CN107563833A - A kind of personalized recommendation method and system based on block chain integration service platform - Google Patents
A kind of personalized recommendation method and system based on block chain integration service platform Download PDFInfo
- Publication number
- CN107563833A CN107563833A CN201710621801.3A CN201710621801A CN107563833A CN 107563833 A CN107563833 A CN 107563833A CN 201710621801 A CN201710621801 A CN 201710621801A CN 107563833 A CN107563833 A CN 107563833A
- Authority
- CN
- China
- Prior art keywords
- user
- commodity
- vector information
- product features
- targeted customer
- 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
- 230000010354 integration Effects 0.000 title claims abstract description 37
- 238000000034 method Methods 0.000 title claims abstract description 29
- 239000011159 matrix material Substances 0.000 claims abstract description 56
- 238000012545 processing Methods 0.000 claims abstract description 34
- 238000011524 similarity measure Methods 0.000 claims abstract description 15
- 238000012937 correction Methods 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000013519 translation Methods 0.000 claims description 7
- 238000004458 analytical method Methods 0.000 claims description 6
- 238000010276 construction Methods 0.000 claims description 6
- 238000012163 sequencing technique Methods 0.000 claims description 4
- 238000005516 engineering process Methods 0.000 abstract description 11
- 230000006399 behavior Effects 0.000 description 11
- 238000010586 diagram Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 2
- 230000004927 fusion Effects 0.000 description 2
- 238000011478 gradient descent method Methods 0.000 description 2
- 238000003058 natural language processing Methods 0.000 description 2
- 230000003542 behavioural effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000001394 metastastic effect Effects 0.000 description 1
- 206010061289 metastatic neoplasm Diseases 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 235000015170 shellfish Nutrition 0.000 description 1
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of personalized recommendation method and system based on block chain integration service platform, wherein, the personalized recommendation method includes:The commodity data of user's purchase is pre-processed, obtains pre-processed results;Commodity user is built according to pre-processed results to fall to sort and user's commodity association table;Fall to sort according to commodity user and carry out user's Similarity Measure processing, obtain user's similarity matrix;Characteristic vector pickup processing is carried out according to user's commodity association table, obtains user's product features vector information;User's product features vector information of targeted customer is modified according to user's similarity matrix and user's product features vector information, the user's product features vector information for obtaining revised targeted customer carries out commodity push according to user's product features vector information of revised targeted customer to targeted customer.In embodiments of the present invention, the data edge that block chain technology brings is effectively utilized by the embodiment of the present invention and realizes that more accurate individual character is recommended.
Description
Technical field
The present invention relates to big data technical field, more particularly to a kind of personalization based on block chain integration service platform to push away
Recommend method and system.
Background technology
As new technology, block chain technology is not yet ripe in the stability of a system, application security, business model etc.,
Be primarily adapted for use at present non real-time nature, lightweight information, transaction handling capacity is smaller and information sensitivity is relatively low business field
Scape.The agreement of block chain is hashing algorithm safe to use (Secure Hash Algorithm, abbreviation SHA) and superencipherment
Standard (Advanced Encryption Standard, abbreviation AES) algorithm etc. develop, therefore it can be used as storage with
The security platform of transmission integration ownership.In addition, other attributes such as immutableness of block chain, be easy to metastatic, along with examining
Transparency and ease for use are counted, is allowed to more valuable.
Integration managing is one of important application scene of block chain technology.Block chain integrating system is to use block chain bottom
The integration distribution and management system that technology is made, safeguard an authentic data storehouse by block chain, are ensureing platform integration data just
Often operation, have safety, it is open, can expand and the characteristic such as easy to maintain.Block chain can be integration publisher, application developers
Disintermediation is provided between consumer, decouples the task about integration managing, such as distribution, trading processing, ensures user's fund security.
At present, the design and research of domestic several block chain company incision block chain integrating systems, wherein what is put it into commercial operation has
Sound of laughing integration, number shellfish pocket and a too cloud.
The existing technology on accumulated point exchanging push is according to the demand of the behavioural analysis user of user, is chosen
At present much-sought-after item is recommended to user;Contacting between user and project is excavated to project scoring based on user, then carried out
Information pushes, and such scheme is used primarily in library book recommendation, film scoring recommendation etc.;Integration platform carries out personalization and pushed away
Recommend, the transaction of commerce system internal integral, accumulated point exchanging.
Above-mentioned technically there are several weak points, such as:It is keen to the recommendation of much-sought-after item, coverage rate is low, unfavorable
In the Recycle mechanism of whole platform;Commerce system inside panel can recommended range it is small, can not effectively excavate long-tail information;In foundation
User scores and carried out in the system of project recommendation, and system depends on user's scoring unduly, and many users do not have score data, this
Sample can cause to recommend not accurate enough because of Sparse.
