CN110162697A - User browsing behavior analysis method, device, medium and terminal based on big data - Google Patents
User browsing behavior analysis method, device, medium and terminal based on big data Download PDFInfo
- Publication number
- CN110162697A CN110162697A CN201910295880.2A CN201910295880A CN110162697A CN 110162697 A CN110162697 A CN 110162697A CN 201910295880 A CN201910295880 A CN 201910295880A CN 110162697 A CN110162697 A CN 110162697A
- Authority
- CN
- China
- Prior art keywords
- browsing behavior
- group
- user
- browsing
- historical viewings
- 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
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3346—Query execution using probabilistic model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
Abstract
The invention belongs to field of computer technology more particularly to a kind of user browsing behavior analysis method, device, computer readable storage medium and terminal devices based on big data.The method obtains browsing behavior data of the user in preset each analysis dimension;According to the browsing behavior feature vector of user user described in the browsing behavior data configuration in each analysis dimension;Each historical viewings record in preset browsing behavior database is divided into more than two browsing behavior groups;Analysis probability value is calculated according to the browsing behavior feature vector of each browsing behavior group and the user, it is the probability value of preset first result that the analysis probability value, which is the browsing result of the user,;The interactive operation with the user is executed according to the analysis probability value.Sufficient mining analysis is carried out by browsing behavior to user, understand the demand of user, and interaction corresponding with user's progress according to the demand of user in time, thus the significant increase experience of user.
Description
Technical field
The invention belongs to field of computer technology more particularly to a kind of user browsing behavior analysis sides based on big data
Method, device, computer readable storage medium and terminal device.
Background technique
Browsing behavior of the user in webpage or application program often means that user has certain demand, for example, if
A certain user continually browses a certain product in webpage or application program, then means that the user may have the product
Stronger purchase intention, and sufficient mining analysis is not carried out to the browsing behavior of user in the prior art, it can not timely root
According to the interaction corresponding with user's progress of the demand of user, cause user experience poor.
Summary of the invention
In view of this, the user browsing behavior analysis method that the embodiment of the invention provides a kind of based on big data, device,
Computer readable storage medium and terminal device, to solve in the prior art not dig the browsing behavior of user adequately
Pick analysis, can not interaction corresponding with user's progress according to the demand of user in time, lead to the problem that user experience is poor.
The first aspect of the embodiment of the present invention provides a kind of user browsing behavior analysis method based on big data, can be with
Include:
Browsing behavior data of the user in preset each analysis dimension are obtained, the browsing behavior data are in webpage
Or generated behavioral data when browsing objective object in application program;
According to the browsing behavior feature of user user described in the browsing behavior data configuration in each analysis dimension
Vector;
Each historical viewings record in preset browsing behavior database is divided into more than two browsing behavior groups
Group, wherein every historical viewings record is made of the browsing behavior feature vector and browsing result of a historical viewings behavior;
It is calculated according to the browsing behavior feature vector of each browsing behavior group and the user and analyzes probability value, described point
It is the probability value of preset first result that analysis probability value, which is the browsing result of the user,;
The interactive operation with the user is executed according to the analysis probability value.
The second aspect of the embodiment of the present invention provides a kind of user browsing behavior analytical equipment, may include:
Browsing behavior data acquisition module, for obtaining browsing behavior number of the user in preset each analysis dimension
It is generated behavioral data when browsing objective object in webpage or application program according to, browsing behavior data;
Feature vector constructing module, for the browsing behavior data configuration institute according to the user in each analysis dimension
State the browsing behavior feature vector of user;
Group division module, for each historical viewings record in preset browsing behavior database to be divided into two
Above browsing behavior group, wherein every historical viewings record from the browsing behavior feature of a historical viewings behavior to
Amount and browsing result composition;
Probability evaluation entity, for being calculated according to the browsing behavior feature vector of each browsing behavior group and the user
Probability value is analyzed, it is the probability value of preset first result that the analysis probability value, which is the browsing result of the user,;
Interactive operation execution module, for executing the interactive operation with the user according to the analysis probability value.
The third aspect of the embodiment of the present invention provides a kind of computer readable storage medium, the computer-readable storage
Media storage has computer-readable instruction, and the computer-readable instruction realizes following steps when being executed by processor:
Browsing behavior data of the user in preset each analysis dimension are obtained, the browsing behavior data are in webpage
Or generated behavioral data when browsing objective object in application program;
According to the browsing behavior feature of user user described in the browsing behavior data configuration in each analysis dimension
Vector;
Each historical viewings record in preset browsing behavior database is divided into more than two browsing behavior groups
Group, wherein every historical viewings record is made of the browsing behavior feature vector and browsing result of a historical viewings behavior;
It is calculated according to the browsing behavior feature vector of each browsing behavior group and the user and analyzes probability value, described point
It is the probability value of preset first result that analysis probability value, which is the browsing result of the user,;
The interactive operation with the user is executed according to the analysis probability value.
The fourth aspect of the embodiment of the present invention provides a kind of terminal device, including memory, processor and is stored in
In the memory and the computer-readable instruction that can run on the processor, the processor executes the computer can
Following steps are realized when reading instruction:
Browsing behavior data of the user in preset each analysis dimension are obtained, the browsing behavior data are in webpage
Or generated behavioral data when browsing objective object in application program;
According to the browsing behavior feature of user user described in the browsing behavior data configuration in each analysis dimension
Vector;
Each historical viewings record in preset browsing behavior database is divided into more than two browsing behavior groups
Group, wherein every historical viewings record is made of the browsing behavior feature vector and browsing result of a historical viewings behavior;
It is calculated according to the browsing behavior feature vector of each browsing behavior group and the user and analyzes probability value, described point
It is the probability value of preset first result that analysis probability value, which is the browsing result of the user,;
The interactive operation with the user is executed according to the analysis probability value.
Existing beneficial effect is the embodiment of the present invention compared with prior art: when user is in webpage or application program
When browsing objective object, the embodiment of the present invention obtains browsing behavior data of the user in preset each analysis dimension first,
And the browsing behavior feature vector of user is constructed accordingly, then by each historical viewings in preset browsing behavior database
Record be divided into more than two browsing behavior groups, using these historical datas be used as to user progress big data analysis according to
According to calculating the browsing result of user according to the browsing behavior feature vector of each browsing behavior group and user is preset first
As a result probability value carries out sufficient mining analysis by the browsing behavior to user, understand the demand of user, and root in time
According to the interaction corresponding with user's progress of the demand of user, to meet the needs of users as far as possible, thus the significant increase body of user
It tests.
Detailed description of the invention
It to describe the technical solutions in the embodiments of the present invention more clearly, below will be to embodiment or description of the prior art
Needed in attached drawing be briefly described, it should be apparent that, the accompanying drawings in the following description is only of the invention some
Embodiment for those of ordinary skill in the art without any creative labor, can also be according to these
Attached drawing obtains other attached drawings.
Fig. 1 is a kind of one embodiment stream of the user browsing behavior analysis method based on big data in the embodiment of the present invention
Cheng Tu;
Fig. 2 is that each historical viewings record in preset browsing behavior database is divided into more than two browsings to go
For the schematic flow diagram of group;
Fig. 3 is a kind of one embodiment structure chart of user browsing behavior analytical equipment in the embodiment of the present invention;
Fig. 4 is a kind of schematic block diagram of terminal device in the embodiment of the present invention.
Specific embodiment
In order to make the invention's purpose, features and advantages of the invention more obvious and easy to understand, below in conjunction with the present invention
Attached drawing in embodiment, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that disclosed below
Embodiment be only a part of the embodiment of the present invention, and not all embodiment.Based on the embodiments of the present invention, this field
Those of ordinary skill's all other embodiment obtained without making creative work, belongs to protection of the present invention
Range.
Referring to Fig. 1, an a kind of reality of the user browsing behavior analysis method based on big data in the embodiment of the present invention
Applying example may include:
Step S101, browsing behavior data of the user in preset each analysis dimension are obtained.
The browsing behavior data are generated behavioral data when browsing objective object in webpage or application program,
The target object can be product for sale shown in the webpage or application program.
Generally, user mainly passes through two kinds of approach and browses product on network.One is the webpages for passing through the end PC and Web
It is browsed, in which case it is possible to use the mode of JavaScript label, since user issues http request,
JavaScript label, which returns in the web page code of user, to work as page presentation comprising one section of special JavaScript code
While this section of code be also carried out.This section of code can obtain details (access time, browsing from the Cookie of user
Device information, tool manufacturer assign userID of active user etc.) and it is sent to specified analysing terminal, which is this
The subject of implementation of embodiment.Another kind is browsed by the end application program (APP), in such a case, it is possible to be based on SDK
Point is embedded, acquisition user carries out parameter passback, be sent in analysing terminal in the click behavior of the page.
In the present embodiment, each analysis dimension includes but is not limited to following specific analysis dimension:
(1) browse the period: browsing period, that is, user executes the period locating when browse operation, by a large amount of statistical
Analysis, for user within the different periods, the behavior of purchase product has great difference.It in the present embodiment can be according to reality
Situation carries out Time segments division, for example, early, middle and late three periods can be divided by one day according to the actual situation, it can also be by one
It is divided into 24 periods, and each hour is a period, can also carry out Time segments division according to working hour, rest period,
Or the Time segments division of other forms can be carried out.
For different Time segments division modes, it need to be handled, one day is divided into early, middle and late using corresponding numeralization
It is illustrated for three periods, if the browsing period is the early period, the value numerical value of the dimension is turned to 1, if the browsing period
For the middle period, then the value numerical value of the dimension is turned to 2, if the browsing period is the late period, the value of the dimension is quantized
It is 3, the Time segments division of other forms can carry out numeralization processing with analogy.
(2) add up the browsing frequency: the behavior that same user browses identical product can be accumulated record, when user has browsed certain
After one product, to be accumulated be 1 to the browsing frequency, later, if after it browses the product again, the browsing frequency can be accumulated for
2, and so on.
It further, in the present embodiment, can be preparatory in order to avoid data excessively remote interfere current results
One data validity interval is set, and the user browsing behavior more than the data validity interval will no longer add up, and only have to the data
User browsing behavior in the effect phase is added up, which can be configured according to the actual situation, for example, can be with
It is set to one week, one month, a season, half a year, 1 year or other values.
This sentences a specific example and is illustrated:
All previous browsing serial number | The all previous browsing date | The accumulative browsing frequency |
1 | April 1 | 1 |
2 | April 13 | 2 |
3 | April 27 | 3 |
4 | May 4 | 3 |
5 | May 17 | 3 |
…… | …… | …… |
Assuming that data validity interval is 1 month, then after the 4th (May 4) browsing behavior, count from April 5 to May 4
The accumulative browsing frequency in day this month is 3, and the browsing behavior on April 1 has been more than the data validity interval, no longer to its into
Row is accumulative, and similarly, after the 5th (May 17) browsing behavior, statistics is accumulative in this month from April 18 to May 17
Browsing the frequency is 3, and the browsing behavior in April 1 and April 13 has been more than the data validity interval, is no longer added up to it.
(3) it browses duration: for browsing this data dimension of duration, and this browsing can be divided into and held
Continuous duration and accumulative browsing the two amounts of duration, generally, with the second as measurement unit.Wherein, this browsing is held
Continuous duration only records user to the duration of the browsing behavior of a certain product, and accumulative browsing duration then records user couple
The cumulative amount of all previous browsing behavior of a certain product.
For example, the duration that user browses a certain product for the first time is 20s, then when this browsing of user continues at this time
Long value is 20, and the accumulative browsing duration value of user is also 20;The duration that user browses the product for the second time is
30s, then this browsing duration value of user is 30 at this time, and the accumulative browsing duration value of user is then 50, according to
It is secondary to analogize.
Similarly, for browsing duration, the setting of above-mentioned data validity interval can be equally continued to use, is more than the number
It will no longer add up according to the user browsing behavior of validity period, only the user browsing behavior in the data validity interval be carried out tired
Meter.
This sentences a specific example and is illustrated:
Assuming that data validity interval is 1 month, then after the 4th (May 4) browsing behavior, count from April 5 to May 4
Accumulative browsing duration in this month of day is 55s, and the browsing behavior on April 1 has been more than the data validity interval, no longer
It is added up, similarly, after the 5th (May 17) browsing behavior, statistics is from April 18 to May 17 in this month
Accumulative browsing duration be 50s, and the browsing behavior in April 1 and April 13 has been more than the data validity interval, no longer right
It is added up.
Other than the above analyzes dimension, it is also necessary to obtain the browsing of user as a result, to judge the demand of user
Whether met, for user browses the behavior of a certain product on network, final browsing result can be divided into
It buys product and does not buy two kinds of situations of product.If after user has browsed a certain product, browsing result is to have purchased the product,
It may be considered that its demand has been met, no longer need to carry out subsequent browsing behavior analytic process, and if user browses
After a certain product, browsing result is not buy the product, then needs further to carry out browsing behavior to it by subsequent process
Analysis, to decide whether to execute the interactive operation with it and execute which kind of interactive operation.
It should be noted that the behavior of the browsing product for user each time, can all collect in its each analysis dimension
Browsing behavior data can acquire multiple browsing behavior data respectively if user has carried out multiple browsing to a certain product,
It is as follows:
Step S102, according to the browsing of user user described in the browsing behavior data configuration in each analysis dimension
Behavioural characteristic vector.
For example, the browsing behavior feature vector of the user can be constructed according to the following formula:
CrBehvVec=(CrData1,CrData2,CrData3,...CrDatad,...CrDataDN)
Wherein, d is the serial number of each analysis dimension, and 1≤d≤DN, DN are the sum for analyzing dimension, CrDatadIt is described
The browsing behavior data that user analyzes in dimension at d-th, CrBehvVec are the browsing behavior feature vector of the user.
Step S103, each historical viewings record in preset browsing behavior database is divided into more than two clear
Look at behavior group.
In order to accurately excavate the demand of user, the present embodiment can pre-establish other historical users' including magnanimity
The database namely the browsing behavior database of historical viewings record, carry out big data analysis to it, as to active user
The foundation predicted of behavior.
Every historical viewings record is made of the browsing behavior feature vector and browsing result of a historical viewings behavior.
Herein, the browsing behavior feature vector by each secondary historical viewings behavior is denoted as respectively:
BehvVecl=(Datal,1,Datal,2,Datal,3,...Datal,d,...Datal,DN)
Wherein, l is the serial number of each historical viewings record, and 1≤l≤LN, LN are going through in the browsing behavior database
The sum of history browsing record, Datal,dFor the l articles historical viewings record browsing behavior feature vector in d-th of analysis dimension
Component namely the l articles historical viewings be recorded in the browsing behavior data in d-th of analysis dimension, BehvVeclIt is gone through for the l articles
The browsing behavior feature vector of history browsing record.
In the present embodiment, the division of browsing behavior group can be completed by way of taking turns iteration more.
It goes firstly, being chosen in preset space coordinates GN o'clock respectively as each browsing after the 0th wheel group division
For core point namely the initial cores point of group.
The space coordinates are the space coordinates of a various dimensions, and total number of dimensions is DN, in the space coordinates
Each Spatial Dimension both correspond to an above-mentioned analysis dimension.
GN is the sum of browsing behavior group, and the specific location of each initial cores point can randomly select, but should be as far as possible
Guarantee that it is dispersed in space coordinates, the space length between any two initial cores point should be greater than it is preset away from
From threshold value.
Then, more wheel iterative calculation are carried out to the core point of each browsing behavior group and re-starts group division, directly
Until the stabilization core point for meeting preset decision condition.
As shown in Fig. 2, the detailed process of as t wheel (t is positive integer) group division:
Step S1031, the core point according to each browsing behavior group after t-1 wheel group division is clear to each article of history
Record of looking at carries out t wheel group division, obtains t wheel group division result.
It herein, can be by the core point of each browsing behavior group after t wheel group division in the space coordinates
Position is indicated with vector form are as follows:
CoreVect,g=(CtDatat,g,1,CtDatat,g,2,CtDatat,g,3,...CtDatat,g,d,
...CtDatat,g,DN)
Wherein, CtDatat,g,dCore point for g-th of browsing behavior group after t wheel group division is tieed up in d-th of analysis
Component on degree, CoreVect,gThe core point of g-th of browsing behavior group after group division is taken turns for t.
It is then possible to calculate separately the browsing behavior feature vector and t-1 of the l articles historical viewings record according to the following formula
Space length after wheel group division between the core point of each browsing behavior group:
Wherein, serial number of the g for each browsing behavior group, the serial number that 1≤g≤GN, l record for each historical viewings, 1≤
L≤LN, LN are the sum of the historical viewings record in the browsing behavior database, and d is the serial number of each analysis dimension, 1≤d
≤ DN, DN are the sum for analyzing dimension, Datal,dFor the l articles historical viewings record browsing behavior feature vector at d-th point
Analyse the component in dimension, CtDatat-1,g,dFor t-1 take turns group division after g-th of browsing behavior group core point in d
Component in a analysis dimension, CoreDist,g,lIt is taken turns for the browsing behavior feature vector and t-1 of the l articles historical viewings record
Space length after group division between the core point of each browsing behavior group.
Finally, can determine according to the following formula the l articles historical viewings be recorded in t wheel group division after belonging to browsing behavior
Group:
GroupSqt,l=Arg min (CoreDist,1,l,CoreDist,2,l,...,CoreDist,g,l,...,
CoreDist,GN,l) wherein, Argmin is minimum independent variable function, GroupSqt,lT, which is recorded in, for the l articles historical viewings takes turns group
The serial number of group browsing behavior group affiliated after dividing.
Step S1032, according to each browsing behavior group after t wheel group division result calculating t wheel group division
Core point.
For example, the core point of each browsing behavior group after t wheel group division can be calculated according to the following formula:
Wherein, s is the serial number of the historical viewings record in each browsing behavior group, 1≤s≤SpNumt,g, SpNumt,g
The sum of the historical viewings record after group division in g-th of browsing behavior group is taken turns for t, and
BehvVect,g.sBrowsing behavior for the s articles historical viewings record in g-th of browsing behavior group after t wheel group division is special
Levy vector.
Step S1033, judge whether the core point of each browsing behavior group after t wheel group division meets preset to sentence
Fixed condition.
The decision condition can indicate are as follows:
Wherein, Thresh is preset decision threshold, can be configured according to the actual situation to it, for example, can incite somebody to action
It is set as 10,50,100 or other values.
If being unsatisfactory for the decision condition, S1034 is thened follow the steps, if meeting the decision condition, is thened follow the steps
S1035。
Step S1034, it carries out t+1 and takes turns group division.
The process that t+1 takes turns group division is similar with the t wheel process of group division, and details are not described herein again.
Step S1035, terminate group division process.
After terminating group division, the core point of each browsing behavior group is as each after t is taken turns group division
The stabilization core point of a browsing behavior group.
The core point of each browsing behavior group is indicated in the position of the space coordinates with vector form herein are as follows:
StCoreVecg=(StCtDatag,1,StCtDatag,2,StCtDatag,3,...,StCtDatag,d,...,
StCtDatag,DN)
Wherein, StCtDatag,dDividing in dimension is analyzed at d-th for the stabilization core point of g-th of browsing behavior group
Amount, StCoreVecgFor the stabilization core point of g-th of browsing behavior group.
Step S104, analysis probability is calculated according to the browsing behavior feature vector of each browsing behavior group and the user
Value.
It is the probability value of preset first result that the analysis probability value, which is the browsing result of the user,.Specifically, may be used
To calculate the analysis probability value according to the following formula:
Wherein, PosNumgTo browse the historical viewings record that result is first result in g-th of browsing behavior group
Sum, first result is the product for having purchased browsing, NegNumgIt is to browse result in g-th of browsing behavior group
The sum of the historical viewings record of preset second result, second result is not buy the product of browsing, and Prob is described
Analyze probability value.
Step S105, the interactive operation with the user is executed according to the analysis probability value.
If the analysis probability value is greater than preset first probability threshold value, it is believed that the user is to buy diving for product
In client, can then precision marketing be carried out to it at this time, conversely, not marketing to it then.First probability threshold value can basis
Actual conditions are configured it, for example, 50%, 60%, 70% or other values can be set to.
Further, for the potential customers of different analysis probability values, different marketing modes can be used, for example, if
The analysis probability value of a certain potential customers be greater than preset second probability threshold value, then can to it by the way of telemarketing,
It otherwise, then can be to it by the way of short message marketing.Second probability threshold value is greater than the first probability threshold value, can be according to reality
Situation is configured it, for example, 80%, 85%, 90% or other values can be set to.
In conclusion the embodiment of the present invention obtains first when user's browsing objective object in webpage or application program
Take browsing behavior data of the family in preset each analysis dimension, and construct accordingly the browsing behavior feature of user to
Then each historical viewings record in preset browsing behavior database is divided into more than two browsing behavior groups by amount
Group, using these historical datas as the foundation for carrying out big data analysis to user, according to each browsing behavior group and user
The browsing result that browsing behavior feature vector calculates user is the probability value of preset first result, that is, passes through the browsing to user
Behavior carries out sufficient mining analysis, understands the demand of user, and friendship corresponding with user's progress according to the demand of user in time
Mutually, to meet the needs of users as far as possible, thus the significant increase experience of user.
It should be understood that the size of the serial number of each step is not meant that the order of the execution order in above-described embodiment, each process
Execution sequence should be determined by its function and internal logic, the implementation process without coping with the embodiment of the present invention constitutes any limit
It is fixed.
It is shown corresponding to a kind of user browsing behavior analysis method based on big data, Fig. 3 described in foregoing embodiments
A kind of one embodiment structure chart of user browsing behavior analytical equipment provided in an embodiment of the present invention.
In the present embodiment, a kind of user browsing behavior analytical equipment may include:
Browsing behavior data acquisition module 301, for obtaining browsing behavior of the user in preset each analysis dimension
Data, the browsing behavior data are generated behavioral data when browsing objective object in webpage or application program;
Feature vector constructing module 302, for the browsing behavior data structure according to the user in each analysis dimension
Make the browsing behavior feature vector of the user;
Group division module 303, for each historical viewings record in preset browsing behavior database to be divided into
More than two browsing behavior groups, wherein every historical viewings record is special by the browsing behavior of a historical viewings behavior
Levy vector and browsing result composition;
Probability evaluation entity 304, for the browsing behavior feature vector according to each browsing behavior group and the user
Analysis probability value is calculated, it is the probability value of preset first result that the analysis probability value, which is the browsing result of the user,;
Interactive operation execution module 305, for executing the interactive operation with the user according to the analysis probability value.
Further, the group division module may include:
Initial cores point selection unit, for choosing in preset space coordinates at GN o'clock respectively as the 0th wheel group
The core point of each browsing behavior group after group divides, GN are the sum of browsing behavior group;
Group division unit, for according to t-1 take turns group division after each browsing behavior group core point to each article
Historical viewings record carries out t and takes turns group division, obtains t wheel group division as a result, t is positive integer;
Core point computing unit is gone for taking turns each browsing after group division result calculates t wheel group division according to t
For the core point of group;
First processing units, if the core point for each browsing behavior group after t wheel group division is unsatisfactory for presetting
Decision condition, then execute t+1 wheel group division;
The second processing unit, if for t take turns group division after each browsing behavior group core point meet described in sentence
Fixed condition, then terminate group division process, and after t is taken turns group division each browsing behavior group core point as
The stabilization core point of each browsing behavior group.
Further, the group division unit may include:
Space length computation subunit, the browsing behavior for calculating separately the l articles historical viewings record according to the following formula are special
Space length after sign vector and t-1 wheel group division between the core point of each browsing behavior group:
Wherein, serial number of the g for each browsing behavior group, the serial number that 1≤g≤GN, l record for each historical viewings, 1≤
L≤LN, LN are the sum of the historical viewings record in the browsing behavior database, and d is the serial number of each analysis dimension, 1≤d
≤ DN, DN are the sum for analyzing dimension, Datal,dFor the l articles historical viewings record browsing behavior feature vector at d-th point
Analyse the component in dimension, CtDatat-1,g,dFor t-1 take turns group division after g-th of browsing behavior group core point in d
Component in a analysis dimension, CoreDist,g,lIt is taken turns for the browsing behavior feature vector and t-1 of the l articles historical viewings record
Space length after group division between the core point of each browsing behavior group;
Group determines subelement, affiliated after the l articles historical viewings is recorded in t wheel group division for determining according to the following formula
Browsing behavior group:
GroupSqt,l=Arg min (CoreDist,1,l,CoreDist,2,l,...,CoreDist,g,l,...,
CoreDist,GN,l) wherein, Argmin is minimum independent variable function, GroupSqt,lT, which is recorded in, for the l articles historical viewings takes turns group
The serial number of group browsing behavior group affiliated after dividing.
Further, the core point computing unit be specifically used for calculate according to the following formula t wheel group division after it is each clear
Look at the core point of behavior group:
Wherein, s is the serial number of the historical viewings record in each browsing behavior group, 1≤s≤SpNumt,g, SpNumt,g
The sum of the historical viewings record after group division in g-th of browsing behavior group is taken turns for t, and
BehvVect,g.sBrowsing behavior for the s articles historical viewings record in g-th of browsing behavior group after t wheel group division is special
Levy vector, CoreVect,gThe core point of g-th of browsing behavior group after group division is taken turns for t.
Further, the probability evaluation entity is specifically used for calculating the analysis probability value according to the following formula:
Wherein, CrDatadFor the user browsing behavior feature vector d-th analyze dimension on component,
StCtDatag,dFor the component that the stabilization core point of g-th of browsing behavior group is analyzed in dimension at d-th, PosNumgFor g
The sum that the historical viewings that result is first result record, NegNum are browsed in a browsing behavior groupgIt is browsed for g-th
The sum that the historical viewings that result is preset second result record is browsed in behavior group, Prob is the analysis probability value.
It is apparent to those skilled in the art that for convenience and simplicity of description, the device of foregoing description,
The specific work process of module and unit, can refer to corresponding processes in the foregoing method embodiment, and details are not described herein.
In the above-described embodiments, it all emphasizes particularly on different fields to the description of each embodiment, is not described in detail or remembers in some embodiment
The part of load may refer to the associated description of other embodiments.
The schematic block diagram that Fig. 4 shows a kind of terminal device provided in an embodiment of the present invention is only shown for ease of description
Part related to the embodiment of the present invention.
In the present embodiment, the terminal device 4 can be desktop PC, notebook, palm PC and cloud clothes
Business device etc. calculates equipment.The terminal device 4 can include: processor 40, memory 41 and be stored in the memory 41 simultaneously
The computer-readable instruction 42 that can be run on the processor 40, such as execute the above-mentioned user based on big data and browse row
For the computer-readable instruction of analysis method.The processor 40 is realized above-mentioned each when executing the computer-readable instruction 42
Step in user browsing behavior analysis method embodiment based on big data, such as step S101 to S105 shown in FIG. 1.Or
Person, the processor 40 realize each module/unit in above-mentioned each Installation practice when executing the computer-readable instruction 42
Function, such as the function of module 301 to 305 shown in Fig. 3.
Illustratively, the computer-readable instruction 42 can be divided into one or more module/units, one
Or multiple module/units are stored in the memory 41, and are executed by the processor 40, to complete the present invention.Institute
Stating one or more module/units can be the series of computation machine readable instruction section that can complete specific function, the instruction segment
For describing implementation procedure of the computer-readable instruction 42 in the terminal device 4.
The processor 40 can be central processing unit (Central Processing Unit, CPU), can also be
Other general processors, digital signal processor (Digital Signal Processor, DSP), specific integrated circuit
(Application Specific Integrated Circuit, ASIC), field programmable gate array (Field-
Programmable Gate Array, FPGA) either other programmable logic device, discrete gate or transistor logic,
Discrete hardware components etc..General processor can be microprocessor or the processor is also possible to any conventional processor
Deng.
The memory 41 can be the internal storage unit of the terminal device 4, such as the hard disk or interior of terminal device 4
It deposits.The memory 41 is also possible to the External memory equipment of the terminal device 4, such as be equipped on the terminal device 4
Plug-in type hard disk, intelligent memory card (Smart Media Card, SMC), secure digital (Secure Digital, SD) card dodge
Deposit card (Flash Card) etc..Further, the memory 41 can also both include the storage inside list of the terminal device 4
Member also includes External memory equipment.The memory 41 is for storing the computer-readable instruction and the terminal device 4
Required other instruction and datas.The memory 41 can be also used for temporarily storing the number that has exported or will export
According to.
The functional units in various embodiments of the present invention may be integrated into one processing unit, is also possible to each
Unit physically exists alone, and can also be integrated in one unit with two or more units.Above-mentioned integrated unit both may be used
To use formal implementation of hardware, can also realize in the form of software functional units.
If the integrated unit is realized in the form of SFU software functional unit and sells or use as independent product
When, it can store in a computer readable storage medium.Based on this understanding, technical solution of the present invention substantially or
Person says that all or part of the part that contributes to existing technology or the technical solution can body in the form of software products
Reveal and, which is stored in a storage medium, including several computer-readable instructions are used so that one
Platform computer equipment (can be personal computer, server or the network equipment etc.) executes described in each embodiment of the present invention
The all or part of the steps of method.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory (ROM, Read-
Only Memory), random access memory (RAM, Random Access Memory), magnetic or disk etc. are various can be with
Store the medium of computer-readable instruction.
Embodiment described above is merely illustrative of the technical solution of the present invention, rather than its limitations;Although referring to aforementioned reality
Applying example, invention is explained in detail, those skilled in the art should understand that: it still can be to aforementioned each
Technical solution documented by embodiment is modified or equivalent replacement of some of the technical features;And these are modified
Or replacement, the spirit and scope for technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution.
Claims (10)
1. a kind of user browsing behavior analysis method based on big data characterized by comprising
Obtain browsing behavior data of the user in preset each analysis dimension, the browsing behavior data be in webpage or
Generated behavioral data when browsing objective object in application program;
According to the browsing behavior feature vector of user user described in the browsing behavior data configuration in each analysis dimension;
Each historical viewings record in preset browsing behavior database is divided into more than two browsing behavior groups,
In, every historical viewings record is made of the browsing behavior feature vector and browsing result of a historical viewings behavior;
Analysis probability value is calculated according to the browsing behavior feature vector of each browsing behavior group and the user, the analysis is general
Rate value is the probability value that the browsing result of the user is preset first result;
The interactive operation with the user is executed according to the analysis probability value.
2. user browsing behavior analysis method according to claim 1, which is characterized in that described by preset browsing behavior
Each historical viewings record in database is divided into more than two browsing behavior groups and includes:
The GN o'clock core respectively as each browsing behavior group after the 0th wheel group division is chosen in preset space coordinates
Heart point, GN are the sum of browsing behavior group;
T wheel is carried out to each article of historical viewings record according to the core point of each browsing behavior group after t-1 wheel group division
Group division obtains t wheel group division as a result, t is positive integer;
The core point that group division result calculates each browsing behavior group after t wheel group division is taken turns according to t;
If the core point of each browsing behavior group is unsatisfactory for preset decision condition after t wheel group division, t+1 is executed
Take turns group division;
If the core point of each browsing behavior group meets the decision condition after t wheel group division, terminate group division
Process, and after t is taken turns group division each browsing behavior group core point as the steady of each browsing behavior group
Determine core point.
3. user browsing behavior analysis method according to claim 2, which is characterized in that described to take turns group according to t-1
The core point of each browsing behavior group includes: to each article of historical viewings record progress t wheel group division after division
It calculates separately according to the following formula each after the browsing behavior feature vector and t-1 wheel group division of the l articles historical viewings record
Space length between the core point of a browsing behavior group:
Wherein, serial number of the g for each browsing behavior group, the serial number that 1≤g≤GN, l record for each historical viewings, 1≤l≤
LN, LN be the browsing behavior database in historical viewings record sum, d be it is each analysis dimension serial number, 1≤d≤
DN, DN are the sum for analyzing dimension, Datal,dBrowsing behavior feature vector for the l articles historical viewings record is analyzed at d-th
Component in dimension, CtDatat-1,g,dFor t-1 take turns group division after g-th of browsing behavior group core point at d-th
Analyze the component in dimension, CoreDist,g,lGroup is taken turns for the browsing behavior feature vector and t-1 of the l articles historical viewings record
Space length after group divides between the core point of each browsing behavior group;
Browsing behavior group belonging to being determined after the l articles historical viewings is recorded in t wheel group division according to the following formula:
GroupSqt,l=Argmin (CoreDist,1,l,CoreDist,2,l,...,CoreDist,g,l,...,CoreDist,GN,l)
Wherein, Argmin is minimum independent variable function, GroupSqt,lAfter being recorded in t wheel group division for the l articles historical viewings
The serial number of affiliated browsing behavior group.
4. user browsing behavior analysis method according to claim 2, which is characterized in that described to take turns group stroke according to t
The core point that point result calculates each browsing behavior group after t wheel group division includes:
The core point that t takes turns each browsing behavior group after group division is calculated according to the following formula:
Wherein, s is the serial number of the historical viewings record in each browsing behavior group, 1≤s≤SpNumt,g, SpNumt,gFor t
The sum of historical viewings record after wheel group division in g-th of browsing behavior group, and
BehvVect,g.sBrowsing behavior for the s articles historical viewings record in g-th of browsing behavior group after t wheel group division is special
Levy vector, CoreVect,gThe core point of g-th of browsing behavior group after group division is taken turns for t.
5. user browsing behavior analysis method according to any one of claim 1 to 4, which is characterized in that the basis
The browsing behavior feature vector of each browsing behavior group and the user calculate analysis probability value
The analysis probability value is calculated according to the following formula:
Wherein, CrDatadFor the component that the browsing behavior feature vector of the user is analyzed in dimension at d-th, StCtDatag,d
For the component that the stabilization core point of g-th of browsing behavior group is analyzed in dimension at d-th, PosNumgFor g-th of browsing behavior
The sum that the historical viewings that result is first result record, NegNum are browsed in groupgFor in g-th of browsing behavior group
The sum that the historical viewings that result is preset second result record is browsed, Prob is the analysis probability value.
6. a kind of user browsing behavior analytical equipment characterized by comprising
Browsing behavior data acquisition module, for obtaining browsing behavior data of the user in preset each analysis dimension, institute
Stating browsing behavior data is generated behavioral data when browsing objective object in webpage or application program;
Feature vector constructing module, for according to the user it is each analysis dimension on browsing behavior data configuration described in use
The browsing behavior feature vector at family;
Group division module, for each historical viewings record in preset browsing behavior database to be divided into two or more
Browsing behavior group, wherein every historical viewings record by a historical viewings behavior browsing behavior feature vector and
Browse result composition;
Probability evaluation entity is analyzed for being calculated according to the browsing behavior feature vector of each browsing behavior group and the user
Probability value, it is the probability value of preset first result that the analysis probability value, which is the browsing result of the user,;
Interactive operation execution module, for executing the interactive operation with the user according to the analysis probability value.
7. user browsing behavior analytical equipment according to claim 6, which is characterized in that the group division module packet
It includes:
Initial cores point selection unit is drawn for choosing in preset space coordinates at GN o'clock respectively as the 0th wheel group
The core point of each browsing behavior group after point, GN are the sum of browsing behavior group;
Group division unit, for according to t-1 take turns group division after each browsing behavior group core point to each article of history
Browsing record carries out t and takes turns group division, obtains t wheel group division as a result, t is positive integer;
Core point computing unit, for according to each browsing behavior group after t wheel group division result calculating t wheel group division
The core point of group;
First processing units, if being unsatisfactory for preset sentencing for the t core point for taking turns each browsing behavior group after group division
Fixed condition then executes t+1 wheel group division;
The second processing unit, if the core point for each browsing behavior group after t wheel group division meets the judgement article
Part, then terminate group division process, and after t is taken turns group division each browsing behavior group core point as each
The stabilization core point of browsing behavior group.
8. user browsing behavior analytical equipment according to claim 7, which is characterized in that the group division unit packet
It includes:
Space length computation subunit, for calculate separately according to the following formula the l articles historical viewings record browsing behavior feature to
Space length after amount and t-1 wheel group division between the core point of each browsing behavior group:
Wherein, serial number of the g for each browsing behavior group, the serial number that 1≤g≤GN, l record for each historical viewings, 1≤l≤
LN, LN be the browsing behavior database in historical viewings record sum, d be it is each analysis dimension serial number, 1≤d≤
DN, DN are the sum for analyzing dimension, Datal,dBrowsing behavior feature vector for the l articles historical viewings record is analyzed at d-th
Component in dimension, CtDatat-1,g,dFor t-1 take turns group division after g-th of browsing behavior group core point at d-th
Analyze the component in dimension, CoreDist,g,lGroup is taken turns for the browsing behavior feature vector and t-1 of the l articles historical viewings record
Space length after group divides between the core point of each browsing behavior group;
Group determines subelement, for determine according to the following formula the l articles historical viewings be recorded in t wheel group division after belonging to it is clear
Look at behavior group:
GroupSqt,l=Argmin (CoreDist,1,l,CoreDist,2,l,...,CoreDist,g,l,...,CoreDist,GN,l)
Wherein, Argmin is minimum independent variable function, GroupSqt,lAfter being recorded in t wheel group division for the l articles historical viewings
The serial number of affiliated browsing behavior group.
9. a kind of computer readable storage medium, the computer-readable recording medium storage has computer-readable instruction, special
Sign is, realizes that the user as described in any one of claims 1 to 5 is clear when the computer-readable instruction is executed by processor
Look at behavior analysis method the step of.
10. a kind of terminal device, including memory, processor and storage are in the memory and can be on the processor
The computer-readable instruction of operation, which is characterized in that the processor realizes such as right when executing the computer-readable instruction
It is required that described in any one of 1 to 5 the step of user browsing behavior analysis method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910295880.2A CN110162697A (en) | 2019-04-12 | 2019-04-12 | User browsing behavior analysis method, device, medium and terminal based on big data |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910295880.2A CN110162697A (en) | 2019-04-12 | 2019-04-12 | User browsing behavior analysis method, device, medium and terminal based on big data |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110162697A true CN110162697A (en) | 2019-08-23 |
Family
ID=67639347
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910295880.2A Pending CN110162697A (en) | 2019-04-12 | 2019-04-12 | User browsing behavior analysis method, device, medium and terminal based on big data |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110162697A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111241081A (en) * | 2020-01-09 | 2020-06-05 | 杭州涂鸦信息技术有限公司 | IOT platform data collection method and system, readable storage medium and computer equipment |
-
2019
- 2019-04-12 CN CN201910295880.2A patent/CN110162697A/en active Pending
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111241081A (en) * | 2020-01-09 | 2020-06-05 | 杭州涂鸦信息技术有限公司 | IOT platform data collection method and system, readable storage medium and computer equipment |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111523976B (en) | Commodity recommendation method and device, electronic equipment and storage medium | |
US10417650B1 (en) | Distributed and automated system for predicting customer lifetime value | |
US10453118B2 (en) | Custom user definable keyword bidding system and method | |
JP6412550B2 (en) | System, method and apparatus for implementing data upload, processing and predictive query API publication | |
CN107705183A (en) | Recommendation method, apparatus, storage medium and the server of a kind of commodity | |
CN107391692A (en) | The appraisal procedure and device of a kind of recommendation effect | |
US20110071900A1 (en) | Advertisee-history-based bid generation system and method for multi-channel advertising | |
CN103116582B (en) | A kind of information retrieval method and related system and device | |
CN109961198B (en) | Associated information generation method and device | |
Yeo et al. | Conversion prediction from clickstream: Modeling market prediction and customer predictability | |
WO2020221022A1 (en) | Service object recommendation method | |
US20140337275A1 (en) | Method and system for scaling panels | |
US20210192549A1 (en) | Generating analytics tools using a personalized market share | |
CN112380449B (en) | Information recommendation method, model training method and related device | |
CN112149003B (en) | Commodity community recommendation method and device and computer equipment | |
CN105335518A (en) | Method and device for generating user preference information | |
US20210224351A1 (en) | Method and system for optimizing an objective having discrete constraints | |
CN113065882A (en) | Commodity processing method and device and electronic equipment | |
CN108921587B (en) | Data processing method and device and server | |
US10467654B2 (en) | Forecasting customer channel choice using cross-channel loyalty | |
Jiang et al. | Intertemporal pricing via nonparametric estimation: Integrating reference effects and consumer heterogeneity | |
CN109978594B (en) | Order processing method, device and medium | |
CN110162697A (en) | User browsing behavior analysis method, device, medium and terminal based on big data | |
CN110310162B (en) | Sample generation method and device | |
Koniew | Classification of the User's Intent Detection in Ecommerce systems-Survey and Recommendations. |
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 |