CN104102649A - Method and device for grading website users - Google Patents
Method and device for grading website users Download PDFInfo
- Publication number
- CN104102649A CN104102649A CN201310117776.7A CN201310117776A CN104102649A CN 104102649 A CN104102649 A CN 104102649A CN 201310117776 A CN201310117776 A CN 201310117776A CN 104102649 A CN104102649 A CN 104102649A
- Authority
- CN
- China
- Prior art keywords
- website user
- long value
- normal distribution
- website
- user
- 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
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/958—Organisation or management of web site content, e.g. publishing, maintaining pages or automatic linking
Abstract
The embodiment of the invention discloses a method and a device for grading website users, wherein the grading method comprises the following steps that when the growth values of the website users conform to normal distribution, the value range of the growth values of the website users in each grade in the normal distribution is determined according to the preset generating probability of the website users in each grade; the growth values of the website users to be graded are read; the grade of the website users to be graded is determined according to the value range of the growth values of the website users to be graded. The method and the device according to the embodiment of the invention have the advantage that the stability and the justification of the grading can be ensured.
Description
Technical field
The present invention relates to computer application field, particularly relate to a kind of method and apparatus that website user is carried out to classification.
Background technology
In some WEB application systems, for example, e-commerce website, SNS social network sites or instant message application website etc., be all that website user is provided with into long value, and set website user's grade based on this one-tenth long value, so-called " one-tenth long value " is the concrete numerical value for presentation-entity object with feature in a certain respect or comprehensive characteristics, and meanwhile, this numerical value has in time and the feature that may change.For example, this feature can be: behavior experience, outstanding degree, ability height, accumulation score value, total mark of consumption or aggregative index etc.For the website user of different brackets, website except meeting provides the service and right of differentiated, as, the senior member of Taobao website can enjoy higher discount, website also can be concentrated as each grade Resources allocation (as, user profile storage resources).As, after website user's classification is elementary, intermediate and senior, website is that all elementary website users distribute 2 servers for storing the user profile of all website users under this grade, similarly, for all intermediate website users distribute 4 servers for storing the user profile of all website users under this grade, for all senior website users distribute 1 server for storing the user profile of all website users under this grade.
At present, having a kind of hierarchy plan is to adopt the mode that is fixed into long value interval to carry out classification to website user.For example, website user is divided into three grades by WEB server, and set elementary website user's one-tenth long value interval for [0,100), intermediate website user's one-tenth long value interval be [100,200), senior website user's one-tenth long value interval is [200,400].Visible, under this hierarchy plan, the growth value interval of each rank is fixed.But, in the practical application of this scheme, likely there will be so a kind of classification results, over time, most of website user has been gathered in a grade, that is, and and due to the instability between each classification, cause embodying each website user's grade difference, also cannot embody the effect of classification.
And, if a large amount of website users has been gathered in a grade, and only have a small amount of website user to be gathered in other grade, will cause the storage space of some servers by a large amount of leaving unused, and the unbalance situation that the insufficient memory of other servers is used occurs, therefore, caused storage resources to use unbalanced.
Therefore, a problem in the urgent need to address is at present: how to ensure the stability of classification, realize storage resources and use unbalanced.
Summary of the invention
In order to solve the problems of the technologies described above, the embodiment of the present invention provides a kind of method and apparatus that website user is carried out to classification, to ensure the stability of classification, realizes storage resources and uses unbalanced.
The embodiment of the invention discloses following technical scheme:
A method of website user being carried out to classification, comprising:
In the time of website user's one-tenth long value Normal Distribution, the numerical range according to default each grade website user's the each grade website user's of probability of happening ratio-dependent one-tenth long value in described normal distribution;
Read website user's to be fractionated one-tenth long value;
The numerical range falling into according to website user's to be fractionated one-tenth long value, determines website user's to be fractionated grade.
Preferably, also comprise:
According to classification results, it is website user's distributing user information storage space of different brackets.
Preferably, also comprise:
Judge website user's whether Normal Distribution of one-tenth long value;
If so, enter classification process;
Otherwise, the distribution pattern of described website user's one-tenth long value is converted to normal distribution, enter classification process.
Preferably, the described distribution pattern by website user's one-tenth long value is converted to normal distribution and comprises:
If described website user's one-tenth long value obeys logarithm normal distribution, does log-transformation processing by described website user's one-tenth long value, make the one-tenth long value Normal Distribution of the website user after conversion process.
Preferably, the described distribution pattern by website user's one-tenth long value is converted to distributions and comprises:
If described website user's one-tenth long value is obeyed Poisson distribution or slight skewness distributes, described website user's one-tenth long value is done to square root transformation processing, make the one-tenth long value Normal Distribution of the website user after conversion process.
Preferably, the described distribution pattern by website user's one-tenth long value is converted to distributions and comprises:
If described website user's one-tenth long value is obeyed data two ends, fluctuation larger data distributes, and described website user's one-tenth long value is done to conversion process reciprocal, makes the one-tenth long value Normal Distribution of the website user after conversion process.
Preferably, the described distribution pattern by website user's one-tenth long value is converted to distributions and comprises:
If described website user's one-tenth long value is obeyed ratio or number percent data distribute, described website user's one-tenth long value is first done to square root transformation and do again arcsine transformation, make the one-tenth long value Normal Distribution of the website user after conversion process.
A device that website user is carried out to classification, comprising:
Determination module, for when website user's the one-tenth long value Normal Distribution, the numerical range according to default each grade website user's the each grade website user's of probability of happening ratio-dependent one-tenth long value in described normal distribution;
Read module, for reading website user's to be fractionated one-tenth long value;
Diversity module, for the numerical range falling into according to website user's to be fractionated one-tenth long value, determines website user's to be fractionated grade.
Preferably, also comprise:
Resource distribution module, for according to classification results, is website user's distributing user information storage space of different brackets.
Preferably, also comprise;
Judge module, distributes for judging whether website user's one-tenth long value obeys just too;
Type conversion module, when being no in the judged result of described judge module, is converted to normal distribution by the distribution pattern of described website user's one-tenth long value;
Described determination module, in the judged result of described judge module when being, the numerical range of the one-tenth long value of determining each grade website user according to default each grade website user's probability in described normal distribution.
Preferably, if described website user's one-tenth long value obeys logarithm normal distribution, described type conversion module is:
The first conversion module, for described website user's one-tenth long value is done to log-transformation processing, makes the one-tenth long value Normal Distribution of the website user after conversion process.
Preferably, if described website user's one-tenth long value is obeyed Poisson distribution or slight skewness distributes, described type conversion module is:
The second conversion module, for described website user's one-tenth long value is done to square root transformation processing, makes the one-tenth long value Normal Distribution of the website user after conversion process.
Preferably, if described website user's one-tenth long value is obeyed data two ends, fluctuation larger data distributes, and described type conversion module is:
The 3rd conversion module, for described website user's one-tenth long value is done to conversion process reciprocal, makes the one-tenth long value Normal Distribution of the website user after conversion process.
Preferably, if described website user's one-tenth long value is obeyed ratio or number percent data distribute, described type conversion module is:
The 4th conversion module, does arcsine transformation again for described website user's one-tenth long value is first done to square root transformation, makes the one-tenth long value Normal Distribution of the website user after conversion process.
As can be seen from the above-described embodiment, utilize the feature of website user's one-tenth long value Normal Distribution, website user is divided in different brackets according to the probability of happening of specifying, according to the principle of theory of probability, the ratio that makes each grade website user account for all website users can maintain in the scope of probability of happening of this appointment, has ensured the stability of classification.Based on this classification results, in the time being website user's memory allocated resource of each grade, also can ensure that storage resources uses balanced.
Simultaneously, owing to being the grade of determining website user according to the size that becomes long value, also there will not be into two website users that long value differs greatly and be divided into the situation in same rank and become two identical website users of long value to be divided into the situation of different stage, also ensured the fairness of classification.
Brief description of the drawings
In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, to the accompanying drawing of required use in embodiment or description of the Prior Art be briefly described below, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skill in the art, do not paying under the prerequisite of creative work, can also obtain according to these accompanying drawings other accompanying drawing.
Fig. 1 is a kind of process flow diagram that website user is carried out to the method for classification that the embodiment of the present invention one discloses;
Fig. 2 is a normal distribution curve figure who has divided Website user rank in the present invention;
Fig. 3 is the operation chart of the numerical range of one-tenth long value in described normal distribution of inquiring about each grade website user according to standardized normal distribution probability tables in the present invention;
Fig. 4 is a kind of process flow diagram that website user is carried out to the method for classification that the embodiment of the present invention two discloses;
Fig. 5 is a kind of process flow diagram that website user is carried out to the method for classification that the embodiment of the present invention three discloses;
Fig. 6 is a kind of structural drawing that website user is carried out to the device of classification that the embodiment of the present invention four discloses;
Fig. 7 is another kind that the embodiment of the present invention four discloses carries out the device of classification structural drawing to website user.
Embodiment
The embodiment of the present invention provides the method and the corresponding intrument that website user are carried out to classification.Utilize the feature of website user's one-tenth long value Normal Distribution, according to the principle of normal distribution, in normal distribution curve, horizontal ordinate is website user's one-tenth long value, the area of the closed region that normal distribution curve and abscissa axis surround be gradational website user's probability of happening (gradational website user probability of happening be 1).Preset each grade website user's probability of happening,, above-mentioned closed region is divided into multiple zonules by probability of happening ratio according to the number of grade and each grade, the area of each zonule is exactly a grade website user's probability of happening, and abscissa zone corresponding to each zonule is this grade website user's one-tenth long value interval.
For above-mentioned purpose of the present invention, feature and advantage can be become apparent more, below in conjunction with accompanying drawing, the embodiment of the present invention is described in detail.
Embodiment mono-
Refer to Fig. 1, a kind of process flow diagram that website user is carried out to the method for classification that it discloses for the embodiment of the present invention one, the method comprises the following steps:
Step 101: in the time of website user's one-tenth long value Normal Distribution, the numerical range according to default each grade website user's the each grade website user's of probability of happening ratio-dependent one-tenth long value in described normal distribution;
According to the principle of normal distribution, in normal distribution curve, horizontal ordinate is website user's one-tenth long value, the area of the closed region that normal distribution curve and abscissa axis surround be gradational website user's probability of happening (gradational website user probability of happening be 1).Technical scheme for convenience of description, the grade of supposing to preset website user is 3 grades (being elementary website user, intermediate website user and senior website user) altogether, and the probability of happening of each grade is all identical, the probability of happening ratio of each grade is all about 33%.As shown in Figure 2, the area of the closed region that normal distribution curve and abscissa axis are surrounded is divided into 3 zonules, the area of each zonule is exactly each grade website user's probability of happening, and abscissa zone corresponding to each zonule is this grade website user's one-tenth long value interval.
It should be noted that, website user's number of levels and each grade website user's probability of happening can arrange arbitrarily according to actual user demand, and the present invention does not limit its concrete numerical value.Can also revise arbitrarily user's number of levels and each grade website user's probability of happening according to user's actual user demand.
As shown in Figure 3, known by query criteria normal distribution probability table, in standardized normal distribution, elementary website user's one-tenth long value interval is (∞,-0.43], intermediate website user's one-tenth long value interval be (0.43 ,+0.43], senior website user's one-tenth long value interval is (+0.43 ,+∞).
Step 102: the one-tenth long value that reads website user to be fractionated;
In database, preserve each website user's one-tenth long value, WEB application, in the time that website user is carried out to classification, can directly read each website user's one-tenth long value from database.
Step 103: the numerical range falling into according to website user's to be fractionated one-tenth long value, determine website user's to be fractionated grade.
What obey due to website user's one-tenth long value is normal distribution, and in above-mentioned steps 101 according to standardized normal distribution probability tables determine be the one-tenth long value interval of website users at different levels in standardized normal distribution, therefore, be transfer principle according to mathematical coordinates, need first normal distribution coordinate system to be converted to standardized normal distribution coordinate system.After conversion, can draw:
Wherein, Y
ifor the one-tenth long value of i website user in standardized normal distribution, E
ifor the one-tenth long value of i website user in normal distribution, the average one-tenth long value that μ is website user, the standard deviation of the one-tenth long value that σ is website user.
E
ibe i website user's one-tenth long value, i=1,2,3...n, n is natural number.
For example, after above-mentioned conversion, obtain the one-tenth long value Y of some website users in standardized normal distribution
1=+2.56, this website user's one-tenth long value falls into the numerical range of senior website user's one-tenth long value, and the grade that can determine this website user is senior.
Certainly, when predetermined be the one-tenth long value interval of website users at different levels in normal distribution, carry out above-mentioned coordinate system conversion with regard to no longer needing, but directly determine the numerical range that website user's to be fractionated one-tenth long value falls into.
It should be noted that, WEB application is each website user's one-tenth long value in reading database one by one, and determines one by one the numerical value interval that each website user's one-tenth long value falls into.In addition, WEB application can regularly be carried out above-mentioned classification process, as, every night, realizes the real-time update of Website user rank.
Also it should be noted that, in the present invention, do not limit the execution sequence of step 101 and step 102, that is to say, except can first step 101 performing step again 102, can also first perform step 102 and perform step again 101, or, step 101 and step 102 executed in parallel.This can not affect the realization of technical solution of the present invention.
After having determined each website user's to be fractionated grade, this programme also further comprises: according to classification results, be website user's distributing user information storage space of different brackets.
As can be seen from the above-described embodiment, utilize the feature of website user's one-tenth long value Normal Distribution, website user is divided in different brackets according to the probability of happening of specifying, according to the principle of theory of probability, the ratio that makes each grade website user account for all website users can maintain in the scope of probability of happening of this appointment, has ensured the stability of classification.Based on this classification results, in the time being website user's memory allocated resource of each grade, also can ensure that storage resources uses balanced.
Simultaneously, owing to being the grade of determining website user according to the size that becomes long value, also there will not be into two website users that long value differs greatly and be divided into the situation in same rank and become two identical website users of long value to be divided into the situation of different stage, also ensured the fairness of classification.
Embodiment bis-
The difference of the present embodiment and embodiment mono-is, for website user's the one-tenth long value of disobeying normal distribution, need to before classification, first its distribution pattern be converted to normal distribution.Refer to Fig. 4, a kind of process flow diagram that website user is carried out to the method for classification that it discloses for the embodiment of the present invention two, the method comprises the following steps:
Step 401: judge website user's whether Normal Distribution of one-tenth long value, if so, enter step 402, otherwise, enter step 405;
Conventionally, in dissimilar website, the distribution pattern of website user's one-tenth long value is also different, rule of thumb, can determine the distribution pattern of the one-tenth long value of website user in all types of websites.Once determine the Type of website, can judge the whether Normal Distribution of one-tenth long value of website user in the type website.
Step 402: the numerical range of the one-tenth long value of determining each grade website user according to default each grade website user's probability proportion in described normal distribution;
Step 403: the one-tenth long value that reads website user to be fractionated;
Step 404: the numerical range falling into according to website user's to be fractionated one-tenth long value, determine website user's to be fractionated grade, process ends;
Step 405, is converted to normal distribution by the distribution pattern of described website user's one-tenth long value, enters step 402.
The concrete implementation of above-mentioned steps 403-405 can be referring to the explanation of the step 101-103 in embodiment mono-, and the present embodiment repeats no more.
According to existing theory of probability theory, can utilize the transformational relation between normal distribution and skewed distribution, skewed distribution is converted to normal distribution.Wherein,
If website user's one-tenth long value obeys logarithm normal distribution, does log-transformation processing by this website user's one-tenth long value, make the one-tenth long value Normal Distribution of the website user after conversion process.
If website user's one-tenth long value is obeyed Poisson distribution or slight skewness distributes, this website user's one-tenth long value is done to square root transformation, make the one-tenth long value Normal Distribution of the website user after conversion process.
If website user's one-tenth long value is obeyed data two ends, fluctuation larger data distributes, and this website user's one-tenth long value is done to conversion process reciprocal, makes the one-tenth long value Normal Distribution of the website user after conversion process.
If website user's one-tenth long value is obeyed ratio or number percent data distribute, this website user's one-tenth long value is first done to square root transformation and do again arcsine transformation, make the one-tenth long value Normal Distribution of the website user after conversion process.
In addition, carry out above-mentioned various conversion process and not only can change the distribution pattern of website user's one-tenth long value, and, to becoming long value to also have the effect of assembling, make into the approaching website user of long value and be more easily divided in same grade, thereby further ensure the fairness of classification.
Certainly,, if website user's one-tenth long value is obeyed other skewed distribution except above-mentioned, can do corresponding conversion process according to the transformational relation between the type distribution and normal distribution.
As can be seen from the above-described embodiment, utilize the feature of website user's one-tenth long value Normal Distribution, website user is divided in different brackets according to the probability of happening of specifying, according to the principle of theory of probability, the ratio that makes each grade website user account for all website users can maintain in the scope of probability of happening of this appointment, has ensured the stability of classification.Based on this classification results, in the time being website user's memory allocated resource of each grade, also can ensure that storage resources uses balanced.
Simultaneously, owing to being the grade of determining website user according to the size that becomes long value, also there will not be into two website users that long value differs greatly and be divided into the situation in same rank and become two identical website users of long value to be divided into the situation of different stage, also ensured the fairness of classification.
Embodiment tri-
Below taking website user's one-tenth long value obeys logarithm normal distribution as example, describe WEB application is carried out classification concrete grammar to website user in detail, refer to shown in Fig. 5, a kind of process flow diagram that website user is carried out to the method for classification that it discloses for the embodiment of the present invention three, comprise two parts flow process at this stage division, part flow process (specifically comprising the step 501-508 in Fig. 5) is for calculating the interval definite Website user rank of the numerical value falling at this website user's of standardized normal distribution one-tenth long value and according to this one-tenth long value, another part flow process (specifically comprising the step 509-511 in Fig. 5) is for determining that each grade website user's one-tenth long value is in the numerical range of standardized normal distribution.
Particularly, Part I flow process comprises the following steps:
Step 501: all website users' one-tenth long value in timing ergodic data storehouse, obtains a growth value sequence;
For example, this growth value sequence is X
i(i=1....n), X
i>=0, X
irepresent i website user's one-tenth long value.
Step 502: travel through this growth value sequence, each the one-tenth long value in growth value sequence is done to log-transformation processing, generate the growth value sequence after a conversion;
For example, become long value to do natural logarithm processing to each, that is, and E
i=ln (X
i) (i=1....n).E
irepresent the natural logarithm of i website user's one-tenth long value.
After the processing of step 502, website user's one-tenth long value Normal Distribution.
Step 503: travel through the growth value sequence after this conversion, calculate website user's average one-tenth long value;
For example,, according to formula
calculate website user's average one-tenth long value, the average one-tenth long value that μ is website user.
Step 504: travel through the growth value sequence after this conversion, calculate the standard deviation of website user's one-tenth long value;
For example,, according to formula
Calculate the standard deviation of website user's one-tenth long value, the standard deviation of the one-tenth long value that σ is website user.
Step 505: travel through the growth value sequence after this conversion, according to website user's average one-tenth long value and the standard deviation that becomes long value, calculate the coordinate conversion value of each website user's one-tenth long value, obtain into the coordinate conversion value sequence of long value;
For example,, according to formula
calculate the coordinate conversion value of each website user's one-tenth long value, Y
ibe the coordinate conversion value of i website user's one-tenth long value, that is, and i website user's one-tenth long value in standardized normal distribution.And obtain into the coordinate conversion value sequence Y of long value
i(i=1....n).
Step 506: the numerical range of the one-tenth long value that reads each grade website user in standardized normal distribution;
Step 507: travel through into the coordinate conversion value sequence of long value, the numerical range according to each grade website user's one-tenth long value in standardized normal distribution is determined the numerical range that each becomes the coordinate conversion value of long value to fall into;
Step 508: determine website user's grade according to the numerical range falling into, finish Part I flow process.
Particularly, Part II flow process comprises the following steps:
Step 509: set website user's number of levels and each grade website user's probability of happening;
Step 510: by query criteria normal distribution probability table, the numerical range of the one-tenth long value that obtains each grade website user in described normal distribution;
Step 511: the numerical range of the one-tenth long value of preserving each grade website user in described normal distribution, finishes Part II flow process.
As can be seen from the above-described embodiment, utilize the feature of website user's one-tenth long value Normal Distribution, website user is divided in different brackets according to the probability of happening of specifying, according to the principle of theory of probability, the ratio that makes each grade website user account for all website users can maintain in the scope of probability of happening of this appointment, has ensured the stability of classification.Based on this classification results, in the time being website user's memory allocated resource of each grade, also can ensure that storage resources uses balanced.
Simultaneously, owing to being the grade of determining website user according to the size that becomes long value, also there will not be into two website users that long value differs greatly and be divided into the situation in same rank and become two identical website users of long value to be divided into the situation of different stage, also ensured the fairness of classification.
Embodiment tetra-
With above-mentioned a kind of that website user is carried out to the method for classification is corresponding, the embodiment of the present invention also provides a kind of device that website user is carried out to classification.Refer to Fig. 6, a kind of structural drawing that website user is carried out to the device of classification that it discloses for the embodiment of the present invention four, this device comprises: determination module 601, read module 602 and diversity module 603.Principle of work below in conjunction with this device is further introduced its inner structure and annexation.
Determination module 601, for when website user's the one-tenth long value Normal Distribution, the numerical range according to default each grade website user's the each grade website user's of probability of happening ratio-dependent one-tenth long value in described normal distribution;
Read module 602, for reading website user's to be fractionated one-tenth long value;
Diversity module 603, for the numerical range falling into according to website user's to be fractionated one-tenth long value, determines website user's to be fractionated grade.
Preferably, on the basis of structure shown in Fig. 6, this device also comprises:
Resource distribution module, for according to classification results, is website user's distributing user information storage space of different brackets.
Preferably, on the basis of structure shown in Fig. 6, this device also comprises:
Judge module 604, distributes for judging whether website user's one-tenth long value obeys just too;
Type conversion module 605, when being no in the judged result of described judge module, is converted to normal distribution by the distribution pattern of described website user's one-tenth long value;
Described determination module 601, in the judged result of described judge module when being, the numerical range of the one-tenth long value of determining each grade website user according to default each grade website user's probability in described normal distribution.
Wherein, if described website user's one-tenth long value obeys logarithm normal distribution, type conversion module 605 is:
The first conversion module, for described website user's one-tenth long value is done to log-transformation processing, makes the one-tenth long value Normal Distribution of the website user after conversion process.
If described website user's one-tenth long value is obeyed Poisson distribution or slight skewness distributes, type conversion module 605 is:
The second conversion module, for described website user's one-tenth long value is done to square root transformation processing, makes the one-tenth long value Normal Distribution of the website user after conversion process.
If described website user's one-tenth long value is obeyed data two ends, fluctuation larger data distributes, and type conversion module 605 is:
The 3rd conversion module, for described website user's one-tenth long value is done to conversion process reciprocal, makes the one-tenth long value Normal Distribution of the website user after conversion process.
If described website user's one-tenth long value is obeyed ratio or number percent data distribute, type conversion module 605 is:
The 4th conversion module, does arcsine transformation again for described website user's one-tenth long value is first done to square root transformation, makes the one-tenth long value Normal Distribution of the website user after conversion process.
As can be seen from the above-described embodiment, utilize the feature of website user's one-tenth long value Normal Distribution, website user is divided in different brackets according to the probability of happening of specifying, according to the principle of theory of probability, the ratio that makes each grade website user account for all website users can maintain in the scope of probability of happening of this appointment, has ensured the stability of classification.Based on this classification results, in the time being website user's memory allocated resource of each grade, also can ensure that storage resources uses balanced.
Simultaneously, owing to being the grade of determining website user according to the size that becomes long value, also there will not be into two website users that long value differs greatly and be divided into the situation in same rank and become two identical website users of long value to be divided into the situation of different stage, also ensured the fairness of classification.
It should be noted that, one of ordinary skill in the art will appreciate that all or part of flow process realizing in above-described embodiment method, can carry out the hardware that instruction is relevant by computer program to complete, described program can be stored in a computer read/write memory medium, this program, in the time carrying out, can comprise as the flow process of the embodiment of above-mentioned each side method.Wherein, described storage medium can be magnetic disc, CD, read-only store-memory body (Read-Only Memory, ROM) or random store-memory body (Random Access Memory, RAM) etc.
Above a kind of method and apparatus that website user is carried out to classification provided by the present invention is described in detail, applied specific embodiment herein principle of the present invention and embodiment are set forth, the explanation of above embodiment is just for helping to understand method of the present invention and core concept thereof; , for one of ordinary skill in the art, according to thought of the present invention, all will change in specific embodiments and applications, in sum, this description should not be construed as limitation of the present invention meanwhile.
Claims (14)
1. a method of website user being carried out to classification, is characterized in that, comprising:
In the time of website user's one-tenth long value Normal Distribution, the numerical range according to default each grade website user's the each grade website user's of probability of happening ratio-dependent one-tenth long value in described normal distribution;
Read website user's to be fractionated one-tenth long value;
The numerical range falling into according to website user's to be fractionated one-tenth long value, determines website user's to be fractionated grade.
2. method according to claim 1, is characterized in that, also comprises:
According to classification results, it is website user's distributing user information storage space of different brackets.
3. method according to claim 1, is characterized in that, also comprises:
Judge website user's whether Normal Distribution of one-tenth long value;
If so, enter classification process;
Otherwise, the distribution pattern of described website user's one-tenth long value is converted to normal distribution, enter classification process.
4. method according to claim 3, is characterized in that, the described distribution pattern by website user's one-tenth long value is converted to normal distribution and comprises:
If described website user's one-tenth long value obeys logarithm normal distribution, does log-transformation processing by described website user's one-tenth long value, make the one-tenth long value Normal Distribution of the website user after conversion process.
5. method according to claim 3, is characterized in that, the described distribution pattern by website user's one-tenth long value is converted to distributions and comprises:
If described website user's one-tenth long value is obeyed Poisson distribution or slight skewness distributes, described website user's one-tenth long value is done to square root transformation processing, make the one-tenth long value Normal Distribution of the website user after conversion process.
6. method according to claim 3, is characterized in that, the described distribution pattern by website user's one-tenth long value is converted to distributions and comprises:
If described website user's one-tenth long value is obeyed data two ends, fluctuation larger data distributes, and described website user's one-tenth long value is done to conversion process reciprocal, makes the one-tenth long value Normal Distribution of the website user after conversion process.
7. method according to claim 3, is characterized in that, the described distribution pattern by website user's one-tenth long value is converted to distributions and comprises:
If described website user's one-tenth long value is obeyed ratio or number percent data distribute, described website user's one-tenth long value is first done to square root transformation and do again arcsine transformation, make the one-tenth long value Normal Distribution of the website user after conversion process.
8. a device that website user is carried out to classification, is characterized in that, comprising:
Determination module, for when website user's the one-tenth long value Normal Distribution, the numerical range according to default each grade website user's the each grade website user's of probability of happening ratio-dependent one-tenth long value in described normal distribution;
Read module, for reading website user's to be fractionated one-tenth long value;
Diversity module, for the numerical range falling into according to website user's to be fractionated one-tenth long value, determines website user's to be fractionated grade.
9. device according to claim 8, is characterized in that, also comprises:
Resource distribution module, for according to classification results, is website user's distributing user information storage space of different brackets.
10. device according to claim 8, is characterized in that, also comprises;
Judge module, distributes for judging whether website user's one-tenth long value obeys just too;
Type conversion module, when being no in the judged result of described judge module, is converted to normal distribution by the distribution pattern of described website user's one-tenth long value;
Described determination module, in the judged result of described judge module when being, the numerical range of the one-tenth long value of determining each grade website user according to default each grade website user's probability in described normal distribution.
11. devices according to claim 10, is characterized in that, if described website user's one-tenth long value obeys logarithm normal distribution, described type conversion module is:
The first conversion module, for described website user's one-tenth long value is done to log-transformation processing, makes the one-tenth long value Normal Distribution of the website user after conversion process.
12. devices according to claim 10, is characterized in that, if described website user's one-tenth long value is obeyed Poisson distribution or slight skewness distributes, described type conversion module is:
The second conversion module, for described website user's one-tenth long value is done to square root transformation processing, makes the one-tenth long value Normal Distribution of the website user after conversion process.
13. devices according to claim 10, is characterized in that, if described website user's one-tenth long value is obeyed data two ends, fluctuation larger data distributes, and described type conversion module is:
The 3rd conversion module, for described website user's one-tenth long value is done to conversion process reciprocal, makes the one-tenth long value Normal Distribution of the website user after conversion process.
14. devices according to claim 10, is characterized in that, if described website user's one-tenth long value is obeyed ratio or number percent data distribute, described type conversion module is:
The 4th conversion module, does arcsine transformation again for described website user's one-tenth long value is first done to square root transformation, makes the one-tenth long value Normal Distribution of the website user after conversion process.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310117776.7A CN104102649A (en) | 2013-04-07 | 2013-04-07 | Method and device for grading website users |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310117776.7A CN104102649A (en) | 2013-04-07 | 2013-04-07 | Method and device for grading website users |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104102649A true CN104102649A (en) | 2014-10-15 |
Family
ID=51670808
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310117776.7A Pending CN104102649A (en) | 2013-04-07 | 2013-04-07 | Method and device for grading website users |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104102649A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018006631A1 (en) * | 2016-07-08 | 2018-01-11 | 武汉斗鱼网络科技有限公司 | User level automatic segmentation method and system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1592903A (en) * | 2000-06-10 | 2005-03-09 | Ccr株式会社 | System and method for facilitating internet search by providing web document layout image |
| CN101261604A (en) * | 2008-04-09 | 2008-09-10 | 中兴通讯股份有限公司 | Software quality evaluation apparatus and software quality evaluation quantitative analysis method |
| CN101819587A (en) * | 2010-03-31 | 2010-09-01 | 北京志腾新诺科技有限公司 | Network information processing method and system |
| CN102024347A (en) * | 2001-01-09 | 2011-04-20 | 托普科德公司 | Systems and methods for coding competitions |
-
2013
- 2013-04-07 CN CN201310117776.7A patent/CN104102649A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1592903A (en) * | 2000-06-10 | 2005-03-09 | Ccr株式会社 | System and method for facilitating internet search by providing web document layout image |
| CN102024347A (en) * | 2001-01-09 | 2011-04-20 | 托普科德公司 | Systems and methods for coding competitions |
| CN101261604A (en) * | 2008-04-09 | 2008-09-10 | 中兴通讯股份有限公司 | Software quality evaluation apparatus and software quality evaluation quantitative analysis method |
| CN101819587A (en) * | 2010-03-31 | 2010-09-01 | 北京志腾新诺科技有限公司 | Network information processing method and system |
Non-Patent Citations (3)
| Title |
|---|
| 左任侠: "《教育与心理统计学》", 31 December 1982 * |
| 李晨宇: "中国社交网站盈利模式的可持续性分析", 《中国优秀硕士学位论文全文数据库》 * |
| 杨振明: "《概率论》", 31 December 1999 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018006631A1 (en) * | 2016-07-08 | 2018-01-11 | 武汉斗鱼网络科技有限公司 | User level automatic segmentation method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105900064B (en) | The method and apparatus for dispatching data flow task | |
| WO2018095065A1 (en) | Method and apparatus for allocating data object, and electronic device | |
| CN106959894B (en) | Resource allocation method and device | |
| CN107870981A (en) | Electronic installation, the method and storage medium of tables of data filing processing | |
| CN106899426B (en) | User access number statistical method and system thereof | |
| CN104081392A (en) | Influence scores for social media profiles | |
| CN103473334A (en) | Data storage method, inquiry method and system | |
| CN104239154B (en) | Job scheduling method and job scheduler in a kind of Hadoop clusters | |
| TW201712537A (en) | System, method, and apparatus for computer system resource allocation | |
| CN108805397A (en) | Electronic device, the method and storage medium for distributing task | |
| CN106952085B (en) | Method and device for data storage and service processing | |
| CN101419600A (en) | Data copy mapping method and device based on object-oriented LANGUAGE | |
| CN103853500A (en) | Method, device and system for distributing mass data | |
| CN103617083A (en) | Storage scheduling method and system, job scheduling method and system and management node | |
| CN103713953A (en) | Device and method for transferring data in memory | |
| CN104102649A (en) | Method and device for grading website users | |
| CN103560974B (en) | Method and device for maintaining tokens | |
| CN103002053B (en) | The profit maximization dispatching method of cloud computing and system | |
| TW201619817A (en) | Processing method for time series and system thereof | |
| CN115202890B (en) | Data element production resource space distribution method, system and equipment | |
| CN103902380A (en) | Method, device and equipment for determining resource distribution through sand box | |
| CN115618825A (en) | Financial statement merging method and device, computer readable medium and terminal equipment | |
| CN105511959A (en) | Method and device for distributing virtual resource | |
| CN114626730A (en) | Resource allocation method, device, computer equipment and computer readable storage medium | |
| CN109086309B (en) | Index dimension relation definition method, server and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20141015 |
|
| RJ01 | Rejection of invention patent application after publication |