CN102959539A

CN102959539A - Method and system for item recommendation in service crossing situation

Info

Publication number: CN102959539A
Application number: CN2011800010578A
Authority: CN
Inventors: 杜家春; 汪芳山; 钟杰萍
Original assignee: Huawei Technologies Co Ltd
Current assignee: Huawei Technologies Co Ltd
Priority date: 2011-06-29
Filing date: 2011-06-29
Publication date: 2013-03-06
Anticipated expiration: 2031-06-29
Also published as: WO2012159308A1; CN102959539B

Abstract

A method and system for item recommendation in service crossing situation of digital media or electronic commerce, the method includes: acquiring a service identifier, which a target user is using, of digital media or electronic commerce and a target user identifier, via a computer network interface, and acquiring service source data stored in advance from a memory according to the service identifier; creating a set of candidate recommended items for the target user; acquiring a predicted rating of each candidate recommended item in the set of the candidate recommended items; extracting the qualified candidate recommended items from the candidate recommended items to create a final list of recommended items for the target user, according to the predicted rating of the items to be recommended; sending the final list of recommended items to the client of the target user to display. The method and system can reduce the processing time of item recommendation and therefore improve the efficiency of item recommendation.

Description

Item recommendation method and system during a kind of repeat in work

Technical field

The present invention relates to communication and Internet technical field, more particularly to item recommendation method and system when a kind of Digital Media or electronic commerce affair intersection.

Background technology

With the popularization and fast development of Internet technology, substantial amounts of data message is occurred in that on internet, and user only wants to see oneself data message that is interested or needing when webpage is browsed, but often there are a lot of other unrelated data messages simultaneously in actual scene, although it is more and more that this results in data message, but the utilization rate of data message increasingly lower phenomenon, referred to as information overload phenomenon.

In order to prevent or reduce the influence that information overload phenomenon is come to the interacting strip of user and internet as far as possible, data message can be analyzed and processed before data message is shown to user, such as personalized recommendation technology is to reduce a kind of method of information overload phenomenon by recommending to meet the resource of its interest or demand for user.Personalized recommendation technology has been widely applied in the multiple fields such as ecommerce, digital library, music, video and news at present, this multiple application field also includes multiple business, such as telecom operators are proposed Ring Back Tone service, ring service and complete bent business in music field, for telecom operators, the project in each business（Herein refer to music）Some overlaps again, and the user of each further business may also have coincidence.Such as e-commerce website again, each seller can regard a business as, and the commodity of seller are probably coincidence, and the client of seller is also likely to be coincidence.The business item of multiple business and/or the phenomenon that partially overlaps of user, are repeat in work phenomenon in this same application field.

One traditional technology of personalized recommendation is collaborative filtering, and collaborative filtering carries out personalized recommendation based on user-project score data, and user-project score data can be obtained explicitly and/or implicitly.The openness of data can influence final recommendation effect.In the application field of personalized recommendation technology, user-project score data represents fancy grade of the user for some project in business, and the score value of user-project score data is bigger to represent that user more likes this project.User-project score data can be obtained explicitly or implicitly, and for example user directly carries out scoring operation to project.But in most of application fields, user may only in bulk items set it is blunt it is few with a portion of scoring, therefore user-project score data just has the phenomenon much lacked, and user-project score data also can be than sparse.

In the prior art, it is the sparse sex chromosome mosaicism of solution user-project score data, it is the score data for supplying user to non-scoring item to have a kind of method；For example user can be set to comment to the scoring of non-scoring item Divide the median of scope, or the scoring of non-scoring item is directly set to the average score of user；But the score data of this default setting has certain subjectivity, is generally differed greatly with actual user-project score data；There is a method in which hair is to predict scoring of the user to non-scoring item using some proposed algorithms；But this method based on prediction scoring cannot guarantee that the authentic and valid of score data because being the Rating Model obtained based on former sparse data.

Further, when the above method of application prior art, because being required for dynamic access user-project score data when carrying out project recommendation every time, recommended further according to user-project score data, so also causing the less efficient of project recommendation；Also, because the sparse sex chromosome mosaicism of user-project score data can not be solved well, and the quality of data is also not high enough, and the data message that need not be shown to user originally can be caused to be shown to user, reduces the validity and accuracy of project recommendation.The content of the invention

Item recommendation method and system when offer Digital Media of the embodiment of the present invention or electronic commerce affair intersection, to combine the sparse sex chromosome mosaicism that practical application area solves user-project score data in the prior art, processing time during project recommendation is reduced, so that the efficiency that Improving Project is recommended.

In order to solve the above technical problems, a kind of item recommendation method when intersecting the embodiments of the invention provide Digital Media or electronic commerce affair, this method includes：

Digital Media or electronic commerce affair mark and the targeted customer's mark that targeted customer is using are obtained by computer network interface, and obtain the business source data prestored from memory according to the service identification；

Identified according to the targeted customer, the Digital Media that targeted customer is using or electronic commerce affair are identified and business source data, be targeted customer's generation candidate recommendation project set；

The prediction scoring of each candidate recommendation project in the candidate recommendation project set is obtained according at least to user's similarity in the business source data and/or item similarity；

The consequently recommended bulleted list that qualified candidate recommendation project generates the targeted customer is extracted from the candidate recommendation project according to the prediction scoring of the candidate recommendation project；

Digital Media or electronic commerce affair server, which send the final project recommendation list to the client of the targeted customer, to be shown.

A kind of project recommendation system when intersecting the embodiments of the invention provide Digital Media or electronic commerce affair System, the system includes：

Mark unit is obtained, is identified for the Digital Media or electronic commerce affair mark that are being used by computer network interface acquisition targeted customer and targeted customer；

Acquisition business source data units, for obtaining the business source data prestored from memory according to the service identification；

Candidate collection unit is generated, for being identified according to the targeted customer, the Digital Media that targeted customer is using or electronic commerce affair are identified and business source data, are targeted customer's generation candidate recommendation project set；

Prediction scoring unit is obtained, for according at least to the user's similarity and/or item similarity in the business source data, obtaining the prediction scoring of each candidate recommendation project in the candidate recommendation project set；

Final list cell is generated, the consequently recommended bulleted list that qualified candidate recommendation project generates the targeted customer is extracted from the candidate recommendation project for the prediction scoring according to the candidate recommendation project；

Display unit, is shown for the final project recommendation list to be sent to the client of the targeted customer.

The embodiment of the present invention has advantages below：User's similarity and item similarity, directly can provide data for project recommendation, can thus reduce processing time during project recommendation, so that the efficiency that Improving Project is recommended.Further, before user's similarity and item similarity is prestored, by user-project score data after the mapping of the 4 blunt good selection business, the validity and accuracy of recommendation results can be improved by the user's similarity and item similarity calculated.Therefore, the embodiment of the present invention can be good at solving the sparse sex chromosome mosaicism of user-project score data, processing time during project recommendation can be reduced, the efficiency recommended so as to Improving Project, and the validity and accuracy of recommendation results on increase line can be carried out by lifting the validity and authenticity of user-project score data.Certainly, implement any one embodiment disclosed by the invention to be not necessarily required to while reaching that above-mentioned institute is effective.Brief description of the drawings In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, the required accompanying drawing used in embodiment or description of the prior art will be briefly described below, apparently, drawings in the following description are only some embodiments of the present invention, for those of ordinary skill in the art, without having to pay creative labor, other accompanying drawings can also be obtained according to these accompanying drawings.

The acquisition flow chart of user's similarity and/or item similarity in item recommendation method when Fig. 1 is Digital Media or the electronic commerce affair intersection of the present invention；

Fig. 2 is the flow chart of step 102 in flow chart shown in Fig. 1；

Fig. 3 is the flow chart of step 202 in flow chart shown in Fig. 2；

Fig. 4 is the flow chart of step 304 in flow chart shown in Fig. 3；

The flow chart of item recommendation method one embodiment when Fig. 5 is Digital Media or the electronic commerce affair intersection of the present invention；

The flow chart of item recommendation method another embodiment when Fig. 6 is Digital Media or the electronic commerce affair intersection of the present invention；

The flow chart of item recommendation method another embodiment when Digital Media or electronic commerce affair that Fig. 7 is intersect；

The structural representation of the acquisition embodiment of user's similarity and/or item similarity when Fig. 8 is Digital Media or the electronic commerce affair intersection of the present invention；

Fig. 9 is the structural representation of integral unit 802 in embodiment shown in Fig. 8；

Figure 10 is the structural representation of the second coupling subelement 902 in integral unit 802 shown in Fig. 9；Figure 11 is the structural representation of business coupling subelement 1004 in the second coupling subelement 902 shown in Figure 10；

The structural representation of item recommendation system when Figure 12 is Digital Media or the electronic commerce affair intersection of the present invention.Embodiment

Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is only a part of embodiment of the invention, rather than whole embodiments.Based on the embodiment in the present invention, the every other embodiment that those of ordinary skill in the art are obtained under the premise of creative work is not made belongs to the scope of protection of the invention. The personalized recommendation technology mentioned in background technology is further introduced first.Most widely used in personalized recommendation technology is collaborative filtering.It is that user carries out personalized recommendation that collaborative filtering, which is based on user-project score data,.It is assumed here that scoring of the user to project is, this scoring can explicitly obtain (user carries out scoring operation to project), can also implicitly obtain（By user to the search of project, browse or/and the behavior such as buy and construct score function and calculate and obtain）.Generally, the score value of user-project scoring all can be to limit integer representation within a certain range, and score value is bigger to represent that user more likes the project.The most frequently used in collaborative filtering is the collaborative filtering based on memory, and the collaborative filtering based on memory includes collaborative filtering and project-based collaborative filtering based on user.The general principle of collaborative filtering based on user is the project that the similitude scored using user project may be interested come mutual recommended user.For example, to active user, system passes through its scoring record and specific similarity function, calculate and its nearest-neighbors collection of the behavior most close individual user as user that score, the project for scoring and not scoring by the neighbour user of statistics generates Candidate Recommendation collection, then the prediction scoring that each project is concentrated to Candidate Recommendation is calculated, and takes wherein prediction scoring W project of highest to recommend to collect as-the W of user.The similitude of project-based collaborative filtering then between item compared, and generate Candidate Recommendation collection according to the similar terms of active user's scoring item.

Item recommendation method during Digital Media or the electronic commerce affair intersection of the present invention mainly (can be understood as background system under line）On line（It can be understood as foreground system）Two parts are constituted, wherein, the function of project recommendation is that part is realized on line, and computing is partly carried out under the service source online data used in item recommendation method and is stored into memory, the source data should at least include user's similarity and/or item similarity, so that inline system can carry out project recommendation according to user's similarity and/or item similarity.Also, the acquisition methods of the similarity and/or item similarity can be stored into memory by computing device, and by implementing result；And user items recommend method to be performed in the specific implementation by the Digital Media or e-commerce server being functionally connected with the processor, the Digital Media or e-commerce server can carry out data interaction by human-computer interaction interface and user, and the acquisition of original source data or business source data can be realized by computer network interface and is shared.

In embodiment part, those skilled in the art are better understood upon the user's similarity stored in memory and/or the acquisition modes of item similarity for convenience, so with reference to Fig. 1, to realize the present invention The acquisition methods flow chart of user's similarity and/or item similarity, specifically may comprise steps of in embodiment：

Step 101:The original source data of a variety of different digital media or electronic commerce affair is obtained, the original source data includes：Initial user-project score data of a variety of Digital Medias or electronic commerce affair.

In actual applications, the original source data should at least include initial user-project score data of multiple business；Optionally, the user attribute data and item attribute data of multiple business, and call detailed list data etc. can also be included.

Wherein, the Digital Media or electronic commerce affair, include but is not limited to：Music, using download, internet book store, electronic reading, game and/or shopping online.User-project score data is score data of the user that is produced during business use of the corresponding user of miscellaneous service to project.Call detailed list is the message registration between a period of time interior user, and the data of contact situation between the similar reflection user such as detailed single, the instant messaging record of short message or E-mail communication record can also be used.

Step 102:According to all users between a variety of Digital Medias or electronic commerce affair and the standardization result of initial user-project score data of the matching result of project and a variety of Digital Medias or electronic commerce affair, initial user-project score data of a variety of Digital Medias or electronic commerce affair is integrated into all users including a variety of Digital Medias or electronic commerce affair and unification user-project score data of project.

Wherein, the original source data can also include：The user attribute data and item attribute data of multiple business, with reference to Fig. 2, the step 102 may include steps of again when implementing：Step 201:According to initial user mark, initial user attribute and the initial user property value in the user attribute data, matching obtains unique actual user between a variety of Digital Medias or electronic commerce affair；The initial user mark represents unique user in a certain business；The initial user property value is used to represent all unique user between the multiple business.

Wherein, attribute of the user attribute data to describe each user in miscellaneous service, attribute of the item attribute data to describe each project in miscellaneous service.For example in music field, user attribute data in each business includes telephone number of user etc., and the item attribute data in each business include music name, singer, school, album name, issuing date, region, languages, duration and/or form etc.. What step 201 was performed is the process matched to all users between a variety of Digital Medias or electronic commerce affair in step 102, and user's matching here refers to determination in these different user attribute datas, and which user is same user.User's matching can be completed according to the matching that associates between the user property value of each business is identified with user, wherein user property value energy unique mark user identity；And be same user by associating matching relationship which user in each business found out.

For example in business A, initial user is designated " lawyer ", and in business B, initial user is designated " Zhang San ", but it is all " 1380000000 " that the two initial users, which identify corresponding contact method, it is that user property value is all " 1380000000 ", then it represents that " lawyer " in business A and " Zhang San " in business B are same person.

Step 202:According to the initial item identification in the item attribute data, initial project attribute and initial project property value, matching obtains unique actual items between a variety of Digital Medias or electronic commerce affair；The initial item identification represents unique project in a certain business.

What the step 202 was performed is the process matched to all items between miscellaneous service in step 102.Project matching refers to determine which project is same project in these different business.

With reference to shown in Fig. 3, the step 202 may include steps of again when implementing：Step 301:Utilize actual items attribute different between a variety of Digital Medias of the initial project attributes match of a variety of Digital Medias or electronic commerce affair or electronic commerce affair.

This step is that the Property Name for the project for being included the item attribute data of various Digital Medias or electronic commerce affair using the name-matches relation of initial project attribute is matched, and obtains the actual items attribute between each Digital Media or electronic commerce affair.The name-matches relation of initial project attribute can be that each field is pre-established, and can also be completed by the mode manually participated in.During Property Name is matched, it is believed that " music name " refers to that same item attribute, or " singer " refer to same item attribute with " singer " with " song title ".

Step 302:The item attribute collection registration and each business and the item attribute collection registration average of other business between a variety of Digital Medias or electronic commerce affair are obtained according to the different actual items attribute.

Assuming that entitled of all different attributes obtained after the name-matches of item attribute, the item attribute collection registration between Ze Liangzhong Digital Medias or electronic commerce affair passes through formula（ 1 ) _{C(S S ) =} \ {Attr_t I Attr_t e Sjn {Attr_t \ Attr_t £ S,} | _i≤ ^_≤ ^

',⁷ I {Attr_t I Attr_t e S,.} U {Attr_t | Attr_tE S rivers, (Shang）Where it is assumed that business one has K kinds, and (^≤i ≠ J≤K) represents the business and the business respectively, represents the item attribute set in the business,^AtRepresent the attribute.So, the item attribute collection registration average of the business and other business passes through formula（2 )

Step 303:The a variety of Digital Medias or electronic commerce affair are ranked up according to the size of the item attribute collection registration average.

The various Digital Medias or electronic commerce affair are ranked up, the item attribute collection registration average of the business after sequence before sequence is big, and the item attribute collection registration average of Digital Media or electronic commerce affair behind sequence is small.

Step 304:According to the sequencing of a variety of Digital Medias or electronic commerce affair after sequence, business matching flow is performed by current business of the first business, the business matching flow includes：The matching entries of the current business and other business are determined, and, delete the current business.

Select the first business in sequence, according to the project label order that project is included in the business, each project being followed successively by the first business determines matching entries in other business, the matching entries are same projects with the project in the first business, after each project of the first business, which is matched, to be completed, the first business is deleted, then starts to match each project in second of business, by that analogy, until all business all match completion.

With reference to shown in Fig. 4, in the step 304 it is determined that the current business and other business matching entries when, can carry out as follows：

Step 401:The initial item identification order included according to current business, first project of selection matches flow as current project project implementation. Specifically in Contemporary Digital media or electronic commerce affair（It is the first business）Including each project in, it is necessary to according to initial item identification order first select first project as current project project implementation match flow；Specifically, the project matching flow can include：

Sub-step 4011：Current project and the project matching degree of each project in other business are calculated using initial project property value；

When project matching degree is calculated, it is possible to use initial project property value uses formula（3) handled： Wherein, ' and respectively represent two projects,^w' (i≤r) be the attribute weight, ^^⁽The property value of the attribute of ') represent project '.When project ' sum the attribute property value all exist and it is equal when, function value be 1, be otherwise 0.

Sub-step 4012:For each other business, suitable project matching degree is chosen according to default threshold condition, to form multiple project matching degree set；

For each other business, retain all project matching degrees not less than predetermined threshold, form multiple project matching degree data sets.Here predetermined threshold is related to each business, and its value, can be different according to practical business between 01.

Sub-step 4013:Matching degree highest project is selected in each project matching degree set as the matching entries of current project；

Sub-step 4014:The matching relationship of the matched project of the current project is recorded, and deletes the matching entries that a variety of Digital Medias or electronic commerce affair include；

Sub-step 4015:Delete the current project；

Sub-step 4016:Whether all items collection for judging the Digital Media or electronic commerce affair is empty, if it is terminates, otherwise performs step 402.

Step 402:Using second project in the first described business as current project, the project matching flow is performed, until the project included in the first described business is sky.

It is current using second project after first project matching of the first business is finished Project, performs the project matching flow, until all matching is finished the project included in the first described business.

Step 305:Business matching flow is performed by the current business of second of business, until the business in the sequence is space-time, all unique actual items between a variety of Digital Medias or electronic commerce affair are obtained according to the matching entries and project matching relationship.

After all items that the first described business packet contains, which are all matched, to be finished, then using second of business as current business, business matching flow is performed, until all items in all Digital Media or electronic commerce affair are all matched and finished.

After all items of all Digital Medias or electronic commerce affair, which are all matched, to be finished, all unique actual items between each Digital Media or electronic commerce affair are determined according to matching entries and project matching relationship.

Step 203:According to the user-project scoring score range and the minimum value of the score range of initial user-project score data of a variety of Digital Medias or electronic commerce affair, a variety of Digital Medias or electronic commerce affair, the standardization result of initial user-project score data of a variety of Digital Medias or electronic commerce affair is obtained.

After actual items are determined, the scoring score range in the user of various Digital Medias or electronic commerce affair-project score data is standardized, and calculates user-project score data of various Digital Medias or electronic commerce affair after standardization.Standardization result calculation formula be： Where it is assumed that business one has K kinds,（ ι≤ ≤ ）Scorings of the user ^ to the standardization result of project in the business after standardization is represented,（ 1≤ ≤ ）Represent original scoring of the user to project, g in the business^e( 1≤ ≤ ）The scoring score range of expression the business, min (_rat_e ^w) ( 1≤ ≤ ）Represent the minimum value of the scoring score range of the business.

Step 204:According to the actual user, actual items and standardization result, integrate described The user of a variety of Digital Medias or electronic commerce affair-project score data, unified user-project the score data of generation, the unified user-project score data includes user-project score data in a variety of Digital Medias or electronic commerce affair after the integration of all users and project.

Result after user based on user's matching, project matching and various Digital Medias or electronic commerce affair-project score data standardization, user-project score data of various Digital Medias or electronic commerce affair is integrated, unified user-project score data is generated.Because in original user-project score data, there is polyisomenism between each user and between each project, so original user-project score data is integrated, user and project are all actual user and actual items in unified user-project score data of generation, therefore, score data only one of which score value of the same user to same project.

In this step, unique scoring of the actual user to actual items can use formula

( 5)、（6) or（7) either method in is obtained：

( 6)

(7) wherein, formula（In 6) " (1≤≤）For user ' to the preference weight of the business, scoring number of times of the user in the business can be set in advance, or user uses Duration of the business of kind etc.；It should be noted that when user items attribute data includes call detailed list, formula can be used（7) calculate, formula（7) ^' in) it is linkman set of the user within a period of time, it can be obtained, can also be obtained by detailed single, the instant messaging record of short message or E-mail communication record etc. by call detailed list.It is the tight ness rating of user and user, user and user in a period of time can be set in advance contacts frequency, or contact duration etc..

Step 103:The unification user-project score data is sequentially mapped to a variety of Digital Medias or electronic commerce affair, user-project score data after a variety of Digital Media or electronic commerce affair mappings is generated.

After unified user-project score data is obtained, need that unified user-project score data is sequentially mapped into various Digital Medias or electronic commerce affair again, to generate user-project score data after various Digital Medias or electronic commerce affair mapping.

For every kind of business, mode one：The corresponding all score datas of all projects that every kind of business is included can be extracted in unified user-project score data, the related user-project score data of the miscellaneous service after mapping is constituted；

Mode two：The corresponding all score datas of total user that every kind of business is included can also be extracted in unified user-project score data, the related user-project score data of the miscellaneous service after mapping is constituted；

Mode three：All projects and the corresponding all score datas of total user that every kind of business is included can also be extracted in unified user-project score data, the related user-project score data of the miscellaneous service after mapping is constituted.

Step 104:According to the item similarity between disparity items in the user's similarity and/or same business between different user in user-same business of project score data acquisition after a variety of Digital Media or electronic commerce affair mappings.

Specifically, the item destination aggregation (mda) that the item destination aggregation (mda) and described two different users that are scored jointly according to two different users in the user after the mapping-project score data, same business each score calculates user's similarity between described two different users；And/or,

The item similarity between described two disparity items is calculated according to the set of the user scored jointly two disparity items in the user after the mapping-project score data, same business and the set of the user each scored two disparity items. It should be noted that, at step 104, user's similarity between different user in same business can be only calculated, user's similarity between disparity items in same business can also be only calculated, similarity can also be all calculated between different user and disparity items in same business.Specifically, formula can be used（8) cosine similarity calculates user's similarity between different user in same business：

Wherein, " represent user ' and user comment undue item destination aggregation (mda) jointly,^Au' represent that user comments undue item destination aggregation (mda).Using formula（9) cosine similarity calculates the item similarity between disparity items in same business： ·^rtj Wherein, ^ is represented to project⁷' and project all comment the set of undue user, represent to project ' comment the set of undue user.

Step 105:User's similarity and/or item similarity are stored into the memory.After user's similarity and/or item similarity is got, since it is desired that being used when carrying out project recommendation, so first user's similarity and/or item similarity can be stored into memory, it is so follow-up to carry out project recommendation to user if desired, user's similarity and/or item similarity required for directly can just being got from memory, business source data is provided this makes it possible to the project recommendation directly for subsequent execution, to reduce the recommendation time of project recommendation, so as to improve the efficiency of project recommendation.

Further, the acquisition methods of user's similarity and/or item similarity disclosed in Fig. 1, because being integrated by the specification of user-project score data and mapping to calculate user's similarity or project Similarity, not only can provide data for project recommendation, the user after mapping that can also be by selecting the business well-project score data and corresponding user's similarity and/or item similarity, can improve the validity accuracy of project recommendation.

Those skilled in the art are better understood from the principle on user's similarity and/or item similarity for convenience, with reference to Fig. 5, an instantiation for obtaining user's similarity and/or item similarity is given, its method realized may comprise steps of：

Step 501:The original source data of a variety of Digital Medias or electronic commerce affair is obtained, the original source data includes：The initial user of a variety of Digital Medias or electronic commerce affair-project score data.

Assuming that there are 3 kinds of business in music field, it is designated as respectively in business S, it is assumed that scoring score range is 1-5, and the user in business S-project score data is as shown in table 1：Table 1

The positional representation user for not having data in table 1 does not comment too for corresponding project, i.e. user items score data is not present.

User attribute data in business S is as shown in table 2：

Table 2

User's ID association phone u,

u₂ u₃

U₅Item attribute data in 137** * * * * * * business are as shown in table 3：Table 3

And in business, scoring score range is 1-10, the user in business-project score data is as shown in table 4：

Table 4

User attribute data in business is as shown in table 5: Table 5

User's ID association phone u,

u₂

u₃

u₄ 137** ***** *

150^^^^^^^^ u₆

Item attribute data in business are as shown in table 6：

Table 6

And in business, scoring score range is 1-5, the user in business-project score data is as shown in table 7：

Table 7

u₂ 5

u₃ 3 5

u₄ 3 4

3 5

5 4

u₇ 5

User attribute data in business is as shown in table 8：

Table 8

Item attribute data in business are as shown in table 9：

Table 9

Project label song title, which enters subflow, sends the perfect grand mandarins one of interaction Wang Li of languages to cut plum to take Yuqin's classics old song mandarin GOOD BYE,AUTUMN Wang Qiang mandarin daphne odera Zhou Jielun mandarin garden party Zhou Jielun mandarin steps 502:According to initial user mark, initial user attribute and the initial user property value in the user attribute data, matching obtains unique actual user between a variety of Digital Medias or electronic commerce affair.

User's matching process in multiple business is carried out first.According to the content of table 2, table 5 and table 8, using telephone number identical user as same user, all user's matching relationship data obtained in 3 kinds of business are as shown in table 10：

Table 10

All unique actual user's mark in each business that first row in table 10 is redistributed after representing to match by user, is also illustrated in the user in unified user-project score data.With the second behavior example explanation user's matching relationship of table 10：The row shows business S user and the user of business is same user, to represent in unified user-project score data.

Step 503:According to the initial item identification in the item attribute data, initial project attribute and initial project property value, matching obtains unique actual items between a variety of Digital Medias or electronic commerce affair. It is Property Name matching first in this step.In this example, item attribute " music name " refers to same item attribute with " song title ", and " singer " is also referred to as same item attribute with " singer ".So by formula（1) and（2) the item attribute collection registration average of the miscellaneous service obtained is as shown in table 11：

Table 11

Assuming that in project matching process, the weight for taking music name, singer, album name, school and languages is respectively 0.5,0.3,0.1,0.05 and 0.05, and take business S and business, business S and business, business and business ＆ project matching degree threshold value be respectively 0.8,0.7 and 0.7 in the case of, the project matching relationship obtained in this example is as shown in table 12：

Table 12

The project label s distributed after matching₂ s₃

I, ' All unique actual items mark between miscellaneous service that first row data in table 12 are redistributed after representing to match by project, is also illustrated in the project in unified user-project score data.With the second behavior example explanation project matching relationship of table 12：The row shows business S project and the project of business^lIt is same project, to represent in unified user-project score data.

Step 504:According to the user-project scoring score range and the minimum value of the score range of initial user-project score data of a variety of Digital Medias or electronic commerce affair, a variety of Digital Medias or electronic commerce affair, the standardization result for the initial user-project score data for obtaining a variety of Digital Medias or electronic commerce affair is calculated.

Formula is utilized in this step（4) user of the miscellaneous service after the standardization, obtained-project score data.User-project score data is as shown in table 13 after business S standardization：

Table 13

User-project score data after 2 standardization is as shown in table 14:

u₂ 4 u₃ 5 u₄ 4 5

4 3 u₆User-project score data after the standardization of 54 business is as shown in Table 15:Table 15

Step 505:According to the actual user, actual items and standardization result, integrate user-project score data of a variety of Digital Medias or electronic commerce affair, unified user-project the score data of generation, the unified user-project score data includes user-project score data in a variety of Digital Medias or electronic commerce affair after the integration of all users and project.

Aforementioned formula is specifically utilized in step 505（5) unification user, obtained-project score data, it is specific as shown in table 16：

Table 16

In table 16, user and project are all unique between miscellaneous service respectively, then can as seen from Table 16, in 3 kinds of business, and different actual users has 10, and different actual items have 9.

Step 506:The unification user-project score data is sequentially mapped to a variety of Digital Medias or electronic commerce affair, user-project score data after a variety of Digital Media or electronic commerce affair mappings is generated.

Using aforementioned manner three, the related user-project score data of the business after miscellaneous service mapping can be obtained.User-project score data after being mapped in business S is as shown in table 17：

Table 17

5 5

u₂' 2 5 u₃' 4 3

u₄ 4 5

User-project score data after being mapped in the business of Us 45 is as shown in table 18:

Table 18

User-project score data after being mapped in business is as shown in table 19:

Table 19

Us 5 5 u₂' 5 3 5 u₆ 3 3 4 u₉ 3 5 υ_Ί' 5 4 3 4 u_l'₀5 steps 507:The item similarity in same business between disparity items is calculated according to user-project score data after a variety of Digital Media or electronic commerce affair mappings.

Assume to calculate the similarity in same Digital Media or electronic commerce affair between disparity items in step 507, utilize aforementioned formula（9) item similarity for obtaining various Digital Medias or electronic commerce affair can be calculated.Business S item similarity data are as shown in table 20：

Table 20

The item similarity data of business are as shown in table 21:The ι of table 21₆1, '

1.00 0.44 0.49 0.38 0.24

0.44 1.00 0.00 0.77 0.64 0.49 0.00 1.00 0.00 0.42 i₆ 0.38 0.77 0.00 1.00 0.73

The item similarity data of 0.24 0.64 0.42 0.73 1.00 business are as shown in table 22:Table 22

In the present example, because various Digital Medias or electronic commerce affair belong to music field, specification, integration and the mapping for carrying out user-project score data are rational.Pass through the acquisition methods of this item similarity of the present embodiment, user-project score data of business correlation after obtained mapping is more enriched than user-project score data of original miscellaneous service, and it is with a high credibility, the sparse sex chromosome mosaicism of user-project score data can be solved well, and can also improve the validity and accuracy of project recommendation again by selecting user-project score data and corresponding user's similarity and/or item similarity after the mapping of the business well when carrying out project recommendation.The acquisition flow step 601 of the user's similarity being related in the embodiment of the present invention and/or item similarity is introduced:Service identification and the targeted customer's mark that targeted customer is using are obtained by computer network interface.

Targeted customer is to need the user for its recommended project in this step, and target is obtained first and is used Service identification and its user's mark that family is being used.It should be noted that, goal user mark is unique in same business, it is not necessarily unique in different business, but because a target service can be uniquely determined according to service identification, therefore, target service mark can uniquely determine a user in the target service.

Step 602:The business source data prestored is obtained from memory according to the service identification.Wherein, the business source data can specifically include：Similarity, user's matching relationship data and the project matching relationship data between user-project score data, the disparity items of the business after business mapping；Or, similarity, user's matching relationship data and the project matching relationship data between user-project score data, the different user of the business after business mapping.

The similarity of the user-between project score data and the disparity items of the business after business mapping can be obtained from the result of flow chart of data processing under line according to service identification, and user's matching relationship data and project matching relationship data, the similarity of the user-between project score data and the different user of the business after business mapping can also be obtained from the result of flow chart of data processing under line, user's matching relationship data and project matching relationship data can also be got.

Assuming that needing to recommend a project for the user of foregoing business in practice, the target service then got is designated business, targeted customer is designated the user in business, it is the content shown in table 19 that user-project score data after the mapping got is identified according to target service, similarity in the business between disparity items is the content shown in table 22, user's matching relationship data are the content shown in table 10, and project matching relationship data are the content shown in table 12.

Step 603:The service identification and business source data used according to targeted customer mark, targeted customer, is targeted customer's generation candidate recommendation project set.

Need to be identified according to targeted customer in this step, the service identification that targeted customer is using and business source data are that targeted customer generates candidate recommendation project set.The candidate recommendation project is combined can use the combination of following any mode or two ways in acquisition process：

Mode A:Selection and user's similarity of the targeted customer meet the user of prerequisite, and the scoring for user for selecting user's similarity to meet prerequisite is higher than the item design candidate recommendation project set that predetermined threshold and the targeted customer do not score；

Wherein, the candidate recommendation project in the candidate recommendation project set belongs to Digital Media or electronic commerce affair that the targeted customer is using. Wherein, the candidate recommendation project can include：Digital media content, e-commerce product or uniform resource position mark URL.

Mode B:Selection and project of the user-project scoring higher than predetermined threshold value of the targeted customer, and selection and the user-project scoring meet the item design candidate recommendation project set that prerequisite and the targeted customer do not score higher than the item similarity between the project of predetermined threshold value；Wherein, the candidate recommendation project in the candidate recommendation project set belongs to the business that the targeted customer is using.

Wherein, judge whether a project belongs to Digital Media or electronic commerce affair that targeted customer is using, service identification and the actual items mark that can be being used according to targeted customer are judged.

By taking mode B as an example, according to the table 19 and table 22 got in step 601, for user ^5 (in being table 19）User-project data the high project of score value, meet that similarity between the high project of the score value is high and user does not comment undue item design candidate recommendation project set.

In the present example, it is assumed that think project definition of user's score value not less than 3 for the high project of score value, then the project obtained is⁴With;Assume again that the high implication of similarity is not less than 0.4 for similarity, then, with/₄Similarity is high and user does not comment undue project for 6,^r, and¹With similarity is high and user does not comment undue project for 6 and, therefore candidate recommendation project set includes

^ and 9, the project corresponded in business be and.

Step 604:The prediction scoring of each candidate recommendation project in the candidate recommendation project set is obtained according at least to user's similarity in the business source data and/or item similarity.

In actual applications, user to the prediction of project scoring ' formula can be used（10), (11) and（12) any one mode in is calculated：

( 11 ) ∑ simiU^ U,) - ^ ∑ sim{L,I_k) - r_ik

U_k≡NN_Vi

R a

∑ s U^ X \ sim{I I_k) wherein, formula（10) ^^ in represents the neighbour of the set, i.e. user of the user composition high with user's similarity；^^) represent the similarity of user and user；Formula（11) ^ in represents the similar terms collection of the set of the item design high with item similarity, i.e. project, '^W) represent the similarity of project and project 4；Formula（12) value that " is the parameter between 0 to 1, empirically can manually set, or obtained according to training data study, for example constantly adjustment " in, minimum that of selection final error.

Assuming that utilizing formula（11) prediction scoring is calculated, then prediction scoring of the user to the sum of project 2 is respectively 3.96,3.87 and 3.00.

Step 605:The consequently recommended bulleted list that qualified candidate recommendation project generates the targeted customer is extracted from the candidate recommendation project according to the prediction scoring of the candidate recommendation project.

Final bulleted list is generated for targeted customer according to prediction scoring, the final bulleted list includes several higher projects of prediction scoring, and specifically choosing how many projects can also be adjusted according to actual conditions.Assuming that taking prediction scoring highest candidate recommendation project as final bulleted list in this example, then final bulleted list is project ^.It is of course also possible to select project and as final bulleted list.

Step 606:Digital Media or electronic commerce affair server, which send the consequently recommended bulleted list to the client of the targeted customer, to be shown.

After final bulleted list is generated, Digital Media or electronic commerce affair server are shown in the client that consequently recommended bulleted list is sent to the targeted customer.

In the present embodiment, when carrying out project recommendation, according to targeted customer's mark and service identification, the user after the mapping of the business-project score data and corresponding user's similarity and/or item similarity can be selected, by directly utilizing the user's similarity and/or item similarity stored in memory, it is that preferably have selected the user-project score data and corresponding similarity after the mapping of the business, so reducing the processing time of project recommendation, improve the efficiency of project recommendation, and the validity and accuracy of project recommendation can be improved.With reference to Fig. 7, the embodiment of the invention also discloses item recommendation method during a kind of repeat in work, This method includes user's similarity and/or the acquisition flow and project recommendation flow of item similarity simultaneously；Specifically, the item recommendation method during repeat in work may include steps of：

Step 701:The original source data of a variety of different digital media or electronic commerce affair is obtained, the original source data includes：The initial user of multiple business-project score data.

Step 702:According to the standardization result of initial user-project score data of the user between a variety of Digital Medias or electronic commerce affair and the matching result of project and a variety of Digital Medias or electronic commerce affair, initial user-project score data of a variety of Digital Medias or electronic commerce affair is integrated into the user including each a variety of Digital Medias or electronic commerce affair and unification user-project score data of project.

Step 703:The unification user-project score data is sequentially mapped to a variety of Digital Medias or electronic commerce affair, user-project score data after a variety of Digital Media or electronic commerce affair mappings is generated.

Step 704:According to the item similarity between disparity items in the user's similarity and/or same business between different user in user-same business of project score data acquisition after a variety of Digital Media or electronic commerce affair mappings.

Step 705:User's similarity and/or item similarity are stored into the memory.It should be noted that, the storage user's similarity and/or the process of item similarity that step 701 step 705 is illustrated may be considered preprocessing process, the project recommendation process that can be illustrated with subsequent step 706 711 is independently carried out, and so also can guarantee that the real-time and validity of project recommendation.Preprocessing process and project recommendation process are just subjected to Jie's step 706 in sequence for convenience's sake in the present embodiment:Service identification and the targeted customer's mark that targeted customer is using are obtained by computer network interface.

Step 707:The business source data prestored is obtained from memory according to the service identification.Step 708:Identified according to the targeted customer, the Digital Media that targeted customer is using or electronic commerce affair are identified and source data, be targeted customer's generation candidate recommendation project set.

Step 709:The prediction scoring of each candidate recommendation project in the candidate recommendation project set is obtained according at least to user's similarity in the business source data and/or item similarity.

Step 710:According to the prediction of candidate recommendation project scoring from the candidate recommendation project Extract the consequently recommended bulleted list that qualified candidate recommendation project generates the targeted customer.Step 711:Digital Media or electronic commerce affair server, which send the final project recommendation list to the client of the targeted customer, to be shown.

Because user's similarity and/or the acquisition flow of item similarity and project recommendation flow in embodiment before by the agency of it is very detailed, so the not most part of the present embodiment, may be referred to the related introduction of user's similarity and/or the acquisition flow and project recommendation flow of item similarity.

It should be noted that, for foregoing each method embodiment, in order to be briefly described, therefore it is all expressed as to a series of combination of actions, but those skilled in the art should know, the present invention is not limited by described sequence of movement, because according to the present invention, some steps can be carried out sequentially or simultaneously using other.Secondly, those skilled in the art should also know, embodiment described in this description belongs to preferred embodiment, and involved action and the module not necessarily present invention are necessary.

It is corresponding with the user's similarity and/or the acquisition methods of item similarity that the embodiments of the present invention are provided, referring to Fig. 8, the embodiment of the present invention additionally provides the structural representation of user's similarity and/or the acquisition embodiment of item similarity, can specifically include：

Original source data unit 801 is obtained, the original source data for obtaining different digital media or electronic commerce affair, the original source data includes：The initial user of a variety of Digital Medias or electronic commerce affair-project score data；

Integral unit 802, for the standardization result of the initial user according to the user between a variety of Digital Medias or electronic commerce affair and the matching result of project and a variety of Digital Medias or electronic commerce affair-project score data, initial user-project score data of a variety of Digital Medias or electronic commerce affair is integrated into the user including a variety of Digital Medias or electronic commerce affair and unification user-project score data of project；

The original source data can also include：The user attribute data and item attribute data of a variety of Digital Medias or electronic commerce affair, with reference to shown in Fig. 9, then the integral unit 802 can specifically include：

First coupling subelement 901, for according to initial user mark, initial user attribute and the initial user property value in the user attribute data, matching to obtain unique actual user between a variety of Digital Medias or electronic commerce affair；The initial user mark represents unique in a certain business User；The initial user property value is used to represent all unique user between the multiple business；Second coupling subelement 902, for according to the initial item identification in the item attribute data, initial project attribute and initial project property value, matching to obtain unique actual items between a variety of Digital Medias or electronic commerce affair；The initial item identification represents unique project in a certain business；

With reference to shown in Figure 10, second coupling subelement 902 is specific can be included again：

3rd coupling subelement 1001, for utilizing actual items attribute different between a variety of Digital Medias or a variety of Digital Medias of initial project attributes match or electronic commerce affair of electronic commerce affair；Second obtains subelement 1002, for obtaining item attribute collection registration and each business and the item attribute collection registration average of other business between a variety of Digital Medias or electronic commerce affair according to the different actual items attribute；

Sorted subelement 1003, and a variety of Digital Medias or electronic commerce affair are ranked up for the size according to the item attribute collection registration average；

Business coupling subelement 1004, for the sequencing according to a variety of Digital Medias or electronic commerce affair after sequence, business matching flow is performed by current business of the first business, the business matching flow includes：The matching entries of the current business and other business are determined, and, delete the current business；

With reference to shown in Fig. 11, the business coupling subelement 1004 is specific to include again：Project coupling subelement 1102, for the initial item identification order included according to the first business, first project of selection matches flow as current project project implementation；The project matching flow includes：Calculate current project and the project matching degree of each project in other business；For each other business, suitable project matching degree is chosen according to default threshold condition, to form multiple project matching degree set；Matching degree highest project is selected in each project matching degree set as the matching entries of current project；The matching relationship of the matched project of the current project is recorded, and deletes the matching entries that miscellaneous service includes；Delete the current project；

Subelement 1103 is circulated, using second project in the first described business as current project, the project matching flow is performed, until the project included in the first described business is sky.

3rd obtains subelement 1005, for performing business matching flow by the current business of second of business, until the business in the sequence is space-time, is matched according to the matching entries and project Relation acquisition all unique actual items between a variety of Digital Medias or electronic commerce affair.First obtains subelement 903, for the minimum value of initial user-project score data according to multiple business, the user of multiple business-project scoring score range and the score range, the standardization result of initial user-project score data of the multiple business is obtained；

Integrate subelement 904, for according to the actual user, actual items and standardization result, integrate user-project score data of a variety of Digital Medias or electronic commerce affair, unified user-project the score data of generation, the unified user-project score data includes user-project score data after the integration of user and project in a variety of Digital Medias or electronic commerce affair.

Score data unit 803 is generated, for the unification user-project score data to be sequentially mapped into a variety of Digital Medias or electronic commerce affair, user-project score data after a variety of Digital Media or electronic commerce affair mappings is generated；

In actual applications, the generation score data unit 803, further can be used for：For every kind of business, all projects and/or the corresponding all users-project score data of total user that every kind of Digital Media or electronic commerce affair are included are extracted in unified user-project score data, user-project score data after the mapping of miscellaneous service is constituted.

Similarity unit 804 is obtained, for calculating the similarity in the similarity and/or same business in same business between different user between disparity items according to user-project score data after a variety of Digital Media or electronic commerce affair mappings.

In actual applications, the acquisition similarity unit 804, further can be used for：

The item destination aggregation (mda) that the item destination aggregation (mda) and described two different users scored jointly according to two different users in the user after the mapping-project score data, same business each scores calculates the similarity between described two different users；And/or,

The similarity between described two disparity items is calculated according to the set of the user scored jointly two disparity items in the user after the mapping-project score data, same business and the set of the user each scored two disparity items.

Memory cell 805, for user's similarity and/or item similarity to be stored into the memory.

The acquisition system of user's similarity and/or item similarity disclosed in the embodiment of the present invention, because being integrated by the specification of user-project score data and mapping to calculate user's similarity or project Similarity, not only can provide data for project recommendation, the user after mapping that can also be by selecting the business well-project score data and corresponding user's similarity and/or item similarity, can improve the validity accuracy of project recommendation.With reference to shown in Figure 12, the embodiment of the invention also discloses item recommendation system during a kind of repeat in work, the item recommendation system includes：

Mark unit 1201 is obtained, is identified for the Digital Media or electronic commerce affair mark that are being used by computer network interface acquisition targeted customer and targeted customer；

Acquisition business source data units 1202, for obtaining the business source data prestored from memory according to the service identification；

Candidate collection unit 1203 is generated, for being identified according to the targeted customer, the Digital Media that targeted customer is using or electronic commerce affair are identified and business source data, are targeted customer's generation candidate recommendation project set；

In actual applications, the generation candidate collection unit 1202 further can be used for：Selection and user's similarity of the targeted customer meet the user of prerequisite, and the scoring for user for selecting user's similarity to meet prerequisite is higher than the item design candidate recommendation project set that predetermined threshold and the targeted customer do not score；And/or,

Selection and project of the user-project scoring higher than predetermined threshold value of the targeted customer, and selection and the user-project scoring meet the item design candidate recommendation project set that prerequisite and the targeted customer do not score higher than the item similarity between the project of predetermined threshold value.

Wherein, the candidate recommendation project belongs to the Digital Media or electronic commerce affair that the targeted customer is using.

Prediction scoring unit 1204 is obtained, for according at least to the user's similarity and/or item similarity in the business source data, obtaining the prediction scoring of each candidate recommendation project in the candidate recommendation project set；

Final list cell 1205 is generated, the consequently recommended bulleted list that qualified candidate recommendation project generates the targeted customer is extracted from the candidate recommendation project for the prediction scoring according to the candidate recommendation project；

Display unit 1206, for the final project recommendation list to be sent to the targeted customer's Client is shown.

The item recommendation system of the present embodiment is when carrying out project recommendation, according to targeted customer's mark and service identification, the user after the mapping of the business-project score data and corresponding user's similarity and/or item similarity can be selected, by directly utilizing the user's similarity and/or item similarity stored in memory, it is that preferably have selected the user-project score data and corresponding similarity after the mapping of the business, so reducing the processing time of project recommendation, improve the efficiency of project recommendation, and the validity and accuracy of project recommendation can be improved.

It should be noted that, in the item recommendation system of practical application, perform the system for obtaining user's similarity and/or item similarity and the system for carrying out project recommendation, can independently it work, it can be carried out simultaneously because obtaining the project recommendation of user's similarity and/or item similarity, only the user's similarity calculated and/or item similarity need to can be obtained when carrying out project recommendation, so also can guarantee that the real-time and validity of item recommendation system institute recommended project.

It should be noted that each embodiment in this specification is described by the way of progressive, what each embodiment was stressed be between the difference with other embodiment, each embodiment identical similar part mutually referring to.For system class embodiment, because it is substantially similar to embodiment of the method, so description is fairly simple, the relevent part can refer to the partial explaination of embodiments of method.

It can further be stated that, herein, term " including ", " including " or any other variant thereof is intended to cover non-exclusive inclusion, so that process, method, article or equipment including a series of key elements not only include those key elements, but also other key elements including being not expressly set out, or also include for this process, method, article or the intrinsic key element of equipment.In the absence of more restrictions, the key element limited by sentence " including one ... ... ", it is not excluded that also there is other identical element in the process including the key element, method, article or equipment.

One of ordinary skill in the art will appreciate that all or part of step in the various methods of above-described embodiment can be by program to instruct the hardware of correlation to complete, the program can be stored in a computer-readable recording medium, and storage medium can include：ROM, RAM, disk or CD etc..

The explanation that item recommendation method and system during the repeat in work provided above the embodiment of the present invention have carried out upper embodiment is only intended to help the method for understanding the embodiment of the present invention and its thought；Simultaneously for Those of ordinary skill in the art, according to the thought of the embodiment of the present invention, will change, in summary, this specification content should not be construed as limiting the invention in specific embodiments and applications.

Claims

Claim

1st, item recommendation method when a kind of Digital Media or electronic commerce affair intersect, it is characterised in that including：

Digital Media or electronic commerce affair mark and the targeted customer's mark that targeted customer is using are obtained by computer network interface, and obtain the business source data prestored from memory according to the service identification；

Identified according to the targeted customer, the Digital Media that targeted customer is using or electronic commerce affair are identified and business source data, be targeted customer's generation candidate recommendation project set；

The prediction scoring of each candidate recommendation project in the candidate recommendation project set is obtained according at least to user's similarity in the business source data and/or item similarity；

The consequently recommended bulleted list that qualified candidate recommendation project generates the targeted customer is extracted from the candidate recommendation project according to the prediction scoring of the candidate recommendation project；

Digital Media or electronic commerce affair server, which send the final project recommendation list to the client of the targeted customer, to be shown.

2nd, according to the method described in claim 1, it is characterized in that, the Digital Media used according to targeted customer mark, targeted customer or electronic commerce affair mark and business source data, are targeted customer's generation candidate recommendation project set, including：

Selection and user's similarity of the targeted customer meet the user of prerequisite, and the scoring for user for selecting user's similarity to meet prerequisite is higher than the item design candidate recommendation project that predetermined threshold and the targeted customer do not score；And/or,

Selection and project of the user-project scoring higher than predetermined threshold value of the targeted customer, and selection and the user-project scoring meet the item design candidate recommendation project that prerequisite and the targeted customer do not score higher than the item similarity between the project of predetermined threshold value；

Wherein, the candidate recommendation project belongs to the Digital Media or electronic commerce affair that the targeted customer is using.

3rd, according to the method described in claim 1, it is characterised in that also include：

The original source data of a variety of different digital media or electronic commerce affair is obtained, the original source data includes：Initial user-project score data of a variety of Digital Medias or electronic commerce affair；According to the user between a variety of Digital Medias or electronic commerce affair and the matching result of project And the standardization result of the initial user of a variety of Digital Medias or electronic commerce affair-project score data, initial user-project score data of a variety of Digital Medias or electronic commerce affair is integrated into the user including a variety of Digital Medias or electronic commerce affair and unification user-project score data of project；

The unification user-project score data is sequentially mapped to a variety of Digital Medias or electronic commerce affair, user-project score data after a variety of Digital Media or electronic commerce affair mappings is generated；

According to the item similarity between disparity items in the user's similarity and/or same business between different user in user-same business of project score data acquisition after a variety of Digital Media or electronic commerce affair mappings；

User's similarity and/or item similarity are stored into the memory.

4th, method according to claim 3, it is characterised in that the original source data also includes：The user attribute data and item attribute data of a variety of Digital Medias or electronic commerce affair, then initial user-project the score data by a variety of Digital Medias or electronic commerce affair is integrated into the user including a variety of Digital Medias or electronic commerce affair and unification user-project score data of project, including：

According to initial user mark, initial user attribute and the initial user property value in the user attribute data, matching obtains unique actual user between a variety of Digital Medias or electronic commerce affair；The initial user mark represents unique user in a certain business；The initial user property value is used to represent all unique user between a variety of Digital Medias or electronic commerce affair；

According to the initial item identification in the item attribute data, initial project attribute and initial project property value, matching obtains unique actual items between a variety of Digital Medias or electronic commerce affair；The initial item identification represents unique project in a certain business；

According to the user-project scoring score range and the minimum value of the score range of initial user-project score data of a variety of Digital Medias or electronic commerce affair, a variety of Digital Medias or electronic commerce affair, the standardization result of initial user-project score data of a variety of Digital Medias or electronic commerce affair is obtained；

According to the actual user, actual items and standardization result, user-project score data of a variety of Digital Medias or electronic commerce affair is integrated, unified user-project score data is generated, Unified user-project the score data includes user-project score data after the integration of user and project in a variety of Digital Medias or electronic commerce affair.

5th, method according to claim 4, it is characterized in that, initial item identification, initial project attribute and the initial project property value according in the item attribute data, matching obtains unique actual items between a variety of Digital Medias or electronic commerce affair, including：

Utilize actual items attribute different between Digital Media or ecommerce multiple business described in the initial project attributes match of a variety of Digital Medias or electronic commerce affair；

The item attribute collection registration and each business and the item attribute collection registration average of other business between a variety of Digital Medias or electronic commerce affair are obtained according to the different actual items attribute；

The a variety of Digital Medias or electronic commerce affair are ranked up according to the size of the item attribute collection registration average；

According to the sequencing of a variety of Digital Medias or electronic commerce affair after sequence, business matching flow is performed by current business of the first business, the business matching flow includes：The matching entries of the current business and other business are determined, and, delete the current business；

Business matching flow is performed by the current business of second of business, until the business in the sequence is space-time, all unique actual items between a variety of Digital Medias or electronic commerce affair are obtained according to the matching entries and project matching relationship.

6th, method according to claim 5, it is characterised in that the matching entries of the determination current business and other business, including：

The initial item identification order included according to current business, first project of selection matches flow as current project project implementation；The project matching flow includes：Calculate current project and the project matching degree of each project in other business；For each other business, suitable project matching degree is chosen according to default threshold condition, to form multiple project matching degree set；Matching degree highest project is selected in each project matching degree set as the matching entries of current project；The matching relationship of the matched project of the current project is recorded, and deletes the matching entries that the multiple business includes；Delete the current project；

Using second project in the current business as current project, the project matching flow is performed, until the project included in the current business is sky. 7th, the method according to claim 3, it is characterized in that, it is described that the unification user-project score data is sequentially mapped to a variety of Digital Medias or electronic commerce affair, user-project score data after a variety of Digital Medias or electronic commerce affair mapping is generated, including：

For every kind of business, all projects and/or the corresponding all users-project score data of total user that every kind of business is included are extracted in unified user-project score data, user-project score data after the mapping of miscellaneous service is constituted.

8th, method according to claim 3, it is characterized in that, similarity in similarity and/or same business in the user-same business of project score data acquisition according to after a variety of Digital Media or electronic commerce affair mappings between different user between disparity items, including：

The item destination aggregation (mda) that the item destination aggregation (mda) and described two different users scored jointly according to two different users in the user after the mapping-project score data, same business each scores calculates the similarity between described two different users；And/or,

The similarity between described two disparity items is calculated according to the set of the user scored jointly two disparity items in the user after the mapping-project score data, same business and the set of the user each scored two disparity items.

9th, according to the method described in claim 1, it is characterised in that the Digital Media or electronic commerce affair, include but is not limited to：Music, using download, internet book store, electronic reading, game and/or shopping online.

10th, according to the method described in claim 1, it is characterised in that the candidate recommendation project includes：Digital media content, e-commerce product or uniform resource position mark URL.

11st, item recommendation system when a kind of Digital Media or electronic commerce affair intersect, it is characterised in that including：

Mark unit is obtained, is identified for the Digital Media or electronic commerce affair mark that are being used by computer network interface acquisition targeted customer and targeted customer；

Acquisition business source data units, for obtaining the business source data prestored from memory according to the service identification；

Candidate collection unit is generated, for being identified according to the targeted customer, the Digital Media that targeted customer is using or electronic commerce affair are identified and business source data, are targeted customer's generation candidate recommendation project set； Prediction scoring unit is obtained, for according at least to the user's similarity and/or item similarity in the business source data, obtaining the prediction scoring of each candidate recommendation project in the candidate recommendation project set；

Final list cell is generated, the consequently recommended bulleted list that qualified candidate recommendation project generates the targeted customer is extracted from the candidate recommendation project for the prediction scoring according to the candidate recommendation project；

Display unit, is shown for the final project recommendation list to be sent to the client of the targeted customer.

12nd, system as claimed in claim 11, it is characterised in that the generation candidate collection unit is further used for：

Selection and user's similarity of the targeted customer meet the user of prerequisite, and the scoring for user for selecting user's similarity to meet prerequisite is higher than the item design candidate recommendation project that predetermined threshold and the targeted customer do not score；And/or,

Selection and project of the user-project scoring higher than predetermined threshold value of the targeted customer, and selection and the user-project scoring meet the item design candidate recommendation project that prerequisite and the targeted customer do not score higher than the item similarity between the project of predetermined threshold value；

Wherein, the candidate recommendation project belongs to the Digital Media or electronic commerce affair that the targeted customer is using.

13rd, system as claimed in claim 11, it is characterised in that also include：

Original source data unit is obtained, the original source data for obtaining different digital media or electronic commerce affair, the original source data includes：Initial user-project score data of a variety of Digital Medias or electronic commerce affair；

Integral unit, for the standardization result of the initial user according to the user between a variety of Digital Medias or electronic commerce affair and the matching result of project and a variety of Digital Medias or electronic commerce affair-project score data, initial user-project score data of a variety of Digital Medias or electronic commerce affair is integrated into the user including a variety of Digital Medias or electronic commerce affair and unification user-project score data of project；

Score data unit is generated, for the unification user-project score data to be sequentially mapped into a variety of Digital Medias or electronic commerce affair, a variety of Digital Medias or electronic commerce affair are generated User-project score data after mapping；

Similarity unit is obtained, for obtaining the item similarity in the user's similarity and/or same business in same business between different user between disparity items according to user-project score data after a variety of Digital Media or electronic commerce affair mappings；

Memory cell, for user's similarity and/or item similarity to be stored into the memory.

14th, system as claimed in claim 13, it is characterised in that the original source data also includes：The user attribute data and item attribute data of a variety of Digital Medias or electronic commerce affair, then the integral unit include：

First coupling subelement, for according to initial user mark, initial user attribute and the initial user property value in the user attribute data, matching to obtain unique actual user between a variety of Digital Medias or electronic commerce affair；The initial user mark represents