CN107844584A - Usage mining method, apparatus, electronic equipment and computer-readable recording medium - Google Patents
Usage mining method, apparatus, electronic equipment and computer-readable recording medium Download PDFInfo
- Publication number
- CN107844584A CN107844584A CN201711124056.8A CN201711124056A CN107844584A CN 107844584 A CN107844584 A CN 107844584A CN 201711124056 A CN201711124056 A CN 201711124056A CN 107844584 A CN107844584 A CN 107844584A
- Authority
- CN
- China
- Prior art keywords
- user
- candidate user
- data
- history
- training
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2462—Approximate or statistical queries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
- G06Q30/0202—Market predictions or forecasting for commercial activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0255—Targeted advertisements based on user history
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Finance (AREA)
- Physics & Mathematics (AREA)
- Development Economics (AREA)
- Accounting & Taxation (AREA)
- Theoretical Computer Science (AREA)
- Entrepreneurship & Innovation (AREA)
- General Physics & Mathematics (AREA)
- Probability & Statistics with Applications (AREA)
- Data Mining & Analysis (AREA)
- Game Theory and Decision Science (AREA)
- Economics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Databases & Information Systems (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Fuzzy Systems (AREA)
- General Engineering & Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the present disclosure discloses a kind of usage mining method, apparatus, electronic equipment and computer-readable recording medium, wherein, the usage mining method includes:Obtain the candidate user collection of service provider;Calculate the probable value that the candidate user that the candidate user is concentrated is performed generation pre-set user behavior after predetermined registration operation;The candidate user that the probable value is met to preparatory condition confirms as targeted customer.The disclosure more specific aim, success rate height, while also reduce and develop the cost that new user is spent in terms of new user is developed, promote user's order.The maximization of user satisfaction can be thus realized, the risk for losing user is reduced, gets more validated users.
Description
Technical field
This disclosure relates to technical field of information processing, and in particular to a kind of usage mining method, apparatus, electronic equipment and meter
Calculation machine readable storage medium storing program for executing.
Background technology
With the development of Internet technology, increasing businessman or service provider are promoted by internet channels
Products & services, and make every effort to strive for more user's orders on the basis of products & services are promoted, to lift existing resource
Utilization rate, it is that businessman or service provider create more values.Many businessmans or service provider's generally use at present
Attract user to place an order to the popular random form for performing excitation operation, but this mode lack of targeted, success rate it is low, it is necessary to
The cost of cost is higher.
The content of the invention
The embodiment of the present disclosure provides a kind of usage mining method, apparatus, electronic equipment and computer-readable recording medium.
In a first aspect, a kind of usage mining method is provided in the embodiment of the present disclosure.
Specifically, the usage mining method, including:
Obtain the candidate user collection of service provider;
Calculate the probability that the candidate user that the candidate user is concentrated is performed generation pre-set user behavior after predetermined registration operation
Value;
The candidate user that the probable value is met to preparatory condition confirms as targeted customer.
With reference in a first aspect, the disclosure in the first implementation of first aspect, it is described to obtain service provider
Candidate user collection, including:
Obtain user's history service data;
Obtain service provider history service data;
Calculate the similarity between the user's history service data and service provider history service data;
Similarity is formed to the candidate user collection of the service provider more than the user of default similarity threshold.
With reference in a first aspect, the disclosure in the first implementation of first aspect, it is described acquisition user's history service
Data, including:
Obtain user's history order data;
Extraction obtains user's history and orders service data from the user's history order data;
Statistical history orders the order frequency of service, obtains the label vector that corresponding history orders service, by the label to
Measure the history service data as the user.
With reference in a first aspect, the disclosure in the first implementation of first aspect, it is described acquisition service provider go through
History service data, including:
Obtain service provider History Order data;
Service provider history, which is obtained, from the service provider History Order extracting data orders service data;
Statistical history orders the order frequency of service, obtains the label vector that corresponding history orders service, by the label to
Measure the history service data as the service provider.
With reference in a first aspect, the disclosure in the first implementation of first aspect, it is described calculating candidate user concentrate
Candidate user be performed the probable value that pre-set user behavior occurs after predetermined registration operation, including:
Obtain training characteristics data;
Train to obtain training characteristics weight based on the training characteristics data;
Obtain that pre-set user behavior occurs after candidate user is performed predetermined registration operation based on the training characteristics weight calculation
Probable value.
With reference in a first aspect, the disclosure in the first implementation of first aspect, it is described acquisition training characteristics data,
Including:
The historic user training characteristics data for performing predetermined registration operation and pre-set user behavior occurring are obtained, as training
Positive sample;
Obtain and perform predetermined registration operation but the historic user training characteristics data of pre-set user behavior do not occur, training will be used as
Negative sample.
With reference in a first aspect, the disclosure in the first implementation of first aspect, it is described to be based on the training characteristics
Data train to obtain training characteristics weight, including:
It is trained based on the training positive sample and training negative sample, obtains feature weight forecast model;
Weight corresponding to the fixed reference feature data is predicted based on the feature weight forecast model.
With reference in a first aspect, the disclosure is based on the training in the first implementation of first aspect, using following formula
Candidate user u is calculated in feature weightiIt is performed the probable value p (u that pre-set user behavior occurs after predetermined registration operationi):
Wherein, fiRepresent the ith feature of a certain candidate user, λiFor feature weight corresponding to ith feature.
With reference in a first aspect, the disclosure in the first implementation of first aspect, it is described probable value is met it is default
The candidate user of condition confirms as targeted customer, including:
Expected revenus value after pre-set user behavior occurs for candidate user is calculated based on the probable value;
The candidate user that expected revenus value is more than to default revenue threshold confirms as targeted customer.
With reference to the first of first aspect and first aspect implementation, the disclosure is in second of realization side of first aspect
In formula, methods described also includes:Predetermined registration operation is performed to the targeted customer.
Second aspect, a kind of usage mining device is provided in the embodiment of the present disclosure.
Specifically, the usage mining device, including:
Acquisition module, it is configured as obtaining the candidate user collection of service provider;
Computing module, it is pre- to be configured as calculating generation after the candidate user that the candidate user is concentrated is performed predetermined registration operation
If the probable value of user behavior;
Confirm module, the candidate user for being configured as meeting the probable value preparatory condition confirms as targeted customer.
With reference to second aspect, in the first implementation of second aspect, the acquisition module includes the disclosure:
First acquisition submodule, it is configured as obtaining user's history service data;
Second acquisition submodule, it is configured as obtaining service provider history service data;
First calculating sub module, it is configured as calculating the user's history service data and service provider history service number
Similarity between;
First confirms submodule, and the user for being configured as similarity being more than default similarity threshold forms the service and provided
The candidate user collection of side.
With reference to second aspect, the disclosure is in the first implementation of second aspect, the first acquisition submodule bag
Include:
First acquisition unit, it is configured as obtaining user's history order data;
First extraction unit, it is configured as the extraction from the user's history order data and obtains user's history order service
Data;
First statistic unit, the order frequency that statistical history orders service is configured as, obtains corresponding history and order service
Label vector, the history service data using the label vector as the user.
With reference to second aspect, the disclosure is in the first implementation of second aspect, the second acquisition submodule bag
Include:
Second acquisition unit, it is configured as obtaining service provider History Order data;
Second extraction unit, it is configured as obtaining service provider from the service provider History Order extracting data
History orders service data;
Second statistic unit, the order frequency that statistical history orders service is configured as, obtains corresponding history and order service
Label vector, the history service data using the label vector as the service provider.