The content of the invention
It is an object of the invention to overcome the deficiencies in the prior art, the invention provides one kind based on the integration service of block chain
The personalized recommendation method and system of platform, the data edge that block chain technology brings is effectively utilized by the embodiment of the present invention
Realize that more accurate individual character is recommended.
In order to solve the above-mentioned technical problem, the embodiments of the invention provide a kind of based on block chain integration service platform
Property personalized recommendation method described in recommendation method includes:
The commodity data of user's purchase is pre-processed, obtains pre-processed results;
Commodity user is built according to the pre-processed results to fall to sort and user's commodity association table;
Fall to sort according to the commodity user and carry out user's Similarity Measure processing, obtain user's similarity matrix;
Characteristic vector pickup processing is carried out according to user's commodity association table, obtains user's product features vector information;
User's commodity according to user's similarity matrix and user's product features vector information to targeted customer
Eigenvector information is modified, and obtains user's product features vector information of revised targeted customer;
Commodity push is carried out to targeted customer according to user's product features vector information of the revised targeted customer.
Preferably, described sorted according to the commodity user carries out user's Similarity Measure processing, including:
Improving cosine similarity formula using much-sought-after item penalty term, sequence progress user is similar to the commodity user
Spend calculating processing, obtain user's similarity matrix and user it is neighbouring user collection.
Preferably, it is described that characteristic vector pickup processing is carried out according to user's commodity association table, including:
Carry out building user's commodity matrix disposal using user's commodity association table, obtain user's commodity matrix;
Line translation is entered to user's commodity matrix, integrated value penalty term and hot topic are added during commodity matrixing
Commodity penalty term, obtain integrated value corresponding to user's commodity;
The much-sought-after item diversity for not having behavior according to the user of the article set formed objects interested with user is built
Sample set;
Implicit doctrine analysis is carried out according to integrated value and the sample set corresponding to user's commodity, obtains user's commodity
Eigenvector information;
User's product features vector information includes user characteristics vector information and product features vector information.
Preferably, it is described according to user's similarity matrix and user's product features vector information to targeted customer
User's product features vector information be modified, including:
Using K similarity highest user set, family eigenvector information and commodity in user's similarity matrix
Eigenvector information is modified to user's product features vector information of targeted customer, obtains the use of revised targeted customer
Family product features vector information.
Preferably, user's product features vector information according to the revised targeted customer is entered to targeted customer
Product of doing business push, including:
Interested value of the targeted customer to all commodity is carried out using correction value and user's product features vector information
Calculate, obtain interested value of the targeted customer to all commodity;
Descending sort processing is carried out to the value interested of all commodity to the targeted customer, acquisition order is gathered;
The article for not having behavior to targeted customer according to the order set carries out commodity push to targeted customer.
In addition, the embodiment of the present invention additionally provides a kind of personalized recommendation system based on block chain integration service platform,
The personalized recommendation system includes:
Pretreatment module:For being pre-processed to the commodity data that user buys, pre-processed results are obtained;
Build module:Fall to sort and user's commodity association table for building commodity user according to the pre-processed results;
Similarity calculation module:User's Similarity Measure processing is carried out for falling to sort according to the commodity user, is obtained
User's similarity matrix;
Characteristic vector pickup module:For carrying out characteristic vector pickup processing according to user's commodity association table, obtain
User's product features vector information;
Correcting module:For being used according to user's similarity matrix and user's product features vector information target
User's product features vector information at family is modified, and obtains user's product features vector information of revised targeted customer;
Recommending module:For user's product features vector information according to the revised targeted customer to targeted customer
Carry out commodity push.
Preferably, the similarity calculation module includes:
Similarity calculated:For improving cosine similarity formula to the commodity user using much-sought-after item penalty term
Sequence carry out user's Similarity Measure processing, obtain user's similarity matrix and user it is neighbouring user collection.
Preferably, the characteristic vector pickup module includes:
Matrix construction unit:For carrying out building user's commodity matrix disposal using user's commodity association table, obtain
User's commodity matrix;
Matrixing unit:For entering line translation to user's commodity matrix, added during commodity matrixing
Integrated value penalty term and much-sought-after item penalty term, obtain integrated value corresponding to user's commodity;
Sample set construction unit:There is not behavior for the user according to the article set formed objects interested with user
Much-sought-after item diversity structure sample set;
Implicit doctrine analytic unit:Implied for integrated value and the sample set according to corresponding to user's commodity
Doctrine is analyzed, and obtains user's product features vector information;
User's product features vector information includes user characteristics vector information and product features vector information.
Preferably, the correcting module includes:
Amending unit:For being gathered using K similarity highest user in user's similarity matrix, family feature to
Amount information and product features vector information are modified to user's product features vector information of targeted customer, are obtained revised
User's product features vector information of targeted customer.
Preferably, the recommending module includes
Computing unit:For carrying out targeted customer to all business using correction value and user's product features vector information
The value interested of product calculates, and obtains interested value of the targeted customer to all commodity;
Sequencing unit:For carrying out descending sort processing to the value interested of all commodity to the targeted customer, obtain
Order is gathered;
Push unit:For not there is the article of behavior to be carried out to targeted customer to targeted customer according to the order set
Commodity push.
In embodiments of the present invention, the data edge that block chain technology brings is effectively utilized by the embodiment of the present invention,
The authenticity of user integral transaction data and can not tamper;Platform integration " multi-party distribution, free flow " simultaneously, ensure that number
According to the popularity in source and comprehensive;Neighbouring model of the method based on user is similar using the cosine for adding much-sought-after item penalty term
Degree improves similarity definition, and the hidden semantic model for using for reference natural language processing excavates user and commodity hiding information, and fusion is based on
The advantages of neighbouring model of user and hidden semantic model, obtained using the hidden semantic model of neighbour's user characteristics amendment of targeted customer
User characteristics, realize that more accurate individual character is recommended.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
There is the required accompanying drawing used in technology description to be briefly described, it is clear that, drawings in the following description are only this
Some embodiments of invention, for those of ordinary skill in the art, on the premise of not paying creative work, can be with
Other accompanying drawings are obtained according to these accompanying drawings.
Fig. 1 is the method flow of the personalized recommendation method based on block chain integration service platform in the embodiment of the present invention
Schematic diagram;
Fig. 2 is the system architecture of the personalized recommendation system based on block chain integration service platform in the embodiment of the present invention
Composition schematic diagram.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation describes, it is clear that described embodiment is only part of the embodiment of the present invention, rather than whole embodiments.It is based on
Embodiment in the present invention, those of ordinary skill in the art are obtained all other under the premise of creative work is not made
Embodiment, belong to the scope of protection of the invention.
Block chain integration service platform possesses safe identification safety authentication, and member user possesses reliable unique ID, user
The integration that can be realized between user is represented and cross-merchant folding accumulated point exchanging.Party A-subscriber is integrating caused by a businessmans, can turn
Give party B-subscriber and the accumulated point exchanging service for a businessmans or other businessmans.In the system of block chain, all data are all
It is bright and can not distort, it can be well understood between trade company and user, risk of breaking one's promise reduces.Meanwhile block chain have go to center
The characteristics of change, integration transaction data can be shared between multiple businessmans, realizes depth analysis and the utilization of data.
User in accumulated point exchanging in face of the commodity of magnanimity when, personalized recommendation method can be very good solve information overload
Problem, and long-tail information is fully excavated, improve degree of exchanging of the user with platform.The method of common recommendation much-sought-after item, covering
Rate is low, it is impossible to looks after the most commodity of platform, easily forms " Matthew effect ", much-sought-after item is more popular, unexpected winner commodity
It is more difficult to be realized by user.Often this can cause user to lose interest platform in the past, and the sale demand of businessman is also not being met.This
Intention can solve above mentioned problem, realize a benign exchange between user, platform, businessman, improve user and businessman couple
The viscosity of platform.
Fig. 1 is the method flow of the personalized recommendation method based on block chain integration service platform in the embodiment of the present invention
Schematic diagram, as shown in figure 1, the personalized recommendation method includes:
S11:The commodity data of user's purchase is pre-processed, obtains pre-processed results;
S12:Commodity user is built according to the pre-processed results to fall to sort and user's commodity association table;
S13:Fall to sort according to the commodity user and carry out user's Similarity Measure processing, obtain user's similarity matrix;
S14:Characteristic vector pickup processing is carried out according to user's commodity association table, obtains user's product features vector letter
Breath;
S15:User according to user's similarity matrix and user's product features vector information to targeted customer
Product features vector information is modified, and obtains user's product features vector information of revised targeted customer;
S16:Commodity are carried out to targeted customer according to user's product features vector information of the revised targeted customer
Push.
S11 is described further:
The commodity data of user's purchase is pre-processed, obtains pre-processed results.
The ID in database in each transaction data, commodity ID, the integrated value of transaction, block chain skill are obtained first
Art ensure that the accuracy of each transaction, reduce the work of later data pretreatment;Redundancy is carried out to the data in above-mentioned
The processing such as removal and wrong data removal, and obtain result.
S12 is described further:
Pretreated data in above-mentioned steps are examined, determine that data do not lack and in Limit of J-validity
It is interior, sequencing table is fallen according to commodity-user is established in record to the user of extraction and commodity association table, in order to establish proximal subscribers collection
Model is prepared, and establishes user and matrix of the commodity on integration.
S13 is described further:
Fall to sort according to the commodity user and carry out user's Similarity Measure processing, obtain user's similarity matrix.
Further, cosine similarity formula is improved to the commodity user sequence progress using much-sought-after item penalty term
User's Similarity Measure processing, obtain user's similarity matrix and user it is neighbouring user collection.
Specifically, utilize commodity --- user's inverted list, the similarity between user is calculated, build user's similarity moment
Battle array, as the proximal subscribers collection of targeted customer, the set with K user's composition of targeted customer's similarity highest.
User's Similarity Measure is carried out to user model, much-sought-after item penalty term is added and improves cosine similarity formula, give
Determining user u and user v, make N (u) represent user u article set interested, N (v) represents user v article set interested,
N (i) represents to have commodity i the user of behavior to gather, and user u and user i calculating formula of similarity are as follows:
T (u) represents k similarity highest user set of u user.
S14 is described further:
Characteristic vector pickup processing is carried out according to user's commodity association table, obtains user's product features vector information.
Further, carry out building user's commodity matrix disposal using user's commodity association table, obtain user's commodity
Matrix;Line translation is entered to user's commodity matrix, integrated value penalty term and popular business are added during commodity matrixing
Product penalty term, obtain integrated value corresponding to user's commodity;According to the user of the article set formed objects interested with user not
There is the much-sought-after item diversity structure sample set of behavior;Carried out according to integrated value corresponding to user's commodity and the sample set
Implicit doctrine analysis, obtains user's product features vector information;User's product features vector information include user characteristics to
Measure information and product features vector information.
Specifically, entering line translation to user's commodity matrix, integrated value penalty term and much-sought-after item penalty term are added, gives and uses
Family u, commodity i and corresponding integrated value, N (i) represents to have commodity i the user of behavior to gather, after conversion:
Sampling and the user u of N (u) (article set interested user u) formed objects did not had the much-sought-after item of behavior
It is commodity i unit integrated values as data set as sample set M (u), interest value r of the u user to commodity i in sample setuiDefinition
For:
rui=-log (1+ri)*log(1+|N(i)|),i∈M(u);
It is emerging to commodity i's with commercial productainterests matrix R, user u that user is established using positive sample data set and sample data set
Interesting value is to obtain user --- commodity collection K={ (u, i) } hidden semantic model calculates user u to the emerging of commodity i by equation below
Interest:
Most suitable parameter p and q is found by minimizing loss function C:
Wherein λ | | pu||2+λ||qi||2For regular terms, C is optimized using stochastic gradient descent method, finds user characteristics square
Battle array p and product features matrix q, local derviation is sought to C:
Using stochastic gradient descent method, iterative formula is obtained, wherein being learning rate:
puk=puk+α(qikeui-λpuk);
qik=qik+α(pukeui-λqik)。
S15 is described further:
User's commodity according to user's similarity matrix and user's product features vector information to targeted customer
Eigenvector information is modified, and obtains user's product features vector information of revised targeted customer.
Further, believed using K similarity highest user set, family characteristic vector in user's similarity matrix
Breath and product features vector information are modified to user's product features vector information of targeted customer, obtain revised target
User's product features vector information of user.
Specifically, T (u) is k similarity highest user set of user, using K in above-mentioned user's similarity matrix
The user's product features of similarity highest user set, family eigenvector information and product features vector information to targeted customer
Vector information is modified, that is, utilizes similar users characteristic vector amendment puAnd similarity, computed correction simultaneously correct pu:
According to above formula, correction and user's product features vector information of revised targeted customer are got.
S16 is described further:
Commodity push is carried out to targeted customer according to user's product features vector information of the revised targeted customer.
Further, targeted customer is carried out to all commodity using correction value and user's product features vector information
Value interested calculates, and obtains interested value of the targeted customer to all commodity;It is emerging to sense of the targeted customer to all commodity
Interest value carries out descending sort processing, and acquisition order is gathered;There is not the article of behavior to targeted customer according to the order set
Commodity push is carried out to targeted customer.
Utilize puAnd qiCalculating, user is to the interest values of all commodity, and descending arranges to obtain order set R (u), as
Candidate Recommendation collection;Wherein:
Calculated by above formula, obtain interested value of the targeted customer to all commodity, and to targeted customer to all commodity
Value interested carry out descending sort processing, acquisition order gather, sequentially gathered according to this to targeted customer's Recommendations.
Fig. 2 is the system architecture of the personalized recommendation system based on block chain integration service platform in the embodiment of the present invention
Composition schematic diagram, as shown in Fig. 2 the personalized recommendation system includes:
Pretreatment module 11:For being pre-processed to the commodity data that user buys, pre-processed results are obtained;
Build module 12:Fall to sort and user's commodity association table for building commodity user according to the pre-processed results;
Similarity calculation module 13:User's Similarity Measure processing is carried out for falling to sort according to the commodity user, is obtained
Take family similarity matrix;
Characteristic vector pickup module 14:For carrying out characteristic vector pickup processing according to user's commodity association table, obtain
Take family product features vector information;
Correcting module 15:For according to user's similarity matrix and user's product features vector information to target
User's product features vector information of user is modified, and obtains user's product features vector letter of revised targeted customer
Breath;
Recommending module 16:Used for user's product features vector information according to the revised targeted customer to target
Family carries out commodity push.
Preferably, the similarity calculation module 13 includes:
Similarity calculated:For improving cosine similarity formula to the commodity user using much-sought-after item penalty term
Sequence carry out user's Similarity Measure processing, obtain user's similarity matrix and user it is neighbouring user collection.
Preferably, the characteristic vector pickup module 14 includes:
Matrix construction unit:For carrying out building user's commodity matrix disposal using user's commodity association table, obtain
User's commodity matrix;
Matrixing unit:For entering line translation to user's commodity matrix, added during commodity matrixing
Integrated value penalty term and much-sought-after item penalty term, obtain integrated value corresponding to user's commodity;
Sample set construction unit:There is not behavior for the user according to the article set formed objects interested with user
Much-sought-after item diversity structure sample set;
Implicit doctrine analytic unit:Implied for integrated value and the sample set according to corresponding to user's commodity
Doctrine is analyzed, and obtains user's product features vector information;
User's product features vector information includes user characteristics vector information and product features vector information.
Preferably, the correcting module 15 includes:
Amending unit:For being gathered using K similarity highest user in user's similarity matrix, family feature to
Amount information and product features vector information are modified to user's product features vector information of targeted customer, are obtained revised
User's product features vector information of targeted customer.
Preferably, the recommending module 16 includes
Computing unit:For carrying out targeted customer to all business using correction value and user's product features vector information
The value interested of product calculates, and obtains interested value of the targeted customer to all commodity;
Sequencing unit:For carrying out descending sort processing to the value interested of all commodity to the targeted customer, obtain
Order is gathered;
Push unit:For not there is the article of behavior to be carried out to targeted customer to targeted customer according to the order set
Commodity push.
Specifically, the operation principle of the system related functions module of the embodiment of the present invention can be found in the correlation of embodiment of the method
Description, is repeated no more here.
In embodiments of the present invention, the data edge that block chain technology brings is effectively utilized by the embodiment of the present invention,
The authenticity of user integral transaction data and can not tamper;Platform integration " multi-party distribution, free flow " simultaneously, ensure that number
According to the popularity in source and comprehensive;Neighbouring model of the method based on user is similar using the cosine for adding much-sought-after item penalty term
Degree improves similarity definition, and the hidden semantic model for using for reference natural language processing excavates user and commodity hiding information, and fusion is based on
The advantages of neighbouring model of user and hidden semantic model, obtained using the hidden semantic model of neighbour's user characteristics amendment of targeted customer
User characteristics, realize that more accurate individual character is recommended.
One of ordinary skill in the art will appreciate that all or part of step in the various methods of above-described embodiment is can
To instruct the hardware of correlation to complete by program, the program can be stored in a computer-readable recording medium, storage
Medium can include:Read-only storage (ROM, Read Only Memory), random access memory (RAM, Random
Access Memory), disk or CD etc..
A kind of in addition, personalized recommendation based on block chain integration service platform provided above the embodiment of the present invention
Method and system are described in detail, and should employ specific case herein and the principle and embodiment of the present invention are carried out
Illustrate, the explanation of above example is only intended to help the method and its core concept for understanding the present invention;Meanwhile for this area
Those skilled in the art, according to the thought of the present invention, there will be changes in specific embodiments and applications, to sum up
Described, this specification content should not be construed as limiting the invention.
Claims (10)
- A kind of 1. personalized recommendation method based on block chain integration service platform, it is characterised in that the personalized recommendation side Method includes:The commodity data of user's purchase is pre-processed, obtains pre-processed results;Commodity user is built according to the pre-processed results to fall to sort and user's commodity association table;Fall to sort according to the commodity user and carry out user's Similarity Measure processing, obtain user's similarity matrix;Characteristic vector pickup processing is carried out according to user's commodity association table, obtains user's product features vector information;User's product features according to user's similarity matrix and user's product features vector information to targeted customer Vector information is modified, and obtains user's product features vector information of revised targeted customer;Commodity push is carried out to targeted customer according to user's product features vector information of the revised targeted customer.
- 2. the personalized recommendation method according to claim 1 based on block chain integration service platform, it is characterised in that institute State to be fallen to sort according to the commodity user and carry out user's Similarity Measure processing, including:Improving cosine similarity formula using much-sought-after item penalty term, sequence carries out user's similarity meter to the commodity user Calculation handle, obtain user's similarity matrix and user it is neighbouring user collection.
- 3. the personalized recommendation method according to claim 1 based on block chain integration service platform, it is characterised in that institute State and characteristic vector pickup processing is carried out according to user's commodity association table, including:Carry out building user's commodity matrix disposal using user's commodity association table, obtain user's commodity matrix;Line translation is entered to user's commodity matrix, integrated value penalty term and much-sought-after item are added during commodity matrixing Penalty term, obtain integrated value corresponding to user's commodity;The much-sought-after item diversity for not having behavior according to the user of the article set formed objects interested with user builds sample Collection;Implicit doctrine analysis is carried out according to integrated value and the sample set corresponding to user's commodity, obtains user's product features Vector information;User's product features vector information includes user characteristics vector information and product features vector information.
- 4. the personalized recommendation method according to claim 1 based on block chain integration service platform, it is characterised in that institute State according to user's similarity matrix and user's product features vector information to user's product features of targeted customer to Amount information is modified, including:Using K similarity highest user set, family eigenvector information and product features in user's similarity matrix Vector information is modified to user's product features vector information of targeted customer, obtains the user business of revised targeted customer Product eigenvector information.
- 5. the personalized recommendation method according to claim 1 based on block chain integration service platform, it is characterised in that institute State and commodity push is carried out to targeted customer according to user's product features vector information of the revised targeted customer, including:Targeted customer is carried out using correction value and user's product features vector information to calculate the value interested of all commodity, Obtain interested value of the targeted customer to all commodity;Descending sort processing is carried out to the value interested of all commodity to the targeted customer, acquisition order is gathered;The article for not having behavior to targeted customer according to the order set carries out commodity push to targeted customer.
- A kind of 6. personalized recommendation system based on block chain integration service platform, it is characterised in that the personalized recommendation system System includes:Pretreatment module:For being pre-processed to the commodity data that user buys, pre-processed results are obtained;Build module:Fall to sort and user's commodity association table for building commodity user according to the pre-processed results;Similarity calculation module:User's Similarity Measure processing is carried out for falling to sort according to the commodity user, obtains user Similarity matrix;Characteristic vector pickup module:For carrying out characteristic vector pickup processing according to user's commodity association table, user is obtained Product features vector information;Correcting module:For according to user's similarity matrix and user's product features vector information to targeted customer's User's product features vector information is modified, and obtains user's product features vector information of revised targeted customer;Recommending module:Carried out for user's product features vector information according to the revised targeted customer to targeted customer Commodity push.
- 7. the personalized recommendation system according to claim 6 based on block chain integration service platform, it is characterised in that institute Stating similarity calculation module includes:Similarity calculated:The commodity user is fallen to arrange for improving cosine similarity formula using much-sought-after item penalty term Sequence carry out user's Similarity Measure processing, obtain user's similarity matrix and user it is neighbouring user collection.
- 8. the personalized recommendation system according to claim 6 based on block chain integration service platform, it is characterised in that institute Stating characteristic vector pickup module includes:Matrix construction unit:For carrying out building user's commodity matrix disposal using user's commodity association table, user is obtained Commodity matrix;Matrixing unit:For entering line translation to user's commodity matrix, integration is added during commodity matrixing It is worth penalty term and much-sought-after item penalty term, obtains integrated value corresponding to user's commodity;Sample set construction unit:For not there is the heat of behavior according to the user of the article set formed objects interested with user Door commodity diversity structure sample set;Implicit doctrine analytic unit:Implicit doctrine is carried out for integrated value and the sample set according to corresponding to user's commodity Analysis, obtain user's product features vector information;User's product features vector information includes user characteristics vector information and product features vector information.
- 9. the personalized recommendation method according to claim 6 based on block chain integration service platform, it is characterised in that institute Stating correcting module includes:Amending unit:For using K similarity highest user set, family characteristic vector letter in user's similarity matrix Breath and product features vector information are modified to user's product features vector information of targeted customer, obtain revised target User's product features vector information of user.
- 10. the personalized recommendation method according to claim 6 based on block chain integration service platform, it is characterised in that The recommending module includesComputing unit:For carrying out targeted customer to all commodity using correction value and user's product features vector information Value interested calculates, and obtains interested value of the targeted customer to all commodity;Sequencing unit:For carrying out descending sort processing, acquisition order to the value interested of all commodity to the targeted customer Set;Push unit:For not there is the article of behavior to carry out commodity to targeted customer to targeted customer according to the order set Push.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710621801.3A CN107563833A (en) | 2017-07-27 | 2017-07-27 | A kind of personalized recommendation method and system based on block chain integration service platform |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710621801.3A CN107563833A (en) | 2017-07-27 | 2017-07-27 | A kind of personalized recommendation method and system based on block chain integration service platform |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107563833A true CN107563833A (en) | 2018-01-09 |
Family
ID=60974877
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710621801.3A Pending CN107563833A (en) | 2017-07-27 | 2017-07-27 | A kind of personalized recommendation method and system based on block chain integration service platform |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107563833A (en) |
Cited By (18)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104966219A (en) * | 2015-07-21 | 2015-10-07 | Tcl集团股份有限公司 | Personalized collaborative filtering recommendation method and system based on word frequency weighted technology |
CN108256965A (en) * | 2018-01-11 | 2018-07-06 | 杭州秘猿科技有限公司 | A kind of distributed electronic business plateform system based on block chain |
CN108389039A (en) * | 2018-02-26 | 2018-08-10 | 深圳智乾区块链科技有限公司 | Value system management method, device and storage medium based on block chain |
CN108647996A (en) * | 2018-04-11 | 2018-10-12 | 中山大学 | A kind of personalized recommendation method and system based on Spark |
CN109523341A (en) * | 2018-10-12 | 2019-03-26 | 广西师范大学 | The cross-domain recommended method of anonymity based on block chain technology |
CN109544298A (en) * | 2018-11-23 | 2019-03-29 | 丁娜 | A kind of information dissemination method and system based on block chain technology |
CN109743323A (en) * | 2019-01-08 | 2019-05-10 | 中国石油大学(华东) | A kind of Resources Sharing based on block chain technology |
CN109829761A (en) * | 2019-01-31 | 2019-05-31 | 广州视源电子科技股份有限公司 | A kind of commodity selection method, apparatus, equipment and storage medium |
CN109840800A (en) * | 2018-12-14 | 2019-06-04 | 深圳壹账通智能科技有限公司 | A kind of Products Show method, apparatus, storage medium and server based on integral |
CN110827372A (en) * | 2018-08-09 | 2020-02-21 | 普华云创科技(北京)有限公司 | Method, system and storage medium for constructing person label portrait based on block chain |
CN110852811A (en) * | 2019-11-19 | 2020-02-28 | 华扬联众数字技术股份有限公司 | Method, apparatus and machine-readable storage medium for information processing |
CN111292164A (en) * | 2020-01-21 | 2020-06-16 | 上海风秩科技有限公司 | Commodity recommendation method and device, electronic equipment and readable storage medium |
CN112003883A (en) * | 2020-10-29 | 2020-11-27 | 浙江微能科技有限公司 | System for realizing integral accounting by using block chain technology |
CN112215663A (en) * | 2020-10-29 | 2021-01-12 | 广州机不凡信息科技有限公司 | User point currency method, equipment and medium |
CN112487024A (en) * | 2020-12-10 | 2021-03-12 | 广东电力通信科技有限公司 | Power information inquiry and evaluation system |
CN112669083A (en) * | 2020-12-30 | 2021-04-16 | 杭州趣链科技有限公司 | Commodity recommendation method and device and electronic equipment |
CN113609381A (en) * | 2021-07-13 | 2021-11-05 | 杭州网易云音乐科技有限公司 | Work recommendation method, device, medium and computing equipment |
CN114782076A (en) * | 2022-03-28 | 2022-07-22 | 武汉圣男品牌管理有限公司 | Online shopping mall consumption platform lottery and point exchange intelligent management method and system and computer storage medium |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102426686A (en) * | 2011-09-29 | 2012-04-25 | 南京大学 | Internet information product recommending method based on matrix decomposition |
-
2017
- 2017-07-27 CN CN201710621801.3A patent/CN107563833A/en active Pending
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102426686A (en) * | 2011-09-29 | 2012-04-25 | 南京大学 | Internet information product recommending method based on matrix decomposition |
Non-Patent Citations (3)
Title |
---|
FLEDER D: "Recommender Systems and their Impact", 《PROCEEDINGS OF THE 8TH ACM CONFERENCE ON ELECTRONIC COMMERCE》 * |
陈梅梅: "基于改进张量分解模型的个性化推荐算法研究", 《数据分析与知识发现》 * |
陈清浩: "基于SVD的协同过滤推荐算法研究", 《中国优秀硕士学位论文全文数据库 工程科技辑》 * |
Cited By (23)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104966219A (en) * | 2015-07-21 | 2015-10-07 | Tcl集团股份有限公司 | Personalized collaborative filtering recommendation method and system based on word frequency weighted technology |
CN108256965A (en) * | 2018-01-11 | 2018-07-06 | 杭州秘猿科技有限公司 | A kind of distributed electronic business plateform system based on block chain |
CN108389039A (en) * | 2018-02-26 | 2018-08-10 | 深圳智乾区块链科技有限公司 | Value system management method, device and storage medium based on block chain |
CN108647996B (en) * | 2018-04-11 | 2022-04-19 | 中山大学 | Spark-based personalized recommendation method and system |
CN108647996A (en) * | 2018-04-11 | 2018-10-12 | 中山大学 | A kind of personalized recommendation method and system based on Spark |
CN110827372A (en) * | 2018-08-09 | 2020-02-21 | 普华云创科技(北京)有限公司 | Method, system and storage medium for constructing person label portrait based on block chain |
CN109523341A (en) * | 2018-10-12 | 2019-03-26 | 广西师范大学 | The cross-domain recommended method of anonymity based on block chain technology |
CN109544298A (en) * | 2018-11-23 | 2019-03-29 | 丁娜 | A kind of information dissemination method and system based on block chain technology |
CN109840800A (en) * | 2018-12-14 | 2019-06-04 | 深圳壹账通智能科技有限公司 | A kind of Products Show method, apparatus, storage medium and server based on integral |
CN109743323A (en) * | 2019-01-08 | 2019-05-10 | 中国石油大学(华东) | A kind of Resources Sharing based on block chain technology |
CN109829761A (en) * | 2019-01-31 | 2019-05-31 | 广州视源电子科技股份有限公司 | A kind of commodity selection method, apparatus, equipment and storage medium |
CN110852811A (en) * | 2019-11-19 | 2020-02-28 | 华扬联众数字技术股份有限公司 | Method, apparatus and machine-readable storage medium for information processing |
CN111292164A (en) * | 2020-01-21 | 2020-06-16 | 上海风秩科技有限公司 | Commodity recommendation method and device, electronic equipment and readable storage medium |
CN112003883A (en) * | 2020-10-29 | 2020-11-27 | 浙江微能科技有限公司 | System for realizing integral accounting by using block chain technology |
CN112215663A (en) * | 2020-10-29 | 2021-01-12 | 广州机不凡信息科技有限公司 | User point currency method, equipment and medium |
CN112003883B (en) * | 2020-10-29 | 2022-03-11 | 浙江微能科技有限公司 | System for realizing integral accounting by using block chain technology |
CN112487024A (en) * | 2020-12-10 | 2021-03-12 | 广东电力通信科技有限公司 | Power information inquiry and evaluation system |
CN112487024B (en) * | 2020-12-10 | 2023-10-31 | 广东电力通信科技有限公司 | Electric power information inquiry and evaluation system |
CN112669083A (en) * | 2020-12-30 | 2021-04-16 | 杭州趣链科技有限公司 | Commodity recommendation method and device and electronic equipment |
CN113609381A (en) * | 2021-07-13 | 2021-11-05 | 杭州网易云音乐科技有限公司 | Work recommendation method, device, medium and computing equipment |
CN113609381B (en) * | 2021-07-13 | 2023-12-12 | 杭州网易云音乐科技有限公司 | Work recommendation method, device, medium and computing equipment |
CN114782076A (en) * | 2022-03-28 | 2022-07-22 | 武汉圣男品牌管理有限公司 | Online shopping mall consumption platform lottery and point exchange intelligent management method and system and computer storage medium |
CN114782076B (en) * | 2022-03-28 | 2023-03-31 | 广东圣千科技有限公司 | Online mall consumption platform lottery integral exchange intelligent management method, system and computer storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107563833A (en) | A kind of personalized recommendation method and system based on block chain integration service platform | |
White et al. | Characterizing the OpenSea NFT marketplace | |
Sari et al. | History of Islamic bank in Indonesia: Issues behind its establishment | |
CN103678672B (en) | Method for recommending information | |
Xu et al. | Use of machine learning to detect wildlife product promotion and sales on Twitter | |
CN102184364A (en) | Semi-supervised learning-based recommendation system shilling attack detection method | |
CN104899229A (en) | Swarm intelligence based behavior clustering system | |
CN106484919A (en) | A kind of industrial sustainability sorting technique based on webpage autonomous word and system | |
CN108268439A (en) | The processing method and processing device of text emotion | |
CN109033132A (en) | The method and device of text and the main body degree of correlation are calculated using knowledge mapping | |
CN109993538A (en) | Identity theft detection method based on probability graph model | |
Sakız et al. | Blockchain beyond cryptocurrency: non-fungible tokens | |
Chang et al. | Detecting fake review with rumor model—Case study in hotel review | |
Liang et al. | Personalized recommender system based on item taxonomy and folksonomy | |
Ke et al. | An unsupervised detection framework for Chinese jargons in the darknet | |
Taylor III et al. | The Reverse-Morals Clause: The Unique Way to Save Talent's Reputation and Money in a New Era of Corporate Crimes and Scandals | |
Prasetyo et al. | Evaluation of feature extraction TF-IDF in Indonesian hoax news classification | |
Allahbakhsh et al. | Detecting, representing and querying collusion in online rating systems | |
CN104035969B (en) | Feature Words base construction method and system in social networks | |
Hokianto | Non-fungible tokens: A literature review | |
Yuxian et al. | Investors are social animals: Predicting investor behavior using social network features via supervised learning approach | |
Wu et al. | An improved contextual advertising matching approach based on wikipedia knowledge | |
Wang et al. | An analysis of Appalachian hardwood products in the Chinese market | |
Zhang | Analysis of love-hate shilling attack against e-commerce recommender system | |
Li et al. | Towards understanding and characterizing the arbitrage bot scam in the wild |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180109 |