CN106980989B - Merchant recommendation method based on user behavior characteristic analysis - Google Patents
Merchant recommendation method based on user behavior characteristic analysis Download PDFInfo
- Publication number
- CN106980989B CN106980989B CN201611243256.0A CN201611243256A CN106980989B CN 106980989 B CN106980989 B CN 106980989B CN 201611243256 A CN201611243256 A CN 201611243256A CN 106980989 B CN106980989 B CN 106980989B
- Authority
- CN
- China
- Prior art keywords
- merchant
- user
- behavior
- merchants
- characteristic analysis
- 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
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0255—Targeted advertisements based on user history
-
- 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
- G06Q30/0282—Rating or review of business operators or products
Landscapes
- Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Engineering & Computer Science (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Finance (AREA)
- Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a merchant recommendation method based on user behavior characteristic analysis, which comprises the following steps: collecting historical behavior data of a user and constructing a user-merchant scoring matrix based on the historical behavior data; calculating the transfer probability of the user from the source commercial tenant to the target commercial tenant based on the user-commercial tenant scoring matrix aiming at all different commercial tenant pairs consisting of the source commercial tenant and the target commercial tenant; for each user, calculating a current rating for each merchant based on the calculated transition probability, and then ranking all merchants in order from large to small based on the current rating for each merchant to obtain a merchant recommendation list. The merchant recommendation method based on the user behavior characteristic analysis disclosed by the invention has high matching degree and accuracy.
Description
Technical Field
The invention relates to a merchant recommendation method, in particular to a merchant recommendation method based on user behavior characteristic analysis.
Background
Currently, with the increasing popularity of computer and network applications and the increasing abundance of business categories in different fields, it becomes more and more important to analyze the behavior habits of a user using the user's historical behavior data to recommend a matching target merchant to the user.
The existing technical scheme generally adopts simple attribute-based classification of historical behavior data of users and determines target merchants matched with different types of users according to classification results.
However, the above prior art solutions have the following problems: the matching degree and accuracy of the target merchant are low because deeper analysis cannot be performed according to the characteristics of the historical behavior data of the user.
Therefore, there is a need for: the merchant recommendation method based on the user behavior characteristic analysis is high in matching degree and accuracy.
Disclosure of Invention
In order to solve the problems existing in the prior art, the invention provides the merchant recommendation method based on the user behavior characteristic analysis, which has high matching degree and accuracy.
The purpose of the invention is realized by the following technical scheme:
a merchant recommending method based on user behavior characteristic analysis comprises the following steps:
(A1) collecting historical behavior data of a user and constructing a user-merchant scoring matrix based on the historical behavior data, wherein each row in the user-merchant scoring matrix represents all historical behavior records of the user, namely, non-empty elements in the row are all historical behavior records of the user for merchants corresponding to columns of the non-empty elements;
(A2) calculating a transfer probability of the user from the source merchant to the destination merchant based on the user-merchant scoring matrix for all different merchant pairs consisting of the source merchant and the destination merchant, wherein the source merchant and the destination merchant in each merchant pair are different merchants defined in the user-merchant scoring matrix;
(A3) for each user, calculating a current rating for each merchant based on the calculated transition probability, and then ranking all merchants in order from large to small based on the current rating for each merchant to obtain a merchant recommendation list.
In the above-disclosed aspect, preferably, the step (a2) further includes: determining a set of leading states Q of each user's behavior for each merchant based on the user-merchant scoring matrixujIt represents the collection of the pre-states of all behavior records of the user u to the merchant j, wherein the pre-state of the primary behavior record of the user u to the merchant j is defined as follows: regarding the one-time behavior record of the user u to the merchant j, taking the one-time behavior record as the latest one-time behavior record in a preset time range, and taking all behavior records of the user u between the latest one-time behavior record and the last behavior record with the same type as the latest one-time behavior record as the preposed state of the one-time behavior record of the user u to the merchant j.