With reference to second aspect, in the first implementation of second aspect, the computing module includes the disclosure:
3rd acquisition submodule, it is configured as obtaining training characteristics data;
Submodule is trained, is configured as training to obtain training characteristics weight based on the training characteristics data;
Second calculating sub module, be configured as based on the training characteristics weight calculation obtain candidate user be performed it is default
The probable value of pre-set user behavior occurs after operation.
With reference to second aspect, the disclosure is in the first implementation of second aspect, the 3rd acquisition submodule bag
Include:
3rd acquiring unit, it is configured as obtaining the historic user training for performing predetermined registration operation and pre-set user behavior occurring
Characteristic, as training positive sample;
4th acquiring unit, obtain execution predetermined registration operation but the historic user training characteristics number of pre-set user behavior does not occur
According to, will be used as training negative sample.
With reference to second aspect, in the first implementation of second aspect, the training submodule includes the disclosure:
Training unit, it is configured as being trained based on the training positive sample and training negative sample, obtains feature weight
Forecast model;
Predicting unit, it is configured as predicting based on the feature weight forecast model and is weighed corresponding to the fixed reference feature data
Weight.
With reference to second aspect, the disclosure is in the first implementation of second aspect, the second calculating sub module quilt
It is configured to obtain candidate user u based on the training characteristics weight calculation using following formulaiOccur to preset after being performed predetermined registration operation
Probable value p (the u of user behaviori):
Wherein, fiRepresent the ith feature of a certain candidate user, λiFor feature weight corresponding to ith feature.
With reference to second aspect, in the first implementation of second aspect, the confirmation module includes the disclosure:
3rd calculating sub module, it is configured as after calculating candidate user generation pre-set user behavior based on the probable value
Expected revenus value;
Second confirms submodule, and the candidate user for being configured as expected revenus value being more than default revenue threshold confirms as mesh
Mark user.
With reference to the first of second aspect and second aspect implementation, the disclosure is in second of realization side of second aspect
In formula, described device also includes:Execution module, it is configured as performing predetermined registration operation to the targeted customer.
The third aspect, the embodiment of the present disclosure provide a kind of electronic equipment, including memory and processor, the memory
Usage mining device is supported to perform the computer instruction of usage mining method in above-mentioned first aspect for storing one or more,
The processor is configurable for performing the computer instruction stored in the memory.The usage mining device can be with
Including communication interface, for usage mining device and other equipment or communication.
Fourth aspect, the embodiment of the present disclosure provide a kind of computer-readable recording medium, for storing usage mining dress
Computer instruction used is put, it is involved by usage mining device that it, which is included for performing usage mining method in above-mentioned first aspect,
And computer instruction.
The technical scheme that the embodiment of the present disclosure provides can include the following benefits:
Above-mentioned technical proposal, by obtaining the candidate user collection of service provider, calculate the candidate that candidate user is concentrated and use
Family is performed the probable value of generation pre-set user behavior after predetermined registration operation, and the candidate user that probable value is met to preparatory condition thinks
It is the targeted customer for being most likely to occur pre-set user behavior, these targeted customers can be subsequently performed and send reward voucher, cash equivalent
The default excitation operation such as certificate.The technical scheme has more specific aim in terms of new user is developed, promote user's order, and success rate is high,
Also reduce simultaneously and develop the cost that new user is spent.The maximization of user satisfaction can be thus realized, reduces and loses
The risk of user, get more validated users.
It should be appreciated that the general description and following detailed description of the above are only exemplary and explanatory, not
The disclosure can be limited.
Brief description of the drawings
With reference to accompanying drawing, by the detailed description of following non-limiting embodiment, the further feature of the disclosure, purpose and excellent
Point will be apparent.In the accompanying drawings:
Fig. 1 shows the flow chart of the usage mining method according to the embodiment of the disclosure one;
Fig. 2 shows the flow chart of the step S101 according to Fig. 1 illustrated embodiments;
Fig. 3 shows the flow chart of the step S201 according to Fig. 2 illustrated embodiments;
Fig. 4 shows the flow chart of the step S202 according to Fig. 2 illustrated embodiments;
Fig. 5 shows the flow chart of the step S102 according to Fig. 1 illustrated embodiments;
Fig. 6 shows the flow chart of the step S501 according to Fig. 5 illustrated embodiments;
Fig. 7 shows the flow chart of the step S502 according to Fig. 5 illustrated embodiments;
Fig. 8 shows the flow chart of the step S103 according to Fig. 1 illustrated embodiments;
Fig. 9 shows the structured flowchart of the usage mining device according to the embodiment of the disclosure one;
Figure 10 shows the structured flowchart of the acquisition module 901 according to Fig. 9 illustrated embodiments;
Figure 11 shows the structured flowchart of the first acquisition submodule 1001 according to Figure 10 illustrated embodiments;
Figure 12 shows the structured flowchart of the second acquisition submodule 1002 according to Figure 10 illustrated embodiments;
Figure 13 shows the structured flowchart of the computing module 902 according to Fig. 9 illustrated embodiments;
Figure 14 shows the structured flowchart of the 3rd acquisition submodule 1301 according to Figure 13 illustrated embodiments;
Figure 15 shows the structured flowchart of the training submodule 1302 according to Figure 13 illustrated embodiments;
Figure 16 shows the structured flowchart of the confirmation module 903 according to Fig. 9 illustrated embodiments;
Figure 17 shows the structured flowchart of the electronic equipment according to the embodiment of the disclosure one;
Figure 18 is adapted for the knot of the computer system for realizing the usage mining method according to the embodiment of the disclosure one
Structure schematic diagram.
Embodiment
Hereinafter, the illustrative embodiments of the disclosure will be described in detail with reference to the attached drawings, so that those skilled in the art can
Easily realize them.In addition, for the sake of clarity, the portion unrelated with description illustrative embodiments is eliminated in the accompanying drawings
Point.
In the disclosure, it should be appreciated that the term of " comprising " or " having " etc. is intended to refer to disclosed in this specification
Feature, numeral, step, behavior, part, part or presence of its combination, and be not intended to exclude other one or more features,
Numeral, step, behavior, part, part or its combination there is a possibility that or be added.
It also should be noted that in the case where not conflicting, the feature in embodiment and embodiment in the disclosure
It can be mutually combined.Describe the disclosure in detail below with reference to the accompanying drawings and in conjunction with the embodiments.
The technical scheme that the embodiment of the present disclosure provides, by obtaining the candidate user collection of service provider, calculate candidate and use
The candidate user that family is concentrated is performed the probable value of generation pre-set user behavior after predetermined registration operation, and probable value is met into preparatory condition
Candidate user be considered to be most likely to occur the targeted customer of pre-set user behavior, can subsequently perform hair to these targeted customers
Send the predetermined registration operations such as reward voucher, coupons.The technical scheme has more specific aim in terms of new user is developed, promote user's order,
Success rate is high, while also reduces and develop the cost that new user is spent.The maximization of user satisfaction can be thus realized,
The risk for losing user is reduced, gets more validated users
Fig. 1 shows the flow chart of the usage mining method according to the embodiment of the disclosure one.As shown in figure 1, the user
Method for digging comprises the following steps S101-S103:
In step S101, the candidate user collection of service provider is obtained;
In step s 102, calculate after the candidate user that the candidate user is concentrated is performed predetermined registration operation and default use occurs
The probable value of family behavior;
In step s 103, the candidate user for the probable value being met to preparatory condition confirms as targeted customer.
In view of service provider when product or the service of oneself is promoted, if taken traditionally for whole
If user sends reward voucher, the mode for sending preferential popularization short message, popularization cost is higher, and due to user can not be expected
Receive the possibility of oneself product or service, it is low so as to cause to promote success rate, waste marketing money.Therefore, in the implementation
In mode, a kind of method that can excavate the user for probably receiving oneself product or service is proposed, i.e., obtains clothes first
The candidate user collection of business provider, then calculate after the candidate user that candidate user is concentrated is performed predetermined registration operation and default use occurs
The probable value of family behavior, then candidate user that probable value is met to preparatory condition are considered to be most likely to occur pre-set user row
For, for example receive the targeted customer of oneself product or service, these targeted customers can subsequently be performed and send reward voucher, coupons
Etc. predetermined registration operation.The technical scheme has more specific aim in terms of new user is developed, promote user's order, and success rate is high, also simultaneously
Reduce and develop the cost that new user is spent.
In an optional implementation of the present embodiment, as shown in Fig. 2 the step S101, that is, obtain service and provide
The step of candidate user collection of side, including step S201-S204:
In step s 201, user's history service data is obtained;
In step S202, service provider history service data is obtained;
In step S203, the phase between the user's history service data and service provider history service data is calculated
Like degree;
In step S204, the candidate that similarity is formed to the service provider more than the user of default similarity threshold uses
Family collection.
Wherein, user's history service data refers to that a certain user's purchase, the service data ordered, service provide in history
Square history service data refers to the service data that a certain service provider is sold, provided in history, and the service can be certain
The product that one businessman sells, such as the service that clothing, case and bag, diet vegetable etc. or a certain service provider provide, than
Such as carryout service, Courier Service.
In the implementation, user's history service data and service provider history service data are obtained first, then
The similarity between user's history service data and service provider history service data is calculated, similarity is more than default similar
The user of degree threshold value is considered the user than that may buy or receive a certain product or service, and these users constitute the clothes
Candidate user collection corresponding to business provider.
Specifically, when calculating the similarity between user's history service data and service provider history service data,
For the convenience calculated, user's history service data and service provider history service data can be first according to preset rules
Quantized, then the similarity between the service data after evaluation again.Wherein, the rule that concrete numerical value uses
It can be determined by one skilled in the art according to the situation and historical empirical data of practical application, the disclosure is not made any to it
Limit, all reasonable, achievable rules that quantize are each fallen within the protection domain of the disclosure.
In addition, when calculating the similarity between user's history service data and service provider history service data, can
Using a variety of similarity calculating methods, for example COS distance between the two is calculated, or calculate Euclidean distance between the two
Deng can also use other similarity calculating methods certainly, the disclosure will not enumerate.
Wherein, the default similarity threshold can be set according to the needs of practical application, and the disclosure specifically takes to it
Value is not especially limited.
In an optional implementation of the present embodiment, as shown in figure 3, the step S201, that is, obtain user's history
The step of service data, including step S301-S303:
In step S301, user's history order data is obtained;
In step s 302, extraction obtains user's history and orders service data from the user's history order data;
In step S303, statistical history orders the order frequency of service, obtain corresponding history order the label of service to
Amount, the history service data using the label vector as the user.
In an optional implementation of the present embodiment, as shown in figure 4, the step S202, that is, obtain service and provide
The step of square history service data, including step S401-S403:
In step S401, service provider History Order data are obtained;
In step S402, obtain service provider history from the service provider History Order extracting data and order
Service data;
In step S403, statistical history orders the order frequency of service, obtain corresponding history order the label of service to
Amount, the history service data using the label vector as the service provider.
In the implementation, when obtaining user's history service data, the History Order data of the user are obtained first,
Wherein, the History Order data of the user may include:The quantity of order, Order Type, order time, order contents, order
One or more in the data such as interior included service data, order price, are then extracted from user's history order data
Obtain user's history and order information on services, count the order frequency that the user orders service for history afterwards, the frequency will be ordered
Lined up according to the corresponding order for ordering service, obtain history corresponding to the user and order service labels vector, by the label
History service data of the vector as the user.The acquisition of service provider history service data is similar, no longer superfluous here
State.
It should be noted that in the implementation, the history service data of user and the history service of service provider
Data have been numeric forms, with regard to that need not be quantized again, the on the contrary then there is still a need for progress behaviour as described above that quantizes
Make.
In an optional implementation of the present embodiment, as shown in figure 5, the step S102, that is, calculate candidate user
The candidate user of concentration is performed the step of probable value that pre-set user behavior occurs after predetermined registration operation, including step S501-
S503:
In step S501, training characteristics data are obtained;
In step S502, train to obtain training characteristics weight based on the training characteristics data;
In step S503, obtain occurring after candidate user is performed predetermined registration operation based on the training characteristics weight calculation
The probable value of pre-set user behavior.
In the implementation, it is performed using the candidate user of the method calculating candidate user concentration of model training default
The probable value of pre-set user behavior occurs after operation, i.e., after calculating candidate user has been performed the predetermined registration operations such as transmission favor information
Produce the probable value of respective orders.Specifically, training characteristics data are obtained first, are then based on the training characteristics data training
Training characteristics weight is obtained, the training characteristics weight calculation is finally based on and obtains occurring after candidate user is performed predetermined registration operation
The probable value of pre-set user behavior.
Wherein, the training characteristics data can include:Obtained similarity numerical value calculated above, service provider provide
Service, the service that user orders, user attribute data, predetermined registration operation time of origin section, service provider place, where user
One or more in place, the user attribute data include:Name, sex, phone number, age, industry, occupation, people
The raw stage, Long-term Interest, preference, zone of action, place an order or access frequency, for the preference of service provider, for flat
One or more in the potential value of platform.
In an optional implementation of the present embodiment, as shown in fig. 6, the step S501, that is, obtain training characteristics
The step of data, including step S601-S602:
In step s 601, the historic user training characteristics number for performing predetermined registration operation and pre-set user behavior occurring is obtained
According to as training positive sample;
In step S602, obtain execution predetermined registration operation but the historic user training characteristics number of pre-set user behavior does not occur
According to as training negative sample.
In an optional implementation of the present embodiment, as shown in fig. 7, the step S502, i.e., based on the training
The step of characteristic trains to obtain training characteristics weight, including step S701-S702:
In step s 701, it is trained based on the training positive sample and training negative sample, obtains feature weight prediction
Model;
In step S702, weight corresponding to the fixed reference feature data is predicted based on the feature weight forecast model.
In the implementation, such as, can be used in machine learning it is a kind of it is simple efficiently, practical application is very extensive patrols
Regression model is collected to calculate the probable value, will be performed and sent the predetermined registration operations such as favor information and generate going through for respective orders
History user's training characteristics data will be performed to send the predetermined registration operations such as favor information but do not produce and accordingly ordered as training positive sample
Single historic user training characteristics data can obtain feature weight prediction mould as training negative sample using Logic Regression Models
Type, one group of feature weight value corresponding with training characteristics data is further obtained, can as a rule be obtained by optimization algorithm
The feature weight value optimal to one group corresponding with training characteristics data.
Wherein, for the convenience calculated, can will training positive sample and training negative sample be first according to it is same as above or
Different another preset rules are quantized, such as, for a female user, individual features data can be transformed to { " property
Not _ female ":1, " sex _ man ":0}.It is exemplary illustration above, the disclosure is not appointed for the rule that concrete numerical value uses
What is limited.
In an optional implementation of the present embodiment, for step S503, the training characteristics are based on using following formula
Weight calculation obtains candidate user uiIt is performed the probable value p (u that pre-set user behavior occurs after predetermined registration operationi):
Wherein, fiRepresent the ith feature of a certain candidate user, λiFor feature weight corresponding to ith feature.
In an optional implementation of the present embodiment, as shown in figure 8, the step S103, i.e., meet probable value
The step of candidate user of preparatory condition confirms as targeted customer, including step S801-S802:
In step S801, the expected revenus after pre-set user behavior occurs for candidate user is calculated based on the probable value
Value;
In step S802, it would be desirable to which the candidate user that financial value is more than default revenue threshold confirms as targeted customer.
The possibility of pre-set user behavior occurs to further determine that candidate user after predetermined registration operation is performed, in the reality
Apply in mode, introduce the concept that the expected revenus value after pre-set user behavior occurs for candidate user, that is, be primarily based on the probability
Value calculates candidate user and the expected revenus value after pre-set user behavior occurs, and expected revenus value is more than into default revenue threshold afterwards
Candidate user confirm as targeted customer.
In this embodiment, the expected revenus value that candidate user occurs after pre-set user behavior is defined as:αp(ui), its
In, a certain candidate user of α the expressions caused value in a preset time period, preset time for a certain service provider
It one month can also be the longer or shorter time that section, which can be, and candidate user caused value in this period section can recognize
For to be the user in this period section may be profit that service provider is brought, the profit value can be rule of thumb worth to,
Also can estimate to obtain according to big data, the disclosure is not especially limited to it.
Wherein, the default revenue threshold, which could be arranged to perform candidate user, sends the predetermined registration operation institutes such as favor information
Caused cost, the default revenue threshold can be configured and adjust according to the needs of practical application, with control targe user's
Quantity, when excessive more than the candidate user quantity of default revenue threshold, can also desirably financial value size for candidate
User is ranked up, and takes top n candidate user as targeted customer.
In an optional implementation of the present embodiment, methods described also includes performing the targeted customer default behaviour
The step of making, wherein, the predetermined registration operation includes:Favor information is sent, reward voucher is sent, sends coupons, opens and return existing power
Limit plus send integration, promotional items, give one or more in value-added service as an addition.
Following is embodiment of the present disclosure, can be used for performing embodiments of the present disclosure.
Fig. 9 shows the structured flowchart of the usage mining device according to the embodiment of the disclosure one, and the device can be by soft
Part, hardware or both are implemented in combination with as some or all of of electronic equipment.As shown in figure 9, the usage mining dress
Put including:
Acquisition module 901, it is configured as obtaining the candidate user collection of service provider;
Computing module 902, it is configured as calculating after the candidate user that the candidate user is concentrated is performed predetermined registration operation and sends out
The probable value of raw pre-set user behavior;
Confirm module 903, the candidate user for being configured as meeting the probable value preparatory condition confirms as targeted customer.
In view of service provider when product or the service of oneself is promoted, if taken traditionally for whole
If user sends reward voucher, the mode for sending preferential popularization short message, popularization cost is higher, and due to user can not be expected
Receive the possibility of oneself product or service, it is low so as to cause to promote success rate, waste marketing money.Therefore, in the implementation
In mode, a kind of device that can excavate the user for probably receiving oneself product or service is proposed, i.e., first by obtaining
Modulus block 901 obtains the candidate user collection of service provider, and the candidate of candidate user concentration is then calculated by computing module 902
User is performed the probable value of generation pre-set user behavior after predetermined registration operation, meets probable value finally by confirmation module 903
The candidate user of preparatory condition is considered to be most likely to occur pre-set user behavior, for example receives the target of oneself product or service
User, these targeted customers can subsequently be performed and send the predetermined registration operations such as reward voucher, coupons.The technical scheme is in the new use of development
Family, promote to have more specific aim in terms of user's order, success rate is high, while also reduces and develop the cost that new user is spent.
In an optional implementation of the present embodiment, as shown in Figure 10, the acquisition module 901 includes:
First acquisition submodule 1001, it is configured as obtaining user's history service data;
Second acquisition submodule 1002, it is configured as obtaining service provider history service data;
First calculating sub module 1003, it is configured as calculating the user's history service data and service provider history clothes
Similarity between data of being engaged in;
First confirms submodule 1004, and the user for being configured as similarity being more than default similarity threshold forms the service
The candidate user collection of provider.
Wherein, user's history service data refers to that a certain user's purchase, the service data ordered, service provide in history
Square history service data refers to the service data that a certain service provider is sold, provided in history, and the service can be certain
The product that one businessman sells, such as the service that clothing, case and bag, diet vegetable etc. or a certain service provider provide, than
Such as carryout service, Courier Service.
In the implementation, the first acquisition submodule 1001 obtains user's history service data, the second acquisition submodule
1002 obtain service provider history service data, and the first calculating sub module 1003 calculates user's history service data and carried with service
Similarity between supplier's history service data, first confirms that similarity is more than the use of default similarity threshold by submodule 1004
Family is considered the user than that may buy or receive a certain product or service, and these users constitute the service provider pair
The candidate user collection answered.
Specifically, the first calculating sub module 1003 is calculating user's history service data and service provider history service number
During similarity between, for the convenience calculated, user's history service data and service provider history can be serviced number
Quantized according to preset rules are first according to, then the similarity between the service data after evaluation again.Wherein, have
Body quantize use rule can be determined by one skilled in the art according to the situation and historical empirical data of practical application,
The disclosure is not limited in any way to it, and all reasonable, achievable rules that quantize are each fallen within the protection domain of the disclosure.
In addition, calculate user's history service data and service provider history service data in the first calculating sub module 1003
Between similarity when, a variety of similarity calculating methods can be used, for example calculate COS distance between the two, or calculate two
Euclidean distance between person etc., can also use other similarity calculating methods certainly, and the disclosure will not enumerate.
Wherein, the default similarity threshold can be set according to the needs of practical application, and the disclosure specifically takes to it
Value is not especially limited.
In an optional implementation of the present embodiment, as shown in figure 11, first acquisition submodule 1001 includes:
First acquisition unit 1101, it is configured as obtaining user's history order data;
First extraction unit 1102, it is configured as the extraction from the user's history order data and obtains user's history order
Service data;
First statistic unit 1103, the order frequency that statistical history orders service is configured as, obtains corresponding history and order
The label vector of service, the history service data using the label vector as the user.
In an optional implementation of the present embodiment, as shown in figure 12, second acquisition submodule 1002 includes:
Second acquisition unit 1201, it is configured as obtaining service provider History Order data;
Second extraction unit 1202, is configured as obtaining service from the service provider History Order extracting data carrying
Supplier's history orders service data;
Second statistic unit 1203, the order frequency that statistical history orders service is configured as, obtains corresponding history and order
The label vector of service, the history service data using the label vector as the service provider.
In the implementation, the first acquisition submodule 1001 passes through first when obtaining user's history service data
One acquiring unit 1101 obtains the History Order data of the user, wherein, the History Order data of the user may include:Order
Quantity, Order Type, the order time, order contents, service data included in order, one in the data such as order price
Kind is a variety of, and then the first extraction unit 1102 extracts from user's history order data obtains user's history order service letter
Breath, the order frequency that the first statistic unit 1103 statistics user services for history order afterwards, the frequency will be ordered according to phase
The order that service should be ordered is lined up, and is obtained history corresponding to the user and is ordered service labels vector, the label vector is made
For the history service data of the user.Second acquisition submodule 1002 obtain service provider history service data principle and this
It is similar, repeat no more here.
It should be noted that in the implementation, the history service data of user and the history service of service provider
Data have been numeric forms, with regard to that need not be quantized again, the on the contrary then there is still a need for progress behaviour as described above that quantizes
Make.
In an optional implementation of the present embodiment, as shown in figure 13, the computing module 902 includes:
3rd acquisition submodule 1301, it is configured as obtaining training characteristics data;
Submodule 1302 is trained, is configured as training to obtain training characteristics weight based on the training characteristics data;
Second calculating sub module 1303, is configured as obtaining candidate user based on the training characteristics weight calculation being performed
The probable value of pre-set user behavior occurs after predetermined registration operation.
In the implementation, the candidate that computing module 902 calculates candidate user concentration using the method for model training uses
Family is performed the probable value of generation pre-set user behavior after predetermined registration operation, that is, calculates candidate user and be performed transmission favor information
Etc. the probable value that respective orders are produced after predetermined registration operation.Specifically, the 3rd acquisition submodule 1301 first obtains training characteristics number
According to then training submodule 1302 trains to obtain training characteristics weight based on the training characteristics data, and the second last calculates son
Module 1303 obtains that pre-set user behavior occurs after candidate user is performed predetermined registration operation based on the training characteristics weight calculation
Probable value.
Wherein, the training characteristics data can include:Obtained similarity numerical value calculated above, service provider provide
Service, the service that user orders, user attribute data, predetermined registration operation time of origin section, service provider place, where user
One or more in place, the user attribute data include:Name, sex, phone number, age, industry, occupation, people
The raw stage, Long-term Interest, preference, zone of action, place an order or access frequency, for the preference of service provider, for flat
One or more in the potential value of platform.
In an optional implementation of the present embodiment, as shown in figure 14, the 3rd acquisition submodule 1301 includes:
3rd acquiring unit 1401, it is configured as obtaining the historic user for performing predetermined registration operation and pre-set user behavior occurring
Training characteristics data, as training positive sample;
4th acquiring unit 1402, obtain the historic user training spy for performing predetermined registration operation but pre-set user behavior not occurring
Data are levied, training negative sample will be used as.
In an optional implementation of the present embodiment, as shown in figure 15, the training submodule 1302 includes:
Training unit 1501, it is configured as being trained based on the training positive sample and training negative sample, obtains feature
Weight prediction model;
Predicting unit 1502, it is configured as predicting that the fixed reference feature data are corresponding based on the feature weight forecast model
Weight.
In the implementation, such as, a kind of simple efficient, reality in machine learning can be used to answer for training submodule 1302
The probable value is calculated with very extensive Logic Regression Models, will perform and send predetermined registration operation and the generations such as favor information
The historic user training characteristics data of respective orders will perform as training positive sample and send the predetermined registration operations such as favor information
But the historic user training characteristics data for not producing respective orders use logistic regression as training negative sample, training unit 1501
Model can obtain feature weight forecast model, and further predicting unit 1502 obtains one group of feature corresponding with training characteristics data
Weighted value, one group of optimal feature weight corresponding with training characteristics data as a rule can be obtained by optimization algorithm
Value.
Wherein, for the convenience calculated, can will training positive sample and training negative sample be first according to it is same as above or
Different another preset rules are quantized, such as, for a female user, individual features data can be transformed to { " property
Not _ female ":1, " sex _ man ":0}.It is exemplary illustration above, the disclosure is not appointed for the rule that concrete numerical value uses
What is limited.
In an optional implementation of the present embodiment, for the second calculating sub module 1303, it can be configured under utilization
Formula obtains candidate user u based on the training characteristics weight calculationiThe general of pre-set user behavior occurs after being performed predetermined registration operation
Rate value p (ui):
Wherein, fiRepresent the ith feature of a certain candidate user, λiFor feature weight corresponding to ith feature.
In an optional implementation of the present embodiment, as shown in figure 16, the confirmation module 903 includes:
3rd calculating sub module 1601, it is configured as calculating candidate user generation pre-set user behavior based on the probable value
Expected revenus value afterwards;
Second confirms submodule 1602, and the candidate user for being configured as expected revenus value being more than default revenue threshold confirms
For targeted customer.
The possibility of pre-set user behavior occurs to further determine that candidate user after predetermined registration operation is performed, in the reality
Apply in mode, confirm to introduce the concept that the expected revenus value after pre-set user behavior occurs for candidate user in module 903, that is, calculate
Submodule 1601 is primarily based on the probable value and calculates the expected revenus value after candidate user generation pre-set user behavior, Zhi Hou
The candidate user that expected revenus value is more than default revenue threshold by two confirmation submodules 1602 confirms as targeted customer.
In this embodiment, the expected revenus value that candidate user occurs after pre-set user behavior is defined as:αp(ui), its
In, a certain candidate user of α the expressions caused value in a preset time period, preset time for a certain service provider
It one month can also be the longer or shorter time that section, which can be, and candidate user caused value in this period section can recognize
For to be the user in this period section may be profit that service provider is brought, the profit value can be rule of thumb worth to,
Also can be calculated according to big data, the disclosure is not especially limited to it.
Wherein, the default revenue threshold, which could be arranged to perform candidate user, sends the predetermined registration operation institutes such as favor information
Caused cost, the default revenue threshold can be configured and adjust according to the needs of practical application, with control targe user's
Quantity, when excessive more than the candidate user quantity of default revenue threshold, can also desirably financial value size for candidate
User is ranked up, and takes top n candidate user as targeted customer.
In an optional implementation of the present embodiment, described device also includes execution module, the execution module quilt
It is configured to perform predetermined registration operation to the targeted customer, wherein, the predetermined registration operation includes:Transmission favor information, transmission are preferential
Certificate, coupons are sent, is opened and is returned existing authority plus send integration, promotional items, give one or more in value-added service as an addition.
The disclosure also discloses a kind of electronic equipment, and Figure 17 shows the knot of the electronic equipment according to the embodiment of the disclosure one
Structure block diagram, as shown in figure 17, the electronic equipment 1700 include memory 1701 and processor 1702;Wherein,
The memory 1701 is used to store one or more computer instruction, wherein, one or more computer
Instruction is performed by the processor 1702 to realize:
Obtain the candidate user collection of service provider;
Calculate the probability that the candidate user that the candidate user is concentrated is performed generation pre-set user behavior after predetermined registration operation
Value;
The candidate user that the probable value is met to preparatory condition confirms as targeted customer.
One or more computer instruction can be also performed by the processor 1702 to realize:
The candidate user collection for obtaining service provider, including:
Obtain user's history service data;
Obtain service provider history service data;
Calculate the similarity between the user's history service data and service provider history service data;
Similarity is formed to the candidate user collection of the service provider more than the user of default similarity threshold.
The acquisition user's history service data, including:
Obtain user's history order data;
Extraction obtains user's history and orders service data from the user's history order data;
Statistical history orders the order frequency of service, obtains the label vector that corresponding history orders service, by the label to
Measure the history service data as the user.
The acquisition service provider history service data, including:
Obtain service provider History Order data;
Service provider history, which is obtained, from the service provider History Order extracting data orders service data;
Statistical history orders the order frequency of service, obtains the label vector that corresponding history orders service, by the label to
Measure the history service data as the service provider.
The probability for calculating the candidate user that candidate user is concentrated and being performed generation pre-set user behavior after predetermined registration operation
Value, including:
Obtain training characteristics data;
Train to obtain training characteristics weight based on the training characteristics data;
Obtain that pre-set user behavior occurs after candidate user is performed predetermined registration operation based on the training characteristics weight calculation
Probable value.
The acquisition training characteristics data, including:
The historic user training characteristics data for performing predetermined registration operation and pre-set user behavior occurring are obtained, as training
Positive sample;
Obtain and perform predetermined registration operation but the historic user training characteristics data of pre-set user behavior do not occur, training will be used as
Negative sample.
It is described to train to obtain training characteristics weight based on the training characteristics data, including:
It is trained based on the training positive sample and training negative sample, obtains feature weight forecast model;
Weight corresponding to the fixed reference feature data is predicted based on the feature weight forecast model.
Candidate user u is obtained based on the training characteristics weight calculation using following formulaiOccur in advance after being performed predetermined registration operation
If the probable value p (u of user behaviori):
Wherein, fiRepresent the ith feature of a certain candidate user, λiFor feature weight corresponding to ith feature.
The candidate user that probable value is met to preparatory condition confirms as targeted customer, including:
Expected revenus value after pre-set user behavior occurs for candidate user is calculated based on the probable value;
The candidate user that expected revenus value is more than to default revenue threshold confirms as targeted customer.
Also include:
Predetermined registration operation is performed to the targeted customer.
Figure 18 is suitable to be used for realizing that the structure of the computer system of the usage mining method according to disclosure embodiment is shown
It is intended to.
As shown in figure 18, computer system 1800 includes CPU (CPU) 1801, its can according to be stored in only
Read the program in memory (ROM) 1802 or be loaded into from storage part 1808 in random access storage device (RAM) 1803
Program and perform the various processing in the embodiment shown in above-mentioned Fig. 1-8.In RAM1803, also it is stored with system 1800 and grasps
Various programs and data needed for making.CPU1801, ROM1802 and RAM1803 are connected with each other by bus 1804.Input/defeated
Go out (I/O) interface 1805 and be also connected to bus 1804.
I/O interfaces 1805 are connected to lower component:Importation 1806 including keyboard, mouse etc.;Including such as negative electrode
The output par, c 1807 of ray tube (CRT), liquid crystal display (LCD) etc. and loudspeaker etc.;Storage part including hard disk etc.
1808;And the communications portion 1809 of the NIC including LAN card, modem etc..Communications portion 1809 passes through
Communication process is performed by the network of such as internet.Driver 1810 is also according to needing to be connected to I/O interfaces 1805.It is detachable to be situated between
Matter 1811, such as disk, CD, magneto-optic disk, semiconductor memory etc., it is arranged on as needed on driver 1810, so as to
Storage part 1808 is mounted into as needed in the computer program read from it.
Especially, according to embodiment of the present disclosure, computer is may be implemented as above with reference to Fig. 1-8 methods described
Software program.For example, embodiment of the present disclosure includes a kind of computer program product, it includes being tangibly embodied in and its can
The computer program on medium is read, the computer program includes the program code for the usage mining method for being used to perform Fig. 1-8.
In such embodiment, the computer program can be downloaded and installed by communications portion 1809 from network, and/or
It is mounted from detachable media 1811.
Flow chart and block diagram in accompanying drawing, it is illustrated that according to the system, method and computer of the various embodiments of the disclosure
Architectural framework in the cards, function and the operation of program product.At this point, each square frame in course diagram or block diagram can be with
A part for a module, program segment or code is represented, a part for the module, program segment or code includes one or more
For realizing the executable instruction of defined logic function.It should also be noted that some as replace realization in, institute in square frame
The function of mark can also be with different from the order marked in accompanying drawing generation.For example, two square frames succeedingly represented are actual
On can perform substantially in parallel, they can also be performed in the opposite order sometimes, and this is depending on involved function.Also
It is noted that the combination of each square frame and block diagram in block diagram and/or flow chart and/or the square frame in flow chart, Ke Yiyong
Function as defined in execution or the special hardware based system of operation are realized, or can be referred to specialized hardware and computer
The combination of order is realized.
Being described in unit or module involved in disclosure embodiment can be realized by way of software, also may be used
Realized in a manner of by hardware.Described unit or module can also be set within a processor, these units or module
Title do not form restriction to the unit or module in itself under certain conditions.
As on the other hand, the disclosure additionally provides a kind of computer-readable recording medium, the computer-readable storage medium
Matter can be the computer-readable recording medium included in device described in above-mentioned embodiment;Can also be individualism,
Without the computer-readable recording medium in supplying equipment.Computer-readable recording medium storage has one or more than one journey
Sequence, described program is used for performing by one or more than one processor is described in disclosed method.
Above description is only the preferred embodiment of the disclosure and the explanation to institute's application technology principle.People in the art
Member should be appreciated that invention scope involved in the disclosure, however it is not limited to the technology that the particular combination of above-mentioned technical characteristic forms
Scheme, while should also cover in the case where not departing from the inventive concept, carried out by above-mentioned technical characteristic or its equivalent feature
The other technical schemes for being combined and being formed.Such as features described above has similar work(with the (but not limited to) disclosed in the disclosure
The technical scheme that the technical characteristic of energy is replaced mutually and formed.
The present disclosure discloses A1, a kind of usage mining method, methods described includes:Obtain the candidate user of service provider
Collection;Calculate the probable value that the candidate user that the candidate user is concentrated is performed generation pre-set user behavior after predetermined registration operation;Will
The candidate user that the probable value meets preparatory condition confirms as targeted customer.A2, the method according to A1, it is described to obtain clothes
The candidate user collection of business provider, including:Obtain user's history service data;Obtain service provider history service data;Meter
Calculate the similarity between the user's history service data and service provider history service data;Similarity is more than default phase
The candidate user collection of the service provider is formed like the user of degree threshold value.A3, the method according to A2, the acquisition user go through
History service data, including:Obtain user's history order data;Extraction obtains user's history from the user's history order data
Order service data;Statistical history orders the order frequency of service, obtains the label vector that corresponding history orders service, this is marked
History service data of the label vector as the user.A4, the method according to A2, the acquisition service provider history service
Data, including:Obtain service provider History Order data;Taken from the service provider History Order extracting data
Provider's history of being engaged in orders service data;Statistical history orders the order frequency of service, obtains the mark that corresponding history orders service
Label vector, the history service data using the label vector as the service provider.A5, the method according to A2, the meter
The probable value that the candidate user that candidate user is concentrated is performed generation pre-set user behavior after predetermined registration operation is calculated, including:Obtain instruction
Practice characteristic;Train to obtain training characteristics weight based on the training characteristics data;Based on the training characteristics weight calculation
Obtain the probable value that candidate user is performed generation pre-set user behavior after predetermined registration operation.A6, the method according to A5, it is described
Training characteristics data are obtained, including:Obtain the historic user training characteristics number for performing predetermined registration operation and pre-set user behavior occurring
According to as training positive sample;Obtain execution predetermined registration operation but the historic user training characteristics of pre-set user behavior do not occur
Data, training negative sample will be used as.A7, the method according to A6, described trained based on the training characteristics data are instructed
Practice feature weight, including:It is trained based on the training positive sample and training negative sample, obtains feature weight forecast model;
Weight corresponding to the fixed reference feature data is predicted based on the feature weight forecast model.A8, the method according to A5, profit
Candidate user u is obtained based on the training characteristics weight calculation with following formulaiPre-set user behavior occurs after being performed predetermined registration operation
Probable value p (ui):Wherein, fiRepresent the ith feature of a certain candidate user, λiFor i-th
Feature weight corresponding to feature.A9, the method according to A1, the candidate user that probable value is met to preparatory condition confirm
For targeted customer, including:Expected revenus value after pre-set user behavior occurs for candidate user is calculated based on the probable value;By the phase
Financial value is hoped to be more than the candidate user of default revenue threshold and confirm as targeted customer.A10, the method according to A1, in addition to:
Predetermined registration operation is performed to the targeted customer.
The present disclosure discloses B11, a kind of usage mining device, described device includes:Acquisition module, it is configured as obtaining clothes
The candidate user collection of business provider;Computing module, be configured as calculating the candidate user that the candidate user is concentrated be performed it is pre-
If the probable value of pre-set user behavior occurs after operation;Confirm module, be configured as the probable value meeting preparatory condition
Candidate user confirms as targeted customer.B12, the device according to B11, the acquisition module include:First acquisition submodule,
It is configured as obtaining user's history service data;Second acquisition submodule, it is configured as obtaining service provider history service number
According to;First calculating sub module, be configured as calculating the user's history service data and service provider history service data it
Between similarity;First confirms submodule, and the user for being configured as similarity being more than default similarity threshold forms the service
The candidate user collection of provider.B13, the device according to B12, first acquisition submodule include:First acquisition unit,
It is configured as obtaining user's history order data;First extraction unit, it is configured as carrying from the user's history order data
Obtain user's history and order service data;First statistic unit, the order frequency that statistical history orders service is configured as, is obtained
The label vector of service, the history service data using the label vector as the user are ordered to corresponding history.B14, according to B12
Described device, second acquisition submodule include:Second acquisition unit, it is configured as obtaining service provider History Order
Data;Second extraction unit, is configured as obtaining service provider from the service provider History Order extracting data going through
History orders service data;Second statistic unit, the order frequency that statistical history orders service is configured as, obtains corresponding history and order
Purchase the label vector of service, the history service data using the label vector as the service provider.B15, according to B12
Device, the computing module include:3rd acquisition submodule, it is configured as obtaining training characteristics data;Train submodule, by with
It is set to and trains to obtain training characteristics weight based on the training characteristics data;Second calculating sub module, it is configured as based on described
Training characteristics weight calculation obtains the probable value that candidate user is performed generation pre-set user behavior after predetermined registration operation.B16, basis
Device described in B15, the 3rd acquisition submodule include:3rd acquiring unit, be configured as obtain perform predetermined registration operation and
The historic user training characteristics data of pre-set user behavior occur, as training positive sample;4th acquiring unit, acquisition are held
Row predetermined registration operation but the historic user training characteristics data that pre-set user behavior does not occur, training negative sample will be used as.B17, root
According to the device described in B16, the training submodule includes:Training unit, it is configured as based on the training positive sample and training
Negative sample is trained, and obtains feature weight forecast model;Predicting unit, it is configured as being based on the feature weight forecast model
Predict weight corresponding to the fixed reference feature data.B18, the device according to B15, second calculating sub module are configured
To obtain candidate user u based on the training characteristics weight calculation using following formulaiPre-set user occurs after being performed predetermined registration operation
Probable value p (the u of behaviori):Wherein, fiRepresent the ith feature of a certain candidate user, λiFor
Feature weight corresponding to i feature.B19, the device according to B11, the confirmation module include:3rd calculating sub module,
It is configured as calculating the expected revenus value after pre-set user behavior occurs for candidate user based on the probable value;Second confirms submodule
Block, the candidate user for being configured as expected revenus value being more than default revenue threshold confirm as targeted customer.B20, according to B11 institutes
The device stated, in addition to:Execution module, it is configured as performing predetermined registration operation to the targeted customer.
The present disclosure discloses C21, a kind of electronic equipment, including memory and processor;Wherein, the memory is used to deposit
One or more computer instruction is stored up, wherein, one or more computer instruction is by the computing device to realize such as
Method described in any one of A1-A10.
The disclosure also discloses D22, a kind of computer-readable recording medium, is stored thereon with computer instruction, the calculating
The method as described in any one of A1-A10 is realized in machine instruction when being executed by processor.
Claims (10)
- A kind of 1. usage mining method, it is characterised in that methods described includes:Obtain the candidate user collection of service provider;Calculate the probable value that the candidate user that the candidate user is concentrated is performed generation pre-set user behavior after predetermined registration operation;The candidate user that the probable value is met to preparatory condition confirms as targeted customer.
- 2. according to the method for claim 1, it is characterised in that the candidate user collection for obtaining service provider, including:Obtain user's history service data;Obtain service provider history service data;Calculate the similarity between the user's history service data and service provider history service data;Similarity is formed to the candidate user collection of the service provider more than the user of default similarity threshold.
- 3. according to the method for claim 2, it is characterised in that the candidate user for calculating candidate user concentration is performed The probable value of pre-set user behavior occurs after predetermined registration operation, including:Obtain training characteristics data;Train to obtain training characteristics weight based on the training characteristics data;Obtain that the general of pre-set user behavior occurs after candidate user is performed predetermined registration operation based on the training characteristics weight calculation Rate value.
- 4. according to the method for claim 3, it is characterised in that the acquisition training characteristics data, including:The historic user training characteristics data for performing predetermined registration operation and pre-set user behavior occurring are obtained, as the positive sample of training This;Obtain and perform predetermined registration operation but the historic user training characteristics data of pre-set user behavior do not occur, the negative sample of training will be used as This.
- 5. according to the method for claim 4, it is characterised in that described to be trained based on the training characteristics data Feature weight, including:It is trained based on the training positive sample and training negative sample, obtains feature weight forecast model;Weight corresponding to the fixed reference feature data is predicted based on the feature weight forecast model.
- 6. according to the method for claim 3, it is characterised in that obtained using following formula based on the training characteristics weight calculation Candidate user uiIt is performed the probable value p (u that pre-set user behavior occurs after predetermined registration operationi):<mrow> <mi>p</mi> <mrow> <mo>(</mo> <msub> <mi>u</mi> <mi>i</mi> </msub> <mo>)</mo> </mrow> <mo>=</mo> <mfrac> <mn>1</mn> <mrow> <mn>1</mn> <mo>+</mo> <mi>exp</mi> <mrow> <mo>(</mo> <mo>-</mo> <msub> <mi>&Sigma;</mi> <mi>i</mi> </msub> <msub> <mi>&lambda;</mi> <mi>i</mi> </msub> <msub> <mi>f</mi> <mi>i</mi> </msub> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>,</mo> </mrow>Wherein, fiRepresent the ith feature of a certain candidate user, λiFor feature weight corresponding to ith feature.
- 7. according to the method for claim 1, it is characterised in that it is described probable value is met to preparatory condition candidate user it is true Think targeted customer, including:Expected revenus value after pre-set user behavior occurs for candidate user is calculated based on the probable value;The candidate user that expected revenus value is more than to default revenue threshold confirms as targeted customer.
- 8. a kind of usage mining device, it is characterised in that described device includes:Acquisition module, it is configured as obtaining the candidate user collection of service provider;Computing module, it is configured as calculating generation after the candidate user that the candidate user is concentrated is performed predetermined registration operation and presets use The probable value of family behavior;Confirm module, the candidate user for being configured as meeting the probable value preparatory condition confirms as targeted customer.
- 9. a kind of electronic equipment, it is characterised in that including memory and processor;Wherein,The memory is used to store one or more computer instruction, wherein, one or more computer instruction is by institute Computing device is stated to realize the method as described in claim any one of 1-7.
- 10. a kind of computer-readable recording medium, is stored thereon with computer instruction, it is characterised in that the computer instruction quilt The method as described in claim any one of 1-7 is realized during computing device.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711124056.8A CN107844584A (en) | 2017-11-14 | 2017-11-14 | Usage mining method, apparatus, electronic equipment and computer-readable recording medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711124056.8A CN107844584A (en) | 2017-11-14 | 2017-11-14 | Usage mining method, apparatus, electronic equipment and computer-readable recording medium |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107844584A true CN107844584A (en) | 2018-03-27 |
Family
ID=61678888
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711124056.8A Pending CN107844584A (en) | 2017-11-14 | 2017-11-14 | Usage mining method, apparatus, electronic equipment and computer-readable recording medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107844584A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108647990A (en) * | 2018-04-04 | 2018-10-12 | 北京奇艺世纪科技有限公司 | A kind of method, apparatus and electronic equipment of determining target user |
CN109829592A (en) * | 2019-03-07 | 2019-05-31 | 深圳市丰巢科技有限公司 | A kind of attrition prediction method, apparatus, equipment and the storage medium of express delivery cabinet user |
CN110675179A (en) * | 2018-07-03 | 2020-01-10 | 百度在线网络技术(北京)有限公司 | Marketing information processing method and device, electronic equipment and readable storage medium |
CN111241543A (en) * | 2020-01-07 | 2020-06-05 | 中国搜索信息科技股份有限公司 | Method and system for intelligently resisting DDoS attack by application layer |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2002329086A (en) * | 2001-04-27 | 2002-11-15 | Nippon Telegr & Teleph Corp <Ntt> | Method, system and program for selling over internet and its storage medium |
CN106204063A (en) * | 2016-06-30 | 2016-12-07 | 北京奇艺世纪科技有限公司 | A kind of paying customer's method for digging and device |
CN106557954A (en) * | 2016-11-28 | 2017-04-05 | 贵安新区绿谷科技有限公司 | The method and device of customer service marketing |
CN106897905A (en) * | 2017-02-28 | 2017-06-27 | 百度在线网络技术(北京)有限公司 | Method, device and electronic equipment for pushed information |
CN107220852A (en) * | 2017-05-26 | 2017-09-29 | 北京小度信息科技有限公司 | Method, device and server for determining target recommended user |
-
2017
- 2017-11-14 CN CN201711124056.8A patent/CN107844584A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2002329086A (en) * | 2001-04-27 | 2002-11-15 | Nippon Telegr & Teleph Corp <Ntt> | Method, system and program for selling over internet and its storage medium |
CN106204063A (en) * | 2016-06-30 | 2016-12-07 | 北京奇艺世纪科技有限公司 | A kind of paying customer's method for digging and device |
CN106557954A (en) * | 2016-11-28 | 2017-04-05 | 贵安新区绿谷科技有限公司 | The method and device of customer service marketing |
CN106897905A (en) * | 2017-02-28 | 2017-06-27 | 百度在线网络技术(北京)有限公司 | Method, device and electronic equipment for pushed information |
CN107220852A (en) * | 2017-05-26 | 2017-09-29 | 北京小度信息科技有限公司 | Method, device and server for determining target recommended user |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108647990A (en) * | 2018-04-04 | 2018-10-12 | 北京奇艺世纪科技有限公司 | A kind of method, apparatus and electronic equipment of determining target user |
CN108647990B (en) * | 2018-04-04 | 2022-06-03 | 北京奇艺世纪科技有限公司 | Method and device for determining target user and electronic equipment |
CN110675179A (en) * | 2018-07-03 | 2020-01-10 | 百度在线网络技术(北京)有限公司 | Marketing information processing method and device, electronic equipment and readable storage medium |
CN109829592A (en) * | 2019-03-07 | 2019-05-31 | 深圳市丰巢科技有限公司 | A kind of attrition prediction method, apparatus, equipment and the storage medium of express delivery cabinet user |
CN109829592B (en) * | 2019-03-07 | 2023-03-21 | 深圳市丰巢科技有限公司 | Loss prediction method, device, equipment and storage medium for express cabinet users |
CN111241543A (en) * | 2020-01-07 | 2020-06-05 | 中国搜索信息科技股份有限公司 | Method and system for intelligently resisting DDoS attack by application layer |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Feldman et al. | Customer choice models vs. machine learning: Finding optimal product displays on Alibaba | |
CN107578294A (en) | User's behavior prediction method, apparatus and electronic equipment | |
CN107844584A (en) | Usage mining method, apparatus, electronic equipment and computer-readable recording medium | |
CN107944593A (en) | A kind of resource allocation methods and device, electronic equipment | |
Han et al. | Segmentation of telecom customers based on customer value by decision tree model | |
Shyur | COTS evaluation using modified TOPSIS and ANP | |
CN111008858A (en) | Commodity sales prediction method and system | |
CN107784390A (en) | Recognition methods, device, electronic equipment and the storage medium of subscriber lifecycle | |
US20200143439A1 (en) | System and method for integrating retail price optimization for revenue and profit with business rules | |
CN107273436A (en) | The training method and trainer of a kind of recommended models | |
JP6573205B2 (en) | PROCESSING DEVICE, INFORMATION PROCESSING METHOD, PROGRAM, AND MARKETING INFORMATION PROCESSING DEVICE FOR Calculating Predictive Data Regarding Use Of Application Program By One User | |
CN107369052A (en) | User's registration behavior prediction method, apparatus and electronic equipment | |
CN110111090A (en) | A kind of distribution method and device of electronics red packet | |
CN108133390A (en) | For predicting the method and apparatus of user behavior and computing device | |
Yu et al. | The return policy model with fuzzy demands and asymmetric information | |
Lieckens et al. | Network and contract optimization for maintenance services with remanufacturing | |
CN110298707A (en) | A kind of method and device for tariff package scene price | |
CN110069699A (en) | Order models training method and device | |
CN110033324A (en) | Data processing method, device, electronic equipment and computer readable storage medium | |
CN108805613A (en) | Electronic device promotes list recommendation method and computer readable storage medium | |
Chen et al. | Purchase behavior prediction in e-commerce with factorization machines | |
Shi et al. | A generalized count model on customers' purchases in O2O market | |
Liu et al. | Iterative prediction-and-optimization for E-logistics distribution network design | |
CN110326005A (en) | Information processing equipment, information processing method and program | |
Wagh et al. | Customer churn prediction in telecom sector using machine learning techniques |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180327 |
|
RJ01 | Rejection of invention patent application after publication |