Invention content
Embodiment of the present invention is designed to provide a kind of electric business Method of Commodity Recommendation and electric business Platform Server, passes throughThere are associated associated articles with order goods in database, and carry out corresponding commodity according to the incidence coefficient of associated articles and push awayIt recommends, is conducive to provide with targetedly commercial product recommending for client, improves the shopping experience of user, while being also beneficial to increase electricityThe sales volume of quotient's platform.
In order to solve the above technical problems, embodiments of the present invention provide a kind of electric business Method of Commodity Recommendation, including：When detecting single under user, the order goods in order is obtained；According to order goods, obtains in database and exist with order goodsAssociated associated articles, and obtain the incidence coefficient of each associated articles；Wherein, incidence coefficient is at associated articles and order commodityIn the probability of same order；According to the size of incidence coefficient, at least one associated articles are filtered out in the associated articles of acquisition,As pre- Recommendations；The merchandise news of pre- Recommendations is pushed to the terminal interface of user.
Embodiments of the present invention additionally provide a kind of electric business Platform Server, including：First acquisition module, screening mouldBlock, recommending module；First acquisition module is used to when single under detecting user, obtain the order goods in order；First obtainsModule is additionally operable to according to order goods, obtains in database that there are associated associated articles with order goods, and obtains each associationThe incidence coefficient of commodity；Wherein, incidence coefficient is the probability that associated articles are in same order with order commodity；Screening module withFirst acquisition module connects, and for the size according to incidence coefficient, at least one association is filtered out in the associated articles of acquisitionCommodity, as pre- Recommendations；Recommending module is connect with screening module, for the merchandise news of pre- Recommendations to be pushed to useThe terminal interface at family.
Embodiment of the present invention in terms of existing technologies, when single under user, obtains the order quotient in orderProduct, inquire in database that there are associated associated articles with the order goods, and according to the size of incidence coefficient, filter out associationBig commodity are spent as pre- Recommendations, are recommended to the terminal interface of user.It is in same by associated articles and order commodityThe incidence coefficient that the probability of one order obtains can reflect the pass that there is collocation using effect with order commodity to a certain extentJoin commodity, these associated articles are recommended to the terminal of user, is conducive to provide with targetedly commercial product recommending for client, carryThe shopping experience of high user is also beneficial to increase the sales volume of electric business platform in addition while user buys associated articles.
In addition, according to the size of incidence coefficient, at least one associated articles are filtered out in the associated articles of acquisition, specificallyIncluding：Priority ranking is carried out to the associated articles of acquisition, wherein the higher associated articles of incidence coefficient, priority are higher；It willN number of associated articles of highest priority, as the associated articles filtered out；Wherein, N is greater than or equal to 1；By pre- RecommendationsMerchandise news pushes to the terminal interface of the user, specifically includes：According to the priority of pre- Recommendations from high to lowSequentially, the merchandise news of pre- Recommendations is pushed to the terminal interface of user successively.By the commodity to filtering out according to passThe height for contacting number carries out priority ranking, and the high commodity of priority are pushed to the terminal of user, are conducive to provide to the userThe higher commodity of incidence coefficient so that the commodity specific aim of recommendation is stronger, is conducive to the shopping experience for further increasing user.
In addition, when the associated articles to acquisition carry out priority ranking, further include：If there are the identical associations of priorityCommodity, then according to the commodity classification of the identical associated articles of priority, the similarity with the commodity classification of order goods, to preferentialThe identical associated articles of grade carry out two minor sorts of priority.When identical there are the priority of multiple associated articles, that is, it is associated withWhen coefficient is identical, the second minor sort is carried out according to the commodity classification of this multiple commodity, selects and belongs to same point with order goodsThe associated articles of class are conducive to further increase more targetedly commercial product recommending, improve the shopping experience of user.
In addition, before the merchandise news of pre- Recommendations is pushed to the terminal interface of user, further include：Obtain user'sIdentity information；Judge whether user is the user first logged into；When judgement user is non-first log into when, the history for obtaining user is orderedSingle information and/or information browsing；According to the History Order information and/or information browsing of user, selects and meet preset requirementPre- Recommendations；The merchandise news of pre- Recommendations is pushed to the terminal interface of user, specially：The satisfaction selected is pre-If it is required that the merchandise news of pre- Recommendations push to the terminal interface of user.User is determined by detecting the identity of userWhether it is old user, when being determined as old user, comprehensive push away is carried out according to the History Order information of user and/or reading recordIt recommends, is conducive to the shopping experience for further increasing user.
In addition, incidence coefficient, which is specially the associated articles regularly updated, is in the probability of same order with order commodity.It closesContacting number has timeliness, is regularly updated, updated incidence coefficient accuracy higher, by improving incidence coefficientAccuracy, be conducive to provide to the user more accurate specific aim and recommend, be conducive to the shopping experience for further increasing user.
In addition, before the merchandise news of pre- Recommendations is pushed to the terminal interface of user, further include：Obtain pre- recommendThe history evaluation of commodity, and according to the history evaluation of acquisition, pre- Recommendations are filtered；The commodity of pre- Recommendations are believedBreath pushes to the terminal interface of user, specially：The merchandise news of pre- Recommendations after filtering is pushed to the end of userHold interface.According to the history evaluation of pre- Recommendations, pre- Recommendations are once filtered, filter out the underproof quotient of evaluationProduct provide good commercial product recommending to the user, are conducive to the shopping experience for further increasing user, meanwhile, it is also beneficial to enhanceThe service quality of electric business platform.
Specific implementation mode
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with attached drawing to the present inventionEach embodiment be explained in detail.However, it will be understood by those skilled in the art that in each embodiment party of the present inventionIn formula, in order to make the reader understand this application better, many technical details are proposed.But even if without these technical detailsAnd various changes and modifications based on the following respective embodiments, it can also realize the application technical solution claimed.
The first embodiment of the present invention is related to a kind of electric business Method of Commodity Recommendation.The core of present embodiment is thisElectric business Method of Commodity Recommendation is applied to the server of electric business platform, including：When single under detecting user, obtain in orderOrder goods；It according to order goods, obtains in database that there are associated associated articles with order goods, and obtains each association quotientThe incidence coefficient of product；Wherein, incidence coefficient is the probability that associated articles are in same order with order commodity；According to incidence coefficientSize, at least one associated articles are filtered out in the associated articles of acquisition, as pre- Recommendations；By pre- RecommendationsMerchandise news pushes to the terminal interface of user.By the way that there are associated associated articles with order goods in database, and according toThe incidence coefficient of associated articles carries out corresponding commercial product recommending, is conducive to provide with targetedly commercial product recommending for client, carryThe shopping experience of high user, while being also beneficial to increase the sales volume of electric business platform.The electric business commodity of present embodiment are pushed away belowThe realization details for recommending method is specifically described, and the following contents only for convenience of the realization details provided is understood, not implements thisScheme it is necessary.
A kind of detailed process of electric business Method of Commodity Recommendation in present embodiment is as shown in Figure 1, include：
Step 101：When single under detecting user, the order goods in order is obtained.
Specifically, the server of electric business platform is obtained in the user for detecting log-on webpage when lower single in its orderAll order goods.Such as user's first logs in Taobao's page, when user's first is when lower single, then server obtains use automaticallyOrder goods in the order of family first.
Step 102：It according to order goods, obtains in database that there are associated associated articles with order goods, and obtainsThe incidence coefficient of each associated articles.
Specifically, automatic to obtain in database after the server of electric business platform gets the order goods in orderThere are associated commodity with order goods, and obtain its incidence coefficient.It is mentioned here that there are associated associated articles, Ke YishiIncidence coefficient is more than the commodity of certain threshold value, and can also be the associated articles set in advance can adjust automatically in practical applicationsIt is whole.Incidence coefficient is the probability that associated articles are in same order with order commodity, and probability here refers to being closed in a period of timeConnection commodity and order commodity are in the probability of same order, and such as the order of purchase mobile phone shares 100 lists in one week in the past, this 10080 lists all include earphone in list, it is determined that within past one week, the incidence coefficient of mobile phone and earphone is 80%, in practical applicationIn, incidence coefficient here is that the associated articles regularly updated are in the probability of same order with order commodity.Incidence coefficient hasTime-effectiveness is regularly updated, updated incidence coefficient accuracy higher, by improving the accurate of incidence coefficientProperty, be conducive to provide more accurate specific aim recommendation to the user, be conducive to the shopping experience for further increasing user.
Step 103：According to the size of incidence coefficient, at least one associated articles are filtered out in the associated articles of acquisition,As pre- Recommendations.
Specifically, according to the size of incidence coefficient, at least one associated articles are filtered out in the associated articles of acquisition,It specifically includes：Priority ranking is carried out to the associated articles of acquisition, wherein the higher associated articles of incidence coefficient, priority are got overIt is high；By N number of associated articles of highest priority, as the associated articles filtered out；Wherein, N is greater than or equal to 1.According to associationThe size of coefficient determines the priority of associated articles, according to priority height, top n associated articles is filtered out, by this N number of passJoin commodity as pre- Recommendations.
It is noted that when associated articles identical there are multiple priority, according to the identical association quotient of priorityThe commodity classification of product, the similarity with the commodity classification of order goods carry out priority to the identical associated articles of priorityTwo minor sorts.When identical there are the priority of multiple associated articles, i.e., when incidence coefficient is identical, according to the quotient of this multiple commodityProduct classification carries out the second minor sort, selects the associated articles for belonging to same classification with order goods, is conducive to further increaseSpecific aim commercial product recommending improves the shopping experience of user.For convenience of description, a commodity classification situation is enumerated herein, such as 1 institute of tableShow, in practical applications, commodity classification is adjusted according to actual demand, is not construed as limiting herein.Such as, when the order goods of userFor mobile phone, and the incidence coefficient of earphone, rice and mobile phone was all 80% within past one week, then because of the commodity of earphone and mobile phoneClassify similar, the commodity classification of " mobile phone/number/computer office " is belonged to, so preferential recommendation earphone.
Table 1
In addition, under practical situations, when there are abnormal conditions, the server of electric business platform can be different with automatic fitrationNormal associated articles.If current electric business platform is in promotion period, then the association of commodity sales promotion and order goods whithin a period of timeCoefficient can reach a very high value, and in this case, the server of electric business platform can be fallen and be ordered the goods with automatic fitrationThe associated articles in same commodity field are not belonging to, the associated articles left after filtering are subjected to priority row according to field similitudeSequence, to carry out commercial product recommending.Herein, a kind of citing when only occurring to abnormal conditions in practical applications can be according to demandIt adjusts accordingly, is not restricted herein.
Step 104：The merchandise news of pre- Recommendations is pushed to the terminal interface of user.
Specifically, it is according to the sequence from high to low of the priority of pre- Recommendations, successively by pre- RecommendationsMerchandise news pushes to the terminal interface of user.Here terminal interface can be commodity details interface, acknowledgement of orders interface orPerson's payment interface etc..The presentation mode of Recommendations can be connected on the lower section of current browse page, can also be with suspended window for oneClickthrough, it might even be possible to corner is presented on by the window of small scale, specific presentation mode is not intended to be limited in any herein.
Embodiment of the present invention in terms of existing technologies, when single under user, obtains the order quotient in orderProduct, inquire in database that there are associated associated articles with the order goods, and according to the size of incidence coefficient, filter out associationBig commodity are spent as pre- Recommendations, are recommended to the terminal interface of user.It is in same by associated articles and order commodityThe incidence coefficient that the probability of one order obtains can reflect the pass that there is collocation using effect with order commodity to a certain extentJoin commodity, these associated articles are recommended to the terminal of user, is conducive to provide with targetedly commercial product recommending for client, carryThe shopping experience of high user is also beneficial to increase the sales volume of electric business platform in addition while user buys associated articles.Pass throughPriority ranking is carried out according to the height of incidence coefficient to the commodity filtered out, the high commodity of priority are pushed to the end of userEnd, is conducive to provide the higher commodity of incidence coefficient to the user so that the commodity specific aim of recommendation is stronger, is conducive to further carryThe shopping experience of high user.Meanwhile when identical there are the priority of multiple associated articles, i.e., when incidence coefficient is identical, according toThe commodity classification of this multiple commodity carries out the second minor sort, selects the associated articles for belonging to same classification with order goods, hasConducive to specific aim commercial product recommending is further increased, the shopping experience of user is improved.
Second embodiment of the present invention is related to a kind of electric business Method of Commodity Recommendation.Second embodiment is in the first embodiment partyIt is further improved on the basis of formula, mainly thes improvement is that：In second embodiment of the invention, quotient will recommended in advanceBefore the merchandise news of product pushes to the terminal interface of user, further include：Obtain the identity information of user；Judge user whether headed byThe user of secondary login；When judgement user is non-first log into when, obtain the History Order information and/or information browsing of user；According toThe History Order information and/or information browsing of user selects the pre- Recommendations for meeting preset requirement；By pre- RecommendationsMerchandise news pushes to the terminal interface of user, specially believes the commodity for the pre- Recommendations for meeting preset requirement selectedBreath pushes to the terminal interface of user.Identity by detecting user determines whether user is old user, when being determined as using alwaysWhen family, comprehensive recommendation is carried out according to the History Order information of user and/or reading record, is conducive to further increase user'sShopping experience.
A kind of electric business Method of Commodity Recommendation flow in present embodiment is as shown in Fig. 2, specifically include：
Step 201：When single under detecting user, the order goods in order is obtained.
Step 202：It according to order goods, obtains in database that there are associated associated articles with order goods, and obtainsThe incidence coefficient of each associated articles.
Step 203：According to the size of incidence coefficient, at least one associated articles are filtered out in the associated articles of acquisition,As pre- Recommendations.
Step 204：Obtain the identity information of user.
Specifically, user, usually can be there are one register account number when single under electric business platform is done shopping, will be with the noteThe information of volume account binding, is referred to as the identity information of user.
Step 205：Judge whether user is the user first logged into, if it is not, entering step 206；Otherwise, it enters step208。
Specifically, the server of electric business platform can carry out the user information in the identity information and database of userInquiry is compared, whether the identity had login or lower unirecord before judgement, was first logged into if so, then judging that the user is non-User, as old user, enter step 206, if not having, it is determined that the user is new user, enters step 208.
Step 206：Obtain the History Order information and/or information browsing of user.
Specifically, when judgement log in user for old user when, then obtain its passing History Order information and shouldThe information browsing of user.If user's second is lower in the past twice singly, server obtains this merchandise news in order twice, ifUser's second has browsed the relevant information of tablet computer recently, then obtains the browsing record of the user and the commodity letter of dependent merchandiseBreath.
Step 207：According to the History Order information and/or information browsing of user, selects and meet the pre- of preset requirement and push awayRecommend commodity.
Specifically, according to History Order information, information browsing and the preference information of user, combined recommendation is carried out.HereDescribed preset requirement can be user oneself setting, can also be electric business platform according to the History Order information of user and/Or information browsing automatically generates, and in practical applications, is adjusted accordingly according to actual conditions, is not construed as limiting herein.Such asIt says, user third has bought the mobile phone of rose golden before, and recently in browsing earphone, then the server of electric business platform can be according to these lettersBreath, recommends the earphone of rose golden.
Step 208：The merchandise news of pre- Recommendations is pushed to the terminal interface of user.
Specifically, the merchandise news of pre- Recommendations pushes to the terminal interface of user, will specially selectThe merchandise news for meeting the pre- Recommendations of preset requirement pushes to the terminal interface of user.
Due to step 201-203 in present embodiment, 208 roughly the same with step 101-104 in first embodiment, it isIt avoids repeating, details are not described herein again.
The merchandise news of pre- Recommendations in terms of existing technologies, is being pushed to user's by embodiment of the present inventionBefore terminal interface, the identity of login user is first detected, when by determining that the login user is old user, according to the history of userOrder information and/or reading record carry out comprehensive recommendation, are conducive to the shopping experience for further increasing user.
Third embodiment of the present invention is related to a kind of electric business Method of Commodity Recommendation.Third embodiment is in the first embodiment partyIt is further improved on the basis of formula, mainly thes improvement is that：The merchandise news of pre- Recommendations is being pushed into userTerminal interface before, further include：The history evaluation of pre- Recommendations is obtained, and according to the history evaluation of acquisition, to recommending quotient in advanceProduct are filtered；The merchandise news of pre- Recommendations is pushed to the terminal interface of user, specially：After filtering pre- is pushed awayThe merchandise news for recommending commodity pushes to the terminal interface of user.According to the history evaluation of pre- Recommendations, to pre- Recommendations intoThe primary filtering of row, filters out the underproof commodity of evaluation, provides good commercial product recommending to the user, be conducive to further increase useThe shopping experience at family, meanwhile, it is also beneficial to the service quality of enhancing electric business platform.
A kind of electric business Method of Commodity Recommendation flow in present embodiment is as shown in figure 3, specifically include：
Step 301：When single under detecting user, the order goods in order is obtained.
Step 302：It according to order goods, obtains in database that there are associated associated articles with order goods, and obtainsThe incidence coefficient of each associated articles.
Step 303：According to the size of incidence coefficient, at least one associated articles are filtered out in the associated articles of acquisition,As pre- Recommendations.
Step 304：Obtain the history evaluation of pre- Recommendations.
Specifically, the history evaluation that the other users in pre- Recommendations are carried out from after buying is obtained.The history evaluationInclude not only word evaluation, further include picture evaluation and star grade, marking etc., is not limited in any way herein.
Step 305：According to the history evaluation of acquisition, pre- Recommendations are filtered.
Such as be filtered according to the star grade in history evaluation, to be not up to the pre- Recommendations of three stars intoRow filtering is rejected.Herein for convenience of description, processing is filtered with star grade, in practical applications, can be adjusted flexibly,It is not intended to be limited in any in the present embodiment.
Step 306：The merchandise news of pre- Recommendations is pushed to the terminal interface of user.
Specifically, the merchandise news of pre- Recommendations is pushed to the terminal interface of user, it specially will after filteringThe merchandise news of pre- Recommendations push to the terminal interface of user.
Due to step 301-303 in present embodiment, 306 roughly the same with step 101-104 in first embodiment, it isIt avoids repeating, details are not described herein again.
The merchandise news of pre- Recommendations in terms of existing technologies, is being pushed to user's by embodiment of the present inventionBefore terminal interface, according to the history evaluation of pre- Recommendations, pre- Recommendations are once filtered, evaluation is filtered out and does not conform toThe commodity of lattice provide good commercial product recommending to the user, are conducive to the shopping experience for further increasing user, meanwhile, also favorablyIn the service quality of enhancing electric business platform.
The step of various methods divide above, be intended merely to describe it is clear, when realization can be merged into a step orCertain steps are split, multiple steps are decomposed into, as long as including identical logical relation, all in the protection domain of this patentIt is interior；To either adding inessential modification in algorithm in flow or introducing inessential design, but its algorithm is not changedCore design with flow is all in the protection domain of the patent.
Four embodiment of the invention is related to a kind of electric business Platform Server, as shown in figure 4, including：First acquisition module401, screening module 402, recommending module 403；First acquisition module 401 is used to when single under detecting user, obtain in orderOrder goods；First acquisition module 401 is additionally operable to according to order goods, obtains in database that there are associated with order goodsAssociated articles, and obtain the incidence coefficient of each associated articles；Wherein, incidence coefficient is that associated articles are in same with order commodityThe probability of order；Screening module 402 is connect with the first acquisition module 401, for the size according to incidence coefficient, in the pass of acquisitionAt least one associated articles are filtered out in connection commodity, as pre- Recommendations；Recommending module 403 is connect with screening module 402, is usedIn the terminal interface that the merchandise news of pre- Recommendations is pushed to user.
Wherein, screening module 402 is specifically used for carrying out priority ranking to the associated articles of acquisition, wherein incidence coefficientHigher associated articles, priority are higher；By N number of associated articles of highest priority, as the associated articles filtered out；ItsIn, N is greater than or equal to 1；Recommending module 403 is specifically used for the sequence from high to low of the priority according to pre- Recommendations, according toThe secondary merchandise news by pre- Recommendations pushes to the terminal interface of user.
It is not difficult to find that present embodiment is device embodiment corresponding with first embodiment, present embodiment can be withFirst embodiment is worked in coordination implementation.The relevant technical details mentioned in first embodiment still have in the present embodimentEffect, in order to reduce repetition, which is not described herein again.Correspondingly, the relevant technical details mentioned in present embodiment are also applicable inIn first embodiment.
Fifth embodiment of the invention is related to a kind of electric business Platform Server, and structural representation is as shown in Figure 5.5th embodiment partyFormula is being further improved in the 4th embodiment, is mainly theed improvement is that：In the 5th embodiment, this electric business commodityRecommendation apparatus further includes：Second acquisition module 501, judgment module 502 and selecting module 503；Second acquisition module 501 and screeningModule 402 connects, the identity information for obtaining the user；Judgment module 502 is connect with the second acquisition module 501, is used forJudge whether user is the user first logged into；When judgement user is non-first log into when, obtain user History Order information and/Or information browsing；Selecting module 503 is connect with judgment module 502, for being believed according to the History Order information and/or reading of userBreath, selects the pre- Recommendations for meeting preset requirement；Recommending module 403, which is specifically used for presetting the satisfaction selected, to be wantedThe merchandise news for the pre- Recommendations asked pushes to the terminal interface of the user.
Since second embodiment is corresponded with present embodiment, present embodiment can be mutual with second embodimentMatch implementation.The relevant technical details mentioned in second embodiment are still effective in the present embodiment, implement secondThe attainable technique effect of institute similarly may be implemented in the present embodiment in mode, no longer superfluous here in order to reduce repetitionIt states.Correspondingly, the relevant technical details mentioned in present embodiment are also applicable in second embodiment.
It is noted that each module involved in the 4th embodiment and the 5th embodiment is logic mouldBlock, in practical applications, a logic unit can be a physical units, can also be a part for a physical unit,It can also be realized with the combination of multiple physical units.In addition, in order to protrude the innovative part of the present invention, in present embodiment notHave and introduce the unit less close with technical problem relationship proposed by the invention is solved, but this does not indicate present embodimentIn there is no other units.
It will be understood by those skilled in the art that it is that can pass through to implement the method for the above embodimentsProgram is completed to instruct relevant hardware, which is stored in a storage medium, including some instructions are used so that oneA equipment (can be microcontroller, chip etc.) or processor (processor) execute each embodiment the method for the applicationAll or part of step.And storage medium above-mentioned includes：USB flash disk, mobile hard disk, read-only memory (ROM, Read-OnlyMemory), random access memory (RAM, Random Access Memory), magnetic disc or CD etc. are various can store journeyThe medium of sequence code.
It will be understood by those skilled in the art that the respective embodiments described above are to realize specific embodiments of the present invention,And in practical applications, can to it, various changes can be made in the form and details, without departing from the spirit and scope of the present invention.