In the above-disclosed aspect, preferably, the step (a2) further includes: based on the set of prepositions QujThe calculation is for user u, who is at merchant iAll behavior records of (2) belonging to a set Q of leading states of a user u for a merchant jujTotal number of times of behavior records CuijI.e. the behavior of user u at merchant i is recorded in the set of leading states QujWherein u, i, j are all positive integers smaller than a predetermined threshold, and corresponding C is calculated for all combinations of u, i, j valuesuij。
In the above-disclosed aspect, preferably, the step (a2) further includes: based on calculated CuijCalculating each pre-state reference value S as followsuij：
Suij＝log(Cuij+e-1)
In the above-disclosed aspect, preferably, the step (a2) further includes: based on calculated SuijThe transition probability P for each merchant pair consisting of a source merchant and a destination merchant is calculated as followsij：
Where U is a set of users, and n (i) and n (j) represent the total number of times merchant i and merchant j appear in the pre-state of all users, respectively.
In the above-disclosed aspect, preferably, the step (a3) further includes: based on calculated PijThe current rating P of each merchant for each user u is calculated as followsuj：
Where N (u) is the preposition state of user u, S (j, K) is the set of K merchants with the highest transition probability for merchant j, and W is the number of merchants with the highest transition probability for merchant juiIs the weight.
In the above-disclosed aspect, preferably, the step (a3) further includes: calculate W as followsui：
wui＝log(Cui+e-1)
Wherein, CuiIs the number of times merchant i appears in the pre-state recorded for user u's last activity.
In the above-disclosed solution, preferably, the merchant recommendation list includes only the top N merchants in the ranking result, where N is a positive integer.
In the above-disclosed aspect, preferably, the step (a3) further includes: and performing secondary optimization operation on the sorting result after sorting all the merchants in the order from big to small based on the current score of each merchant.
In the solution disclosed above, preferably, the second optimization operation includes: and judging whether the merchants need to directly promote the ranking or the weighting according to a preset rule aiming at each sequenced merchant, and if the merchants need to directly promote the ranking or the weighting, directly promoting the ranking or the weighting, thereby obtaining a final merchant recommendation list.
In the above-disclosed aspect, preferably, the step (a3) further includes: and after the final merchant recommendation list is obtained, sending the final merchant recommendation list to a user terminal so as to display the final merchant recommendation list to the user terminal and recommend merchants.
The merchant recommendation method based on the user behavior characteristic analysis disclosed by the invention has the following advantages: due to the fact that characteristics of historical behavior data of the user can be deeply analyzed and mined, a target merchant matched with the user with specific characteristics can be more accurately determined.
Drawings
The features and advantages of the present invention will be better understood by those skilled in the art when considered in conjunction with the accompanying drawings, in which:
fig. 1 is a flowchart of a merchant recommendation method based on user behavior characteristic analysis according to an embodiment of the present invention.
Detailed Description
FIG. 1 is a flow diagram of a merchant recommendation method based on user behavior specific analysis, according to an embodiment of the invention. As shown in fig. 1, the merchant recommendation method based on user behavior characteristic analysis disclosed by the present invention includes the following steps: (A1) collecting historical behavior data (for example, historical consumption behavior data) of a user and constructing a user-merchant scoring matrix based on the historical behavior data, wherein each row in the user-merchant scoring matrix represents all historical behavior records of the user, that is, non-empty elements in the row are all historical behavior records of the user for merchants corresponding to columns where the non-empty elements are located; (A2) calculating a transfer probability of the user from the source merchant to the destination merchant based on the user-merchant scoring matrix for all different merchant pairs consisting of the source merchant and the destination merchant, wherein the source merchant and the destination merchant in each merchant pair are different merchants defined in the user-merchant scoring matrix; (A3) for each user, calculating a current rating for each merchant based on the calculated transition probability, and then ranking all merchants in order from large to small based on the current rating for each merchant to obtain a merchant recommendation list.
Preferably, in the merchant recommendation method based on the user behavior characteristic analysis disclosed in the present invention, the step (a2) further includes: determining a set of leading states Q of each user's behavior for each merchant based on the user-merchant scoring matrixujIt represents the collection of the pre-states of all behavior records of the user u to the merchant j, wherein the pre-state of the primary behavior record of the user u to the merchant j is defined as follows: regarding the one-time behavior record of the user u to the merchant j, the one-time behavior record is taken as the latest one-time behavior record in a preset time range, and all behavior records (i.e. the last behavior record and the latest one-time behavior record are not included, and only the behavior records between the last behavior record and the latest one-time behavior record are taken as the advanced state of the one-time behavior record of the user u to the merchant j) between the latest one-time behavior record and the last behavior record with the same type as the latest one-time behavior record (the last behavior record is recorded in the preset time range, for example, the last consumption of the user at the merchant with the same type as the merchant j).
Preferably, in the present invention disclosedIn the merchant recommendation method for user behavior characteristic analysis, the step (a2) further includes: based on the set of prepositions QujCompute a set of pre-states Q for user u for merchant j that belong to user u in all behavior records at merchant iujTotal number of times of behavior records CuijI.e. the behavior of user u at merchant i is recorded in the set of leading states QujWherein u, i, j are all positive integers less than a predetermined threshold (i.e., u is any one positive integer less than the total number of users N, i, j are any two different positive integers less than the total number of merchants M), and corresponding C is calculated for all combinations of u, i, j valuesuij。
Preferably, in the merchant recommendation method based on the user behavior characteristic analysis disclosed in the present invention, the step (a2) further includes: based on calculated CuijCalculating each pre-state reference value S as followsuij：
Suij＝log(Cuij+e-1)
Preferably, in the merchant recommendation method based on the user behavior characteristic analysis disclosed in the present invention, the step (a2) further includes: based on calculated SuijThe transition probability P for each merchant pair consisting of a source merchant and a destination merchant is calculated as followsij(i.e., the probability of the user transitioning from source merchant i to destination merchant j for consumption):
where U is a set of users, and n (i) and n (j) represent the total number of times merchant i and merchant j appear in the pre-state of all users, respectively.
Preferably, in the merchant recommendation method based on the user behavior characteristic analysis disclosed in the present invention, the step (a3) further includes: based on calculated PijThe current rating P of each merchant for each user u is calculated as followsuj：
Where N (u) is the preposition state of user u, S (j, K) is the set of K merchants with the highest transition probability for merchant j, and W is the number of merchants with the highest transition probability for merchant juiIs the weight.
Preferably, in the merchant recommendation method based on the user behavior characteristic analysis disclosed in the present invention, the step (a3) further includes: calculate W as followsui：
wui＝log(Cui+e-1)
Wherein, CuiIs the number of times merchant i appears in the pre-state recorded for user u's last activity.
Preferably, in the merchant recommendation method based on the user behavior characteristic analysis disclosed in the present invention, the merchant recommendation list only includes the top N (e.g. 10) merchants in the ranking result, where N is a positive integer.
Preferably, in the merchant recommendation method based on the user behavior characteristic analysis disclosed in the present invention, the step (a3) further includes: and performing secondary optimization operation on the sorting result after sorting all the merchants in the order from big to small based on the current score of each merchant.
Preferably, in the merchant recommendation method based on the user behavior characteristic analysis disclosed in the present invention, the secondary optimization operation includes: and judging whether the merchants need to directly promote ranking or weighting according to a preset rule for each sorted merchant (for example, the merchants have special benefits to the user and the like), and if the merchants need to directly promote ranking or weighting, performing direct promotion ranking operation or weighting operation on the merchants to obtain a final merchant recommendation list.
Preferably, in the merchant recommendation method based on the user behavior characteristic analysis disclosed in the present invention, the step (a3) further includes: and after the final merchant recommendation list is obtained, sending the final merchant recommendation list to a user terminal so as to display the final merchant recommendation list to the user terminal and recommend merchants.
Therefore, the merchant recommendation method based on the user behavior characteristic analysis disclosed by the invention has the following advantages: due to the fact that characteristics of historical behavior data of the user can be deeply analyzed and mined, a target merchant matched with the user with specific characteristics can be more accurately determined.
Although the present invention has been described in connection with the preferred embodiments, its mode of implementation is not limited to the embodiments described above. It should be appreciated that: various changes and modifications can be made by one skilled in the art without departing from the spirit and scope of the invention.
Claims (7)
1. A merchant recommending method based on user behavior characteristic analysis comprises the following steps:
(A1) collecting historical behavior data of a user and constructing a user-merchant scoring matrix based on the historical behavior data, wherein each row in the user-merchant scoring matrix represents all historical behavior records of the user, namely, non-empty elements in the row are all historical behavior records of the user for merchants corresponding to columns of the non-empty elements;
(A2) calculating, for all different merchant pairs consisting of source merchants and destination merchants, a transfer probability of the user from the source merchant to the destination merchant based on the user-merchant scoring matrix, wherein the source merchant and the destination merchant in each merchant pair are different merchants defined in the user-merchant scoring matrix, and the transfer probability represents a probability of consumption from the source merchant to the destination merchant;
(A3) calculating, for each user, a current rating for each merchant based on the calculated transition probability, and then ranking all merchants in descending order based on the current rating for each merchant to obtain a merchant recommendation list,
wherein the step (A2) further comprises: determining a set of leading states Q of each user's behavior for each merchant based on the user-merchant scoring matrixujWhich represents all behavioral remembers of user u to merchant jA collection of recorded pre-states, wherein the pre-states recorded by the user u for a behavior of the merchant j are defined as follows: regarding the one-time behavior record of the user u to the merchant j, taking the one-time behavior record as the latest one-time behavior record in a preset time range, taking all behavior records of the user u between the latest one-time behavior record and the last behavior record with the same type as the latest one-time behavior record as the preposed state of the one-time behavior record of the user u to the merchant j,
wherein based on the set of leading states QujCompute a set of pre-states Q for user u for merchant j that belong to user u in all behavior records at merchant iujTotal number of times of behavior records CuijI.e. the behavior of user u at merchant i is recorded in the set of leading states QujWherein u, i, j are all positive integers smaller than a predetermined threshold, and corresponding C is calculated for all combinations of u, i, j valuesuij，
Wherein based on the calculated CuijCalculating each pre-state reference value S as followsuij：
Suij＝log(Cuij+e-1)，
Wherein the step (A2) further comprises: based on calculated SuijThe transition probability P for each merchant pair consisting of a source merchant and a destination merchant is calculated as followsij：
Where U is a set of users, and n (i) and n (j) represent the total number of times merchant i and merchant j appear in the pre-state of all users, respectively.
2. The merchant recommendation method based on user behavior characteristic analysis according to claim 1, wherein said step (a3) further comprises: based on calculated PijCalculate per merchant for each user u as followsCurrent score Puj：
Where N (u) is the preposition state of user u, S (j, K) is the set of K merchants with the highest transition probability for merchant j, and W is the number of merchants with the highest transition probability for merchant juiIs the weight.
3. The merchant recommendation method based on user behavior characteristic analysis according to claim 2, wherein said step (a3) further comprises: calculate W as followsui：
wui＝log(Cui+e-1)
Wherein, CuiIs the number of times merchant i appears in the pre-state recorded for user u's last activity.
4. The merchant recommendation method based on user behavior characteristic analysis according to claim 3, wherein the merchant recommendation list includes only the top N merchants in the ranking result, where N is a positive integer.
5. The merchant recommendation method based on user behavior characteristic analysis according to claim 4, wherein said step (A3) further comprises: and performing secondary optimization operation on the sorting result after sorting all the merchants in the order from big to small based on the current score of each merchant.
6. The merchant recommendation method based on user behavior characteristic analysis of claim 5, wherein the secondary optimization operation comprises: and judging whether the merchants need to directly promote the ranking or the weighting according to a preset rule aiming at each sequenced merchant, and if the merchants need to directly promote the ranking or the weighting, directly promoting the ranking or the weighting, thereby obtaining a final merchant recommendation list.
