flowchart and/or block diagram, and a combination of procedures and/or blocks in the flowchart and/or block diagram can be realized by computer program instructions. These computer program instructions may be provided to a general purpose computer, special purpose computer, embedded processor, or processor of other programmable data processing equipment to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing equipment produce a An apparatus for realizing the functions specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to operate in a specific manner, such that the instructions stored in the computer-readable memory produce an article of manufacture comprising instruction means, the instructions The device realizes the function specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions can also be loaded onto a computer or other programmable data processing device, causing a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby The instructions provide steps for implementing the functions specified in the flow chart or blocks of the flowchart and/or the block or blocks of the block diagrams.

The above is only the implementation of the embodiment of the present invention. It should be pointed out that for those skilled in the art, without departing from the principle of the embodiment of the present invention, some improvements and modifications can also be made. These improvements and Retouching should also be regarded as the scope of protection of the embodiments of the present invention.

Claims (12)

1. a big data analysis method, is characterized in that, described method comprises: Based on the input first set of data and second set of data, at least two feature information satisfying preset conditions are acquired; the first set of data and the second set of data are both data in the first communication network; the first set of The set of data satisfies a first preset rule; the second set of data satisfies a second preset rule; Analyzing the first set of data and the second set of data according to the at least two feature information to determine the first type of rules and the second type of rules; According to the first type of rule and the second type of rule, determine the target data that satisfies the first preset rule in the input third set of data; the third set of data is other than the first communication network data in other communication networks. 2. The method according to claim 1, characterized in that, according to the at least two feature information, the first set of data and the second set of data are analyzed to determine the first type of rules and the second set of rules. class rules, including: Using a logistic regression algorithm to analyze the first set of data and the second set of data according to the at least two feature information to determine the first type of rules; A decision tree algorithm is used to analyze the first set of data and the second set of data according to the at least two feature information to determine a second type of rule. 3. The method according to claim 2, wherein the decision tree algorithm is used to analyze the first group of data and the second group of data according to the at least two feature information, and determine The second category of rules includes: Using a decision tree algorithm to analyze the first set of data and the second set of data according to the at least two feature information, and determine N rules; the N is a positive integer greater than or equal to 2; Among the N rules, a second-type rule that satisfies the third preset rule is determined. 4. The method according to any one of claims 1 or 3, characterized in that, according to the first type of rules and the second type of rules, it is determined in the input third group of data that the first type of data is satisfied. Target data for preset rules, including: Analyzing the input third set of data according to the first type of rules and the second type of rules respectively, to obtain the first suspected target data and the second suspected target data; Target data satisfying the first preset rule is determined based on the first suspected target data and the second suspected target data. 5. The method according to claim 4, wherein the second type of rule comprises: a first type of sub-rule; the first type of sub-rule satisfies the first preset rule; Correspondingly, according to the first type of rules and the second type of rules, the input third set of data is analyzed to obtain the first suspected target data and the second suspected target data, including: Analyzing the input third set of data according to the first type of rules to obtain the first suspected target data; According to the first type of sub-rules, the input third set of data is analyzed to obtain the second suspected target data. 6. The method according to claim 5, wherein the second type of rule further comprises: a second type of sub-rule; the second type of sub-rule satisfies a second preset rule; the method further comprises: Analyzing the first suspected target data and the second suspected target data according to the second type of sub-rules to obtain suspected non-target data; Correspondingly, the determining the target data based on the first suspected target data and the second suspected target data includes: The target data is determined based on the first suspected target data, the second suspected target data and the suspected non-target data. 7. A big data analysis device, characterized in that the device comprises: An acquisition unit, configured to acquire at least two feature information satisfying preset conditions based on the input first set of data and second set of data; the first set of data and the second set of data are both data in the first communication network ; The first set of data satisfies a first preset rule; the second set of data satisfies a second preset rule; An analysis unit, configured to analyze the first set of data and the second set of data according to the at least two characteristic information, and determine the first type of rules and the second type of rules; A determining unit, configured to determine, from the input third set of data, target data satisfying the first preset rule according to the first type of rule and the second type of rule; the third set of data is Data in a communication network other than the first communication network. 8. The device according to claim 7, wherein the analysis unit comprises: The first analysis subunit is configured to use a logistic regression algorithm to analyze the first set of data and the second set of data according to the at least two feature information to determine a first type of rule; The second analysis subunit is configured to use a decision tree algorithm to analyze the first set of data and the second set of data according to the at least two feature information to determine a second type of rule. 9. The device according to claim 8, wherein the second analysis subunit is further configured to use a decision tree algorithm to analyze the first group of data and the The second set of data is analyzed to determine N rules; said N is a positive integer greater than or equal to 2; It is also used to determine, among the N rules, a second type of rule that satisfies the third preset rule. 10. The device according to any one of claims 7 to 9, wherein the determining unit includes: The first determination subunit is configured to analyze the input third set of data according to the first type of rules and the second type of rules respectively, to obtain the first suspected target data and the second suspected target data; The second determining subunit is configured to determine target data satisfying the first preset rule based on the first suspected target data and the second suspected target data. 11. The device according to claim 10, wherein the second type of rule comprises: a first type of sub-rule; the first type of sub-rule satisfies the first preset rule; correspondingly, The first determining subunit is further configured to analyze the input third set of data according to the first type of rules to obtain the first suspected target data; It is also used to analyze the input third set of data according to the first type of sub-rules to obtain the second suspected target data. 12. The method according to claim 11, wherein the second type of rule further comprises: a second type of sub-rule; the second type of sub-rule satisfies a second preset rule; The first determination subunit is further configured to analyze the first suspected target data and the second suspected target data according to the second type of sub-rules to obtain suspected non-target data; Correspondingly, the second determination subunit is further configured to determine the target data based on the first suspected target data, the second suspected target data and the suspected non-target data.