CN108921673B - Commodity recommendation method based on big data - Google Patents
Commodity recommendation method based on big data Download PDFInfo
- Publication number
- CN108921673B CN108921673B CN201810779565.2A CN201810779565A CN108921673B CN 108921673 B CN108921673 B CN 108921673B CN 201810779565 A CN201810779565 A CN 201810779565A CN 108921673 B CN108921673 B CN 108921673B
- Authority
- CN
- China
- Prior art keywords
- user
- commodity
- class
- users
- binary classifier
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0255—Targeted advertisements based on user history
Abstract
The invention discloses a commodity recommendation method based on big data, which comprises the following steps: s1: dividing all commodities into a plurality of commodity classes according to the commodity types; s2: extracting the characteristics of commodities in the same commodity class, carrying out SVM machine learning and generating a binary classifier of the commodity class; the number of the binary classifiers is the same as that of the commodity classes, and the binary classifiers correspond to the commodity classes one to one; s3: extracting browsing and transaction records of all users on the platform from the big data as characteristic values of the users; s4: and performing clustering analysis on all users according to the characteristic values of the users to generate a plurality of user classes. According to the commodity recommendation method based on the big data, when real-time recommendation is performed in the using process of a user, the calculation amount required by the system end is only the calculation of the binary classifier, and the calculation amount of the binary classifier after the binary classifier is formed is very small, so that the recommendation speed can be greatly improved compared with a point-to-point recommendation mode in the prior art.
Description
Technical Field
The invention relates to the technical field of commodity recommendation, in particular to a commodity recommendation method based on big data.
Background
With the rapid development of the internet, the amount of information presented on the internet has increased explosively. Over 1000 million items were offered in 2012 on Amazon shopping websites. The increase of the information quantity reduces the utilization rate of the information, so that the user is more difficult to find the required information, and the phenomenon of information overload occurs. At present, commodity personalized recommendation is generally considered to be one of the most effective tools for solving the problem, effective information is recommended to a user through analyzing behavior habits of the user, time for the user to screen the information is saved, and the effective utilization rate of the information is improved.
At present, the method of directly matching and comparing a single user and a single commodity is mainly adopted for commodity recommendation, the calculation amount is increased geometrically for tens of millions of commodities and tens of millions of users at present, and the operation cost of commodity recommendation is greatly improved.
Disclosure of Invention
The invention aims to solve the technical problem that the current commodity recommendation mainly adopts a method for directly matching and comparing a single user with a single commodity, the computation amount of tens of millions of commodities and tens of millions of users at present can increase geometrically, the operation cost of commodity recommendation is greatly improved, and the invention aims to provide a commodity recommendation method based on big data and solve the problems.
The invention is realized by the following technical scheme:
the commodity recommendation method based on big data comprises the following steps: s1: dividing all commodities into a plurality of commodity classes according to the commodity types; s2: extracting the characteristics of commodities in the same commodity class, carrying out SVM machine learning and generating a binary classifier of the commodity class; the number of the binary classifiers is the same as that of the commodity classes, and the binary classifiers correspond to the commodity classes one to one; s3: extracting browsing and transaction records of all users on the platform from the big data as characteristic values of the users; s4: performing clustering analysis on all users according to the characteristic values of the users to generate a plurality of user classes; s5: extracting the characteristic values of the users in the same user class and carrying out weighted average to obtain the characteristic mean value of each user class; s6: substituting the feature mean into each binary classifier; when the output result of any binary classifier is a true value, recommending the commodity class corresponding to the binary classifier to each user in the user class corresponding to the characteristic mean value.