7. The merchant recommendation method based on user behavior characteristic analysis according to claim 6, wherein said step (A3) further comprises: and after the final merchant recommendation list is obtained, sending the final merchant recommendation list to a user terminal so as to display the final merchant recommendation list to the user terminal and recommend merchants.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611243256.0A CN106980989B (en) | 2016-12-29 | 2016-12-29 | Merchant recommendation method based on user behavior characteristic analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611243256.0A CN106980989B (en) | 2016-12-29 | 2016-12-29 | Merchant recommendation method based on user behavior characteristic analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106980989A CN106980989A (en) | 2017-07-25 |
| CN106980989B true CN106980989B (en) | 2021-01-19 |
Family
ID=59340300
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611243256.0A Active CN106980989B (en) | 2016-12-29 | 2016-12-29 | Merchant recommendation method based on user behavior characteristic analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106980989B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107833076A (en) * | 2017-11-16 | 2018-03-23 | 中国银联股份有限公司 | A kind of marketing message method for pushing and device |
| CN108573399B (en) * | 2018-02-28 | 2022-03-18 | 中国银联股份有限公司 | Merchant recommendation method and system based on transition probability network |
| CN118429020A (en) * | 2024-05-16 | 2024-08-02 | 深圳高灯云科技有限公司 | Merchant recommendation method, merchant recommendation device, merchant recommendation computer device, merchant recommendation storage medium and merchant recommendation program product |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101826114A (en) * | 2010-05-26 | 2010-09-08 | 南京大学 | Multi Markov chain-based content recommendation method |
| KR20140031647A (en) * | 2012-09-05 | 2014-03-13 | 에스케이플래닛 주식회사 | Item recommend system and method thereof, apparatus supporting the same |
| CN105677831A (en) * | 2016-01-04 | 2016-06-15 | 拉扎斯网络科技（上海）有限公司 | Method and device for determining recommended merchants |
| CN105740444A (en) * | 2016-02-02 | 2016-07-06 | 桂林电子科技大学 | User score-based project recommendation method |
| CN105809478A (en) * | 2016-03-07 | 2016-07-27 | 合网络技术（北京）有限公司 | Advertisement label marking method and system |
-
2016
- 2016-12-29 CN CN201611243256.0A patent/CN106980989B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101826114A (en) * | 2010-05-26 | 2010-09-08 | 南京大学 | Multi Markov chain-based content recommendation method |
| KR20140031647A (en) * | 2012-09-05 | 2014-03-13 | 에스케이플래닛 주식회사 | Item recommend system and method thereof, apparatus supporting the same |
| CN105677831A (en) * | 2016-01-04 | 2016-06-15 | 拉扎斯网络科技（上海）有限公司 | Method and device for determining recommended merchants |
| CN105740444A (en) * | 2016-02-02 | 2016-07-06 | 桂林电子科技大学 | User score-based project recommendation method |
| CN105809478A (en) * | 2016-03-07 | 2016-07-27 | 合网络技术（北京）有限公司 | Advertisement label marking method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106980989A (en) | 2017-07-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11659050B2 (en) | Discovering signature of electronic social networks | |
| CN104899267B (en) | A kind of integrated data method for digging of social network sites account similarity | |
| CN103678672B (en) | Method for recommending information | |
| CN104281882B (en) | The method and system of prediction social network information stream row degree based on user characteristics | |
| WO2020073534A1 (en) | Pushing method and apparatus based on re-clustering, and computer device and storage medium | |
| US10348550B2 (en) | Method and system for processing network media information | |
| CN105069172B (en) | Interest tags generation method | |
| CN107622072B (en) | Identification method for webpage operation behavior, server and terminal | |
| US20150347591A1 (en) | Information matching and match validation | |
| CN103795612A (en) | Method for detecting junk and illegal messages in instant messaging | |
| CN103077220A (en) | User group correlation degree-based personalized recommendation method and system | |
| CN105069072A (en) | Emotional analysis based mixed user scoring information recommendation method and apparatus | |
| WO2020257991A1 (en) | User identification method and related product | |
| CN103337028B (en) | A kind of recommendation method, device | |
| CN107423335B (en) | Negative sample selection method for single-class collaborative filtering problem | |
| CN106980989B (en) | Merchant recommendation method based on user behavior characteristic analysis | |
| CN105023178B (en) | A kind of electronic commerce recommending method based on ontology | |
| CN105653640A (en) | Collaborative filtering recommendation method based on trust mechanism | |
| WO2022142903A1 (en) | Identity recognition method and apparatus, electronic device, and related product | |
| CN103825784A (en) | Non-public protocol field identification method and system | |
| CN115730152A (en) | Big data processing method and big data processing system based on user portrait analysis | |
| CN114168761A (en) | Multimedia data push method, device, electronic device and storage medium | |
| CN110598126A (en) | A cross-social network user identification method based on behavioral habits | |
| CN110457387A (en) | A kind of method and relevant apparatus determining applied to user tag in network | |
| CN108537569A (en) | The advertisement sending method that interpersonal relationships perceives in online social networks |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |