A kind of intelligent recommendation management system of e-commerce platform
Technical field
The invention belongs to e-commerce platforms, are related to a kind of intelligent recommendation management system of e-commerce platform.
Background technique
In recent years, with the continuous rapid development of information technology and internet, ground of the e-commerce in society and life
Position is more and more significant, and e-commerce system provides more and more selections for user, in the arrival of big data era, e-commerce
The commodity of website are increased with index speed, no matter being all that people are unthinkable in its quantity or in type, this is more increased
The difficulty for oneself wanting commodity is rapidly and accurately obtained, so that user in search commercial articles, need to expend considerable time and effort, and
The commodity oneself wanted are difficult to find that, especially for clothes, as the improvement of people's living standards, people are to spirit and object
Demand in matter is gradually increased, can not be according to the body of user when people's clothes type required for e-commerce platform lookup
The parameters such as height, weight, measurements of the chest, waist and hips and the keyword of user's input are effectively screened, and cause user's waste when searching a large amount of
Time and efforts;
Existing e-commerce platform is simply screened by input keyword, can not combine height, the body of user
The parameters such as weight and measurements of the chest, waist and hips are effectively screened and are recommended, and poor, low efficiency that there are the accuracys of screening and intelligent characteristic are poor
The problem of, in order to solve problem above, now design a kind of intelligent recommendation management system of e-commerce platform.
Summary of the invention
The purpose of the present invention is to provide e-commerce platform intelligent recommendation management system, simultaneously by management server
In conjunction with commerce database, server and info push module are transferred, intelligent recommendation is carried out to user, to solve existing e-commerce
The existing screening accuracy of platform is poor, screening efficiency is low and the problem of intelligent characteristic difference.
The purpose of the present invention can be achieved through the following technical solutions:
A kind of intelligent recommendation management system of e-commerce platform, including user terminal, commerce database, transfer service
Device, management server and info push module;
It transfers server to connect with user terminal, commerce database, management server, management server and commerce database
It is connected with info push module, info push module is connect with user terminal;
User terminal retrieves information for inputting the essential characteristic parameter that need to retrieve information and user, and by the need of input
It is sent to the essential characteristic parameter of user and transfers server, and receive the clothes goodness of fit system that info push module successively pushes
Count each clothes under each clothes classification corresponding from high to low;
Commerce database is used to store the corresponding several keywords of each clothes under different garment classification, counts last quarter
Or the total quantity that each clothes classification is sold in year, and each clothes classification is ranked up according to from more to few sequence, respectively
It is 1,2 ..., i ..., n, each clothes are ranked up according to the sequence of sales volume from high to low in same clothes classification, it is followed successively by 1,
2 ..., g, and same clothes classification includes several keywords, each keyword in same clothes classification is according to shared by keyword
Ratio degree is descending to be ranked up, and respectively 1,2 ..., j ..., m, the corresponding key of each clothes in same clothes classification
Word constitutes standard key set of words Aig(aig1,aig2,...,aigj,...,aigm),aigJ is expressed as g in i-th of clothes classification
Corresponding j-th of the keyword of a clothes, the corresponding specific gravity of each keyword is respectively Sa in standard key set of wordsig1,
Saig2,...,Saigj,...,SaigM, SaiJ is expressed as corresponding j-th of keyword institute of g-th of clothes in i-th of clothes classification
The specific gravity accounted for;
Be stored in commerce database different height grades, weight grade, shoulder breadth grade, bust grade, waistline grade and
The corresponding model's model class of the sizing information of hip circumference grade, each model's model class correspond to different clothes classification sizes, respectively
Model's model class is ranked up according to the sequence of setting, respectively R1, R2 ..., Rv, the corresponding collection of each model's model grade
It is combined into Rv(rv1,rv2,...,rvW), rvW is expressed as the grade of w-th of characteristic parameter in v-th of model's model grade, different moulds
The grade of at least one characteristic parameter is different in special model class, successively to height, weight, shoulder in each model's model class
Wide, bust, waist and hip circumference feature levels are ranked up, and the height grade is arranged according to the sequence of height from high to low
Sequence is followed successively by s1, s2 ..., sf, and sf is expressed as the corresponding height ranges of f-th of height grade, and different height grades are corresponding not
With height ranges, the height ranges that s1 is indicated are 90-95cm, the difference in each height grade between highest height and minimum height
Value is 5cm, and the weight grade is ranked up according to the sequence of weight from high to low, is followed successively by t1, t2 ..., tf, tf table
It is shown as the corresponding weight weight range of f-th of weight grade, the weight range that t1 is indicated is 30-32kg, different weight grade pair
The different weight ranges answered, the shoulder breadth grade are ranked up according to the sequence of shoulder breadth size from high to low, are followed successively by k1,
K2 ..., kf, kf are expressed as the corresponding shoulder breadth size range of f-th of shoulder breadth grade, and the shoulder breadth range that k1 is indicated is 35-36cm,
The corresponding shoulder breadth size range of different shoulder breadth grades is different, the bust grade, waistline grade and hip circumference grade according to size by
It arrives small sequence greatly to be ranked up, respectively x1, x2 ..., xf, y1, y2 ..., yf, u1, u2 ..., uf, and xf, yf and uf
It is expressed as the corresponding size range of f-th of bust grade, the corresponding size range of f-th of waistline grade and f-th of stern
The corresponding size range of grade is enclosed, the bust range that x1 is indicated is 70-72cm, and the waistline range that y1 is indicated is 50-52cm, u1 table
The hip circumference range shown is 55-57cm;
Transfer server for receive user terminal transmission input retrieve information, by received retrieval information respectively with quotient
The corresponding several keywords of each clothes classification stored in business database are compared one by one, obtain the key in retrieval information
The keyword of acquisition is constituted search key set B (b1, b2 ..., bf) by word, and bf is expressed as f-th in retrieval information
Keyword, transfer server by each keyword in search key set respectively with corresponding keyword under different garment classification
Each keyword in set compares, and obtains comparison keyword set A 'ig(a′ig1,a′ig2,...,a′igj,...,a′igM), a 'igJ is expressed as the jth corresponding with g-th of clothes in i-th of clothes classification of each keyword in search key set
The reduced value of a keyword, if each keyword in search key set not with g-th of clothes pair in i-th of clothes classification
J-th of the keyword answered is identical, then takes a 'igJ is equal to 0, on the contrary then take a 'igJ is equal to T, and T is fixed numerical value, transfers server
Obtained comparison keyword set is sent to management server;
Meanwhile transferring server and receiving user's essential characteristic parameter that user terminal is sent, extract user's essential characteristic ginseng
Height, weight, shoulder breadth, bust, waist and hip circumference size in number, and by received height, weight, shoulder breadth, bust, waistline and
Hip circumference size range corresponding from different height grades corresponding in commerce database, the corresponding range of weight grade, shoulder respectively
The corresponding range of wide grade, the corresponding range of bust grade, the corresponding range of waistline grade and the corresponding range of hip circumference grade
Compared one by one, obtain each feature level of the user, transfer server by user's height grade of acquisition, weight grade,
Shoulder breadth grade, bust grade, waistline grade and hip circumference grade are sent to management server;
Management server receives the height grade for transferring the user of server transmission, weight grade, shoulder breadth grade, bust etc.
Grade, waistline grade and hip circumference grade simultaneously constitute user screening class set D (d1, d2 ..., dw), and dw is expressed as ith feature
The grade of parameter, according to the grade of each characteristic parameter in user's screening class set one by one with stored in commerce database it is each
Each characteristic parameter grade in model's model grade compares, and obtains comparison model's model class set R 'v(r′v1,r′v2,...,r′vW), r 'vW is expressed as w-th of characteristic parameter grade and user's screening class collection under v-th of model's model grade
The reduced value of w-th of characteristic parameter grade in conjunction, management server is according to comparison model's model class set R 'vCounting user
Feature level parameter and each model's model class between matching degree FACTOR Pv, management server filter out user characteristics grade ginseng
The maximum model's model class of matching degree coefficient between several and each model's model class;
Management server is according to the matching degree coefficient between the user characteristics class parameter and each model's model class filtered out
Maximum model's model class extracts the clothes classification size between the maximum model's model class of matching degree coefficient, and receives tune
The comparison set of keywords for taking server to send, statistics input are retrieved under information and user characteristics parameter and each clothes classification
Clothes goodness of fit coefficientQvigIt is expressed as user and v-th of clothes classification
In g-th of clothes joint match degree coefficient, c is expressed as influencing changed factor, and the influences changed factor is equal to the clothes
Clothes quantity under classification accounts for the ratio of the clothes total quantity of purchase, a 'igJ is expressed as each keyword in search key set
With the reduced value of j-th of keyword in i-th of clothes classification in g-th of clothes, SaigJ is expressed as in i-th of clothes classification
Specific gravity shared by j-th of keyword in g clothes,It is expressed as between user characteristics class parameter and each model's model class
The maximum model's model class of matching degree coefficient between matching degree coincidence coefficient, management server is by the clothes under each clothes classification
Dress goodness of fit coefficient and the standard garments goodness of fit coefficient of setting compare, and reject the clothes goodness of fit system under each clothes classification
Number is less than the clothes of the standard garments goodness of fit coefficient of setting, and will be greater than each clothes of the standard garments goodness of fit coefficient of setting
Each clothes under classification are successively sent to info push module according to the sequence of clothes goodness of fit coefficient from high to low;
Info push module receives the goodness of fit coefficient of each clothes under each clothes classification that management server is sent, and root
According to received sequencing, each clothes received under the corresponding each clothes classification of the goodness of fit coefficient are successively pushed into user's end
End.
Further, the essential characteristic parameter of the user includes height, weight, shoulder breadth, bust, waistline and the stern of user
It encloses.
Further, specific gravity shared by corresponding j-th of the keyword of g-th of clothes in i-th of clothes classificationIn same clothes classification, Saig1 > Saig2 > ... > Saigj
> ... > SaigM, and Saig1+Saig2+...+Saigj+...+SaigM=1.
Further, the difference in every individual weight grade between highest weight numerical value and minimum weight numerical value is 2kg, each shoulder
Difference in wide grade between maximum shoulder breadth size and minimum shoulder size is identical, is 1cm, highest bust size in each bust grade
Difference in difference, each waistline grade between minimum bust size between highest waist size and minimum waist size and every
Difference in a hip circumference grade between highest hip circumference size and minimum hip circumference size is 2cm.
Further, each feature level of the user includes the height grade, weight grade, shoulder breadth grade, chest of user
Enclose grade, waistline grade and hip circumference grade.
Further, if in w-th characteristic parameter grade and user's screening class set under v-th of model's model grade
W-th of characteristic parameter grade it is identical, then take r 'vW is equal to 0, conversely, w-th of feature under v-th of model's model grade is joined
Number grade is carried out making difference and be taken absolute value with w-th of characteristic parameter grade in user's screening class set, obtains grade difference
Δ w, and Δ w > 0.
Further, the calculation formula of the matching degree coefficient between the feature level parameter of user and each model's model class isZ is expressed as matching impact factor, takes 3.680, r 'vW is expressed as under v-th of model's model grade
The reduced value of w-th of characteristic parameter grade in w characteristic parameter grade and user's screening class set, ΛwIt is expressed as v-th
Numerical value in w-th of characteristic parameter grade under model's model grade between highest size and minimum size.
Beneficial effects of the present invention:
The intelligently pushing management system of e-commerce platform provided by the invention need to retrieve information by user terminal input
And user essential characteristic parameter and combine transfer server, management server, commerce database and info push module, unite
Meter input need to retrieve the corresponding comparison keyword set of information, and the essential characteristic parameter of counting user and each model's model class
Between matching degree coefficient, and screen the maximum model's model class of matching degree coefficient, it is maximum according to the matching degree coefficient of screening
Model's model class and comparison set of keywords, statistics input retrieval information and user's essential characteristic parameter and clothes classification
Under each clothes goodness of fit coefficient, and the corresponding clothes of goodness of fit coefficient are pushed into user terminal through info push module,
It realizes the intelligent clothing push of e-commerce platform, convenient for recommending suitable push content for user, improves user's screening
Accuracy and satisfaction reduce the time and efforts of lookup, have the characteristics that intelligence.
Detailed description of the invention
In order to illustrate the technical solution of the embodiments of the present invention more clearly, will be described below to embodiment required
Attached drawing is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for ability
For the those of ordinary skill of domain, without creative efforts, it can also be obtained according to these attached drawings other attached
Figure.
Fig. 1 is a kind of schematic diagram of the intelligent recommendation management system of e-commerce platform in the present invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts all other
Embodiment shall fall within the protection scope of the present invention.
Refering to Figure 1, a kind of intelligent recommendation management system of e-commerce platform, including user terminal, commercial number
According to library, transfer server, management server and info push module;
It transfers server to connect with user terminal, commerce database, management server, management server and commerce database
It is connected with info push module, info push module is connect with user terminal;
User terminal retrieves information for inputting the essential characteristic parameter that need to retrieve information and user, and by the need of input
It is sent to the essential characteristic parameter of user and transfers server, and receive the clothes goodness of fit system that info push module successively pushes
Number each clothes under corresponding each clothes classification from high to low, the essential characteristic parameter of the user include the height of user,
The information such as weight, shoulder breadth, bust, waist and hip circumference;
Commerce database is used to store the corresponding several keywords of each clothes under different garment classification, counts last quarter
Or the total quantity that each clothes classification is sold in year, and each clothes classification is ranked up according to from more to few sequence, respectively
It is 1,2 ..., i ..., n, the clothes classification includes jeans, casual pants, defends clothing, one-piece dress, half body skirt etc., same clothes
Each clothes are ranked up according to the sequence of sales volume from high to low in classification, are followed successively by 1,2 ..., g, and same clothes classification includes
Several keywords, each keyword in same clothes classification are ranked up according to ratio degree is descending shared by keyword,
Respectively 1,2 ..., j ..., m, the corresponding keyword of each clothes constitutes standard key set of words A in same clothes classificationig
(aig1,aig2,...,aigj,...,aigm),aigJ is expressed as corresponding j-th of the key of g-th of clothes in i-th of clothes classification
Word, the corresponding specific gravity of each keyword is respectively Sa in standard key set of wordsig1,Saig2,...,Saigj,...,SaigM,In same clothes classification, Saig1 > Saig2 > ... > Saigj
> ... > SaigM, and Saig1+Saig2+...+Saigj+...+SaigM=1;
Be stored in commerce database different height grades, weight grade, shoulder breadth grade, bust grade, waistline grade and
The corresponding model's model class of the sizing information of hip circumference grade, each model's model class correspond to different clothes classification sizes, respectively
Model's model class is ranked up according to the sequence of setting, respectively R1, R2 ..., Rv, the corresponding collection of each model's model grade
It is combined into Rv(rv1,rv2,...,rvW), rvW is expressed as the grade of w-th of characteristic parameter in v-th of model's model grade, different moulds
The grade of at least one characteristic parameter is different in special model class, successively to height, weight, shoulder in each model's model class
Wide, bust, waist and hip circumference feature levels are ranked up, and the height grade is arranged according to the sequence of height from high to low
Sequence is followed successively by s1, s2 ..., sf, and sf is expressed as the corresponding height ranges of f-th of height grade, and different height grades are corresponding not
With height ranges, the height ranges that s1 is indicated are 90-95cm, the difference in each height grade between highest height and minimum height
Value is 5cm, and the weight grade is ranked up according to the sequence of weight from high to low, is followed successively by t1, t2 ..., tf, tf table
It is shown as the corresponding weight weight range of f-th of weight grade, the weight range that t1 is indicated is 30-32kg, different weight grade pair
The different weight ranges answered, the difference in every individual weight grade between highest weight numerical value and minimum weight numerical value is 2kg, described
Shoulder breadth grade is ranked up according to the sequence of shoulder breadth size from high to low, is followed successively by k1, k2 ..., kf, and kf is expressed as f-th of shoulder
The corresponding shoulder breadth size range of wide grade, the shoulder breadth range that k1 is indicated are 35-36cm, the corresponding shoulder breadth size of different shoulder breadth grades
Range is different, and the difference in each shoulder breadth grade between maximum shoulder breadth size and minimum shoulder size is identical, is 1cm;Described bust etc.
Grade, waistline grade and hip circumference grade are ranked up according to the descending sequence of size, respectively x1, x2 ..., xf, y1,
Y2 ..., yf, u1, u2 ..., uf, and xf, yf and uf are expressed as the corresponding size range of f-th of bust grade, f-th
The corresponding size range of waistline grade and the corresponding size range of f-th of hip circumference grade, the bust range that x1 is indicated are 70-
The waistline range that 72cm, y1 are indicated is 50-52cm, and the hip circumference range that u1 is indicated is 55-57cm, highest chest in each bust grade
Enclose difference between size and minimum bust size, the difference in each waistline grade between highest waist size and minimum waist size
And the difference in each hip circumference grade between highest hip circumference size and minimum hip circumference size is 2cm.
Transfer server for receive user terminal transmission input retrieve information, by received retrieval information respectively with quotient
The corresponding several keywords of each clothes classification stored in business database are compared one by one, obtain the key in retrieval information
The keyword of acquisition is constituted search key set B (b1, b2 ..., bf) by word, and bf is expressed as f-th in retrieval information
Keyword, transfer server by each keyword in search key set respectively with corresponding keyword under different garment classification
Each keyword in set compares, and obtains comparison keyword set A 'ig(a′ig1,a′ig2,...,a′igj,...,a′igM), a 'igJ is expressed as the jth corresponding with g-th of clothes in i-th of clothes classification of each keyword in search key set
The reduced value of a keyword, if each keyword in search key set not with g-th of clothes pair in i-th of clothes classification
J-th of the keyword answered is identical, then takes a 'igJ is equal to 0, on the contrary then take a 'igJ is equal to T, and T is fixed numerical value, transfers server
Obtained comparison keyword set is sent to management server;
Meanwhile transferring server and receiving user's essential characteristic parameter that user terminal is sent, extract user's essential characteristic ginseng
Height, weight, shoulder breadth, bust, waist and hip circumference size in number, and by received height, weight, shoulder breadth, bust, waistline and
Hip circumference size range corresponding from different height grades corresponding in commerce database, the corresponding range of weight grade, shoulder respectively
The corresponding range of wide grade, the corresponding range of bust grade, the corresponding range of waistline grade and the corresponding range of hip circumference grade
It is compared one by one, obtains each feature level of the user, each feature level of the user includes the height grade of user, body
Weight grade, shoulder breadth grade, bust grade, waistline grade and hip circumference grade, transfer server for user's height grade of acquisition, body
Weight grade, shoulder breadth grade, bust grade, waistline grade and hip circumference grade are sent to management server, wherein if user's is a certain
Characteristic size numerical value then shows that a certain characteristic size of the user is this feature etc. in the corresponding a certain rate range of this feature
Grade.
Management server receives the height grade for transferring the user of server transmission, weight grade, shoulder breadth grade, bust etc.
Grade, waistline grade and hip circumference grade simultaneously constitute user screening class set D (d1, d2 ..., dw), and dw is expressed as ith feature
The grade of parameter, according to the grade of each characteristic parameter in user's screening class set one by one with stored in commerce database it is each
Each characteristic parameter grade in model's model grade compares, and obtains comparison model's model class set R 'v(r′v1,r′v2,...,r′vW), r 'vW is expressed as w-th of characteristic parameter grade and user's screening class collection under v-th of model's model grade
The reduced value of w-th of characteristic parameter grade in conjunction, if w-th of characteristic parameter grade and use under v-th of model's model grade
W-th of characteristic parameter grade in the screening class set of family is identical, then takes r 'vW is equal to 0, conversely, by v-th of model's model etc.
W-th of characteristic parameter grade under grade and w-th of characteristic parameter grade in user's screening class set are carried out making difference and be taken absolutely
To value, grade difference DELTA w, and Δ w > 0 are obtained, management server is according to comparison model's model class set R 'vCounting user
Matching degree coefficient between feature level parameter and each model's model classZ be expressed as matching influence because
Son takes 3.680, r 'vW is expressed as w-th of characteristic parameter grade and user's screening class set under v-th of model's model grade
In w-th of characteristic parameter grade reduced value, ΛwW-th of characteristic parameter grade being expressed as under v-th of model's model grade
Numerical value between middle highest size and minimum size filters out the matching degree between user characteristics class parameter and each model's model class
The maximum model's model class of coefficient;
Management server is according to the matching degree coefficient between the user characteristics class parameter and each model's model class filtered out
Maximum model's model class extracts the clothes classification size between the maximum model's model class of matching degree coefficient, and receives tune
The comparison set of keywords for taking server to send, statistics input are retrieved under information and user characteristics parameter and each clothes classification
Clothes goodness of fit coefficientQvigIt is expressed as user and v-th of clothes classification
In g-th of clothes joint match degree coefficient, c is expressed as influencing changed factor, and the influences changed factor is equal to the clothes
Clothes quantity under classification accounts for the ratio of the clothes total quantity of purchase, a 'igJ is expressed as each keyword in search key set
With the reduced value of j-th of keyword in i-th of clothes classification in g-th of clothes, SaigJ is expressed as in i-th of clothes classification
Specific gravity shared by j-th of keyword in g clothes,It is expressed as between user characteristics class parameter and each model's model class
The maximum model's model class of matching degree coefficient between matching degree coincidence coefficient, management server is by the clothes under each clothes classification
Dress goodness of fit coefficient and the standard garments goodness of fit coefficient of setting compare, and reject the clothes goodness of fit system under each clothes classification
Number is less than the clothes of the standard garments goodness of fit coefficient of setting, and will be greater than each clothes of the standard garments goodness of fit coefficient of setting
Each clothes under classification are successively sent to info push module according to the sequence of clothes goodness of fit coefficient from high to low;
Info push module receives the goodness of fit coefficient of each clothes under each clothes classification that management server is sent, and root
According to received sequencing, each clothes received under the corresponding each clothes classification of the goodness of fit coefficient are successively pushed into user's end
End, shows the clothes for meeting user characteristics parameter, intuitively convenient for user terminal so that user selects.
The intelligently pushing management system of e-commerce platform provided by the invention need to retrieve information by user terminal input
And user essential characteristic parameter and combine transfer server, management server, commerce database and info push module, unite
Meter input need to retrieve the corresponding comparison keyword set of information, and the essential characteristic parameter of counting user and each model's model class
Between matching degree coefficient, and screen the maximum model's model class of matching degree coefficient, it is maximum according to the matching degree coefficient of screening
Model's model class and comparison set of keywords, statistics input retrieval information and user's essential characteristic parameter and clothes classification
Under each clothes goodness of fit coefficient, and the corresponding clothes of goodness of fit coefficient are pushed into user terminal through info push module,
It realizes the intelligent clothing push of e-commerce platform, convenient for recommending suitable push content for user, improves user's screening
Accuracy and satisfaction reduce the time and efforts of lookup, have the characteristics that intelligence.
The above content is just an example and description of the concept of the present invention, affiliated those skilled in the art
It makes various modifications or additions to the described embodiments or is substituted in a similar manner, without departing from invention
Design or beyond the scope defined by this claim, be within the scope of protection of the invention.