In the prior art, the commodity recommendation mainly adopts a method of directly matching and comparing a single user with a single commodity, and for tens of millions of commodities and tens of millions of users at present, the calculation amount is increased geometrically, so that the operation cost of the commodity recommendation is greatly increased. When the method is applied, all commodities are divided into a plurality of commodity classes according to commodity types, a general electronic commerce platform can have perfect commodity classification, so that the calculation amount is basically not needed, then the characteristics of the commodities in the same commodity class are extracted to carry out SVM machine learning and generate a binary classifier of the commodity class, the binary classifier corresponds to the preference of a user, the input information is the characteristic value of the user, the output information is a true value (true) or a false value (false), and the operation of the step is equivalent to the pretreatment carried out by the platform; then, browsing and transaction records of all users on the platform are extracted from the big data to serve as characteristic values of the users, the characteristic values can be provided by the big data on the platform and can also be obtained from the big data on other platforms, with the development of big data technology, the information sharing of the big data platform is more and more developed, and the amount of the obtained information is more and more abundant; and then, performing cluster analysis on all users according to the characteristic values of the users to generate a plurality of user classes, wherein each user class corresponds to a preference which may correspond to one commodity or a plurality of commodities, for example, the user is an automobile industry buyer, and the preference of the user may correspond to a plurality of commodities such as aluminum, steel, rubber and the like. In order to locate the common preference in the user classes, the feature mean is brought into each binary classifier; when the output result of any binary classifier is a true value, the commodity class corresponding to the binary classifier is recommended to each user in the user class corresponding to the characteristic mean value, so that the calculation amount required by the system end is only the calculation of the binary classifier when the user uses the commodity class for real-time recommendation, and the calculation amount of the binary classifier after the binary classifier is formed is very small, so that the recommendation speed can be greatly increased compared with a point-to-point recommendation method in the prior art.
Further, in step S2, the binary classifier is a linear binary classifier.
When the method is applied, in order to further improve the recommendation efficiency, a linear binary classifier is adopted.
Further, the method also comprises the following steps: s7: when a new user joins the platform, extracting browsing and transaction records of the user from the big data as characteristic values of the new user; and obtaining the distances between the new user and all user classes according to the characteristics of the new user, and adding the new user into the user class with the shortest distance to the new user.
When the method is applied, in order to quickly match the preference of the new user, the method of directly comparing the new user with the existing user classes is adopted, the new user is distributed into the existing classes to recommend the commodities, the new user does not need to match and compare the new user with tens of millions of commodities again, the recommendation efficiency is effectively improved, and the calculation amount is reduced.
Further, the distance in step S7 is a cosine distance.
Further, step S5 includes the following steps: s51: obtaining the central point of the user class according to the characteristic value of the user in the same user class; s52: and taking the reciprocal of the distance from the user to the user class central point as a weight to carry out weighted average on the characteristic values of the user to obtain a characteristic mean value of the user class.
When the method is applied, in order to further simplify the operation process, the user carries out cluster analysis to form a user class, so that the inventor adopts an operation mode of the distance from the central point to obtain the weight of the user, the recommendation efficiency is improved, the operation amount is reduced, and the user experience is effectively improved.
Further, the distance in step S52 is a cosine distance.
Compared with the prior art, the invention has the following advantages and beneficial effects:
according to the commodity recommendation method based on the big data, when real-time recommendation is performed in the using process of a user, the calculation amount required by the system end is only the calculation of the binary classifier, and the calculation amount of the binary classifier after the binary classifier is formed is very small, so that the recommendation speed can be greatly improved compared with a point-to-point recommendation mode in the prior art.
Drawings
The accompanying drawings, which are included to provide a further understanding of the embodiments of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the principles of the invention. In the drawings:
FIG. 1 is a schematic diagram of the steps of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to examples and accompanying drawings, and the exemplary embodiments and descriptions thereof are only used for explaining the present invention and are not meant to limit the present invention.
Example 1
As shown in fig. 1, the commodity recommendation method based on big data of the present invention includes the following steps: s1: dividing all commodities into a plurality of commodity classes according to the commodity types; s2: extracting the characteristics of commodities in the same commodity class, carrying out SVM machine learning and generating a binary classifier of the commodity class; the number of the binary classifiers is the same as that of the commodity classes, and the binary classifiers correspond to the commodity classes one to one; s3: extracting browsing and transaction records of all users on the platform from the big data as characteristic values of the users; s4: performing clustering analysis on all users according to the characteristic values of the users to generate a plurality of user classes; s5: extracting the characteristic values of the users in the same user class and carrying out weighted average to obtain the characteristic mean value of each user class; s6: substituting the feature mean into each binary classifier; when the output result of any binary classifier is a true value, recommending the commodity class corresponding to the binary classifier to each user in the user class corresponding to the characteristic mean value.
In the implementation of this embodiment, all the commodities are divided into a plurality of commodity classes according to the commodity types, and a general e-commerce platform has perfect commodity classification, so that no calculation amount is basically needed, and then the characteristics of the commodities in the same commodity class are extracted to perform SVM machine learning and generate a binary classifier of the commodity class, wherein the binary classifier corresponds to the preference of a user, the input information is the characteristic value of the user, the output information is a true value (true) or a false value (false), and the operation of this step is equivalent to the preprocessing performed by the platform; then, browsing and transaction records of all users on the platform are extracted from the big data to serve as characteristic values of the users, the characteristic values can be provided by the big data on the platform and can also be obtained from the big data on other platforms, with the development of big data technology, the information sharing of the big data platform is more and more developed, and the amount of the obtained information is more and more abundant; and then, performing cluster analysis on all users according to the characteristic values of the users to generate a plurality of user classes, wherein each user class corresponds to a preference which may correspond to one commodity or a plurality of commodities, for example, the user is an automobile industry buyer, and the preference of the user may correspond to a plurality of commodities such as aluminum, steel, rubber and the like. In order to locate the common preference in the user classes, the feature mean is brought into each binary classifier; when the output result of any binary classifier is a true value, the commodity class corresponding to the binary classifier is recommended to each user in the user class corresponding to the characteristic mean value, so that the calculation amount required by the system end is only the calculation of the binary classifier when the user uses the commodity class for real-time recommendation, and the calculation amount of the binary classifier after the binary classifier is formed is very small, so that the recommendation speed can be greatly increased compared with a point-to-point recommendation method in the prior art.
Example 2
In this embodiment, on the basis of embodiment 1, the binary classifier in step S2 adopts a linear binary classifier.
In this embodiment, in order to further improve the recommendation efficiency, a linear binary classifier is used.
Example 3
The embodiment further includes the following steps based on embodiment 1: s7: when a new user joins the platform, extracting browsing and transaction records of the user from the big data as characteristic values of the new user; and obtaining the distances between the new user and all user classes according to the characteristics of the new user, and adding the new user into the user class with the shortest distance to the new user.
In the implementation of the embodiment, in order to quickly match the preferences of the new user, the method of directly comparing the preferences of the new user with the existing user classes is adopted, the new user is distributed into the existing classes to recommend the commodities, and the new user does not need to match and compare the new user with tens of millions of commodities again, so that the recommendation efficiency is effectively improved, and the calculation amount is reduced. The distance is a cosine distance in step S7.
Example 4
In this embodiment, on the basis of embodiment 1, step S5 includes the following steps: s51: obtaining the central point of the user class according to the characteristic value of the user in the same user class; s52: and taking the reciprocal of the distance from the user to the user class central point as a weight to carry out weighted average on the characteristic values of the user to obtain a characteristic mean value of the user class.
When the embodiment is implemented, in order to further simplify the operation process, since the user already performs cluster analysis to form the user class, the inventor adopts an operation mode of the distance from the central point to obtain the weight of the user, so that the recommendation efficiency is improved, the operation amount is reduced, and the user experience is effectively improved. The distance is a cosine distance in step S52.
The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are merely exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (6)
1. The commodity recommendation method based on big data is characterized by comprising the following steps:
s1: dividing all commodities into a plurality of commodity classes according to the commodity types;
s2: extracting the characteristics of commodities in the same commodity class, carrying out SVM machine learning and generating a binary classifier of the commodity class; the number of the binary classifiers is the same as that of the commodity classes, and the binary classifiers correspond to the commodity classes one to one;
s3: extracting browsing and transaction records of all users on the platform from the big data as characteristic values of the users;
s4: performing clustering analysis on all users according to the characteristic values of the users to generate a plurality of user classes;
s5: extracting the characteristic values of the users in the same user class and carrying out weighted average to obtain the characteristic mean value of each user class;
s6: substituting the feature mean into each binary classifier; when the output result of any binary classifier is a true value, recommending the commodity class corresponding to the binary classifier to each user in the user class corresponding to the characteristic mean value.
2. The big-data-based commodity recommendation method according to claim 1, wherein said binary classifier in step S2 is a linear binary classifier.
3. The big-data-based commodity recommendation method according to claim 1, further comprising the steps of:
s7: when a new user joins the platform, extracting browsing and transaction records of the user from the big data as characteristic values of the new user; and obtaining the distances between the new user and all user classes according to the characteristics of the new user, and adding the new user into the user class with the shortest distance to the new user.
4. The big-data-based commodity recommendation method according to claim 3, wherein said distance in step S7 is a cosine distance.
5. The big-data-based commodity recommendation method according to claim 1, wherein the step S5 comprises the steps of:
s51: obtaining the central point of the user class according to the characteristic value of the user in the same user class;
s52: and taking the reciprocal of the distance from the user to the user class central point as a weight to carry out weighted average on the characteristic values of the user to obtain a characteristic mean value of the user class.
6. The big-data-based commodity recommendation method according to claim 5, wherein said distance in step S52 is a cosine distance.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810779565.2A CN108921673B (en) | 2018-07-16 | 2018-07-16 | Commodity recommendation method based on big data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810779565.2A CN108921673B (en) | 2018-07-16 | 2018-07-16 | Commodity recommendation method based on big data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108921673A CN108921673A (en) | 2018-11-30 |
| CN108921673B true CN108921673B (en) | 2021-06-01 |
Family
ID=64411929
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810779565.2A Active CN108921673B (en) | 2018-07-16 | 2018-07-16 | Commodity recommendation method based on big data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108921673B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111612557B (en) * | 2019-02-25 | 2023-06-30 | 阿里巴巴集团控股有限公司 | Method and device for providing commodity object information and electronic equipment |
| CN111709812A (en) * | 2020-06-19 | 2020-09-25 | 山东财经大学 | E-commerce platform commodity recommendation method and system based on user dynamic classification |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103399891A (en) * | 2013-07-22 | 2013-11-20 | 百度在线网络技术（北京）有限公司 | Method, device and system for automatic recommendation of network content |
| CN104616178A (en) * | 2015-03-06 | 2015-05-13 | 浪潮集团有限公司 | Recommendation method of E-commerce goods based on big-data multi-label classification method |
| CN105389713A (en) * | 2015-10-15 | 2016-03-09 | 南京大学 | Mobile data traffic package recommendation algorithm based on user historical data |
| CN105787770A (en) * | 2016-04-27 | 2016-07-20 | 上海遥薇(集团)有限公司 | Non-negative matrix factorization (NMF) algorithm-based big data commodity and service recommending method and system |
| CN106157114A (en) * | 2016-07-06 | 2016-11-23 | 商宴通（上海）网络科技有限公司 | Have dinner based on user the homepage proposed algorithm of behavior modeling |
| CN106294783A (en) * | 2016-08-12 | 2017-01-04 | 乐视控股（北京）有限公司 | A kind of video recommendation method and device |
| CN106407379A (en) * | 2016-09-13 | 2017-02-15 | 天津大学 | Hadoop platform based movie recommendation method |
| CN107360222A (en) * | 2017-06-30 | 2017-11-17 | 广东欧珀移动通信有限公司 | Merchandise news method for pushing, device, storage medium and server |
| CN107480719A (en) * | 2017-08-17 | 2017-12-15 | 广东工业大学 | A kind of recommendation method and system of the skin-protection product based on skin properties evaluation |
| CN107578332A (en) * | 2017-09-22 | 2018-01-12 | 深圳乐信软件技术有限公司 | A kind of method, apparatus, equipment and storage medium for recommending cash commodity |
| CN107633430A (en) * | 2017-09-20 | 2018-01-26 | 哈尔滨工业大学 | A kind of Method of Commodity Recommendation based on community of colony |
| CN107888793A (en) * | 2017-11-15 | 2018-04-06 | 中国联合网络通信集团有限公司 | Package recommendation method, apparatus and system |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105357583B (en) * | 2015-10-16 | 2019-08-02 | Tcl集团股份有限公司 | A kind of method and apparatus finding smart television user interest preference |
| CN106127525A (en) * | 2016-06-27 | 2016-11-16 | 浙江大学 | A kind of TV shopping Method of Commodity Recommendation based on sorting algorithm |
| CN107424043B (en) * | 2017-06-15 | 2021-06-25 | 北京三快在线科技有限公司 | Product recommendation method and device and electronic equipment |
| CN107704485A (en) * | 2017-07-31 | 2018-02-16 | 北京拉勾科技有限公司 | A kind of position recommends method and computing device |
-
2018
- 2018-07-16 CN CN201810779565.2A patent/CN108921673B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103399891A (en) * | 2013-07-22 | 2013-11-20 | 百度在线网络技术（北京）有限公司 | Method, device and system for automatic recommendation of network content |
| CN104616178A (en) * | 2015-03-06 | 2015-05-13 | 浪潮集团有限公司 | Recommendation method of E-commerce goods based on big-data multi-label classification method |
| CN105389713A (en) * | 2015-10-15 | 2016-03-09 | 南京大学 | Mobile data traffic package recommendation algorithm based on user historical data |
| CN105787770A (en) * | 2016-04-27 | 2016-07-20 | 上海遥薇(集团)有限公司 | Non-negative matrix factorization (NMF) algorithm-based big data commodity and service recommending method and system |
| CN106157114A (en) * | 2016-07-06 | 2016-11-23 | 商宴通（上海）网络科技有限公司 | Have dinner based on user the homepage proposed algorithm of behavior modeling |
| CN106294783A (en) * | 2016-08-12 | 2017-01-04 | 乐视控股（北京）有限公司 | A kind of video recommendation method and device |
| CN106407379A (en) * | 2016-09-13 | 2017-02-15 | 天津大学 | Hadoop platform based movie recommendation method |
| CN107360222A (en) * | 2017-06-30 | 2017-11-17 | 广东欧珀移动通信有限公司 | Merchandise news method for pushing, device, storage medium and server |
| CN107480719A (en) * | 2017-08-17 | 2017-12-15 | 广东工业大学 | A kind of recommendation method and system of the skin-protection product based on skin properties evaluation |
| CN107633430A (en) * | 2017-09-20 | 2018-01-26 | 哈尔滨工业大学 | A kind of Method of Commodity Recommendation based on community of colony |
| CN107578332A (en) * | 2017-09-22 | 2018-01-12 | 深圳乐信软件技术有限公司 | A kind of method, apparatus, equipment and storage medium for recommending cash commodity |
| CN107888793A (en) * | 2017-11-15 | 2018-04-06 | 中国联合网络通信集团有限公司 | Package recommendation method, apparatus and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108921673A (en) | 2018-11-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106296059B (en) | Method and equipment for determining delivery network points | |
| TWI787196B (en) | Method, device and system for generating business object attribute identification | |
| CN107330752B (en) | Method and device for identifying brand words | |
| TWI508011B (en) | Category information providing method and device | |
| US11281860B2 (en) | Method, apparatus and device for recognizing text type | |
| TWI582619B (en) | Method and apparatus for providing referral words | |
| CN107516235B (en) | Commodity preference estimation method and device | |
| CN109816482B (en) | Knowledge graph construction method, device and equipment of e-commerce platform and storage medium | |
| CN103870507B (en) | Method and device of searching based on category | |
| CN110827112B (en) | Deep learning commodity recommendation method and device, computer equipment and storage medium | |
| CN105426528A (en) | Retrieving and ordering method and system for commodity data | |
| CN103246980A (en) | Information output method and server | |
| US11741094B2 (en) | Method and system for identifying core product terms | |
| CN106296368A (en) | A kind of vehicle commending system and method | |
| CN108073667B (en) | Method for generating user browsing attributes, and non-transitory computer readable medium | |
| CN109460519B (en) | Browsing object recommendation method and device, storage medium and server | |
| TW201939378A (en) | Resource recommendation method and device | |
| CN104751354A (en) | Advertisement cluster screening method | |
| CN103886486A (en) | Electronic commerce recommending method based on support vector machine (SVM) | |
| US20190340176A1 (en) | System and method for data mining and similarity estimation | |
| JP2015525418A (en) | Search method and apparatus | |
| CN108921673B (en) | Commodity recommendation method based on big data | |
| CN107886382B (en) | Method, device and system for analyzing channel drainage effect in website | |
| CN103426089B (en) | Commodity purchase guiding system and method | |
| CN106372956B (en) | Method and system for identifying intention entity based on user search log |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20200806 Address after: 510000 room 711, floor 7, No. 733 (1), Dongfeng East Road, Yuexiu District, Guangzhou City, Guangdong Province (only for office use) Applicant after: GUANGZHOU TENTCOO SOFTWARE TECHNOLOGY Co.,Ltd. Address before: 510000 Guangdong Guangzhou Haizhuqu District Xingang East Road 1000 1103, 1104, 1105, 1106 rooms Applicant before: GUANGZHOU YOUMI TECHNOLOGY Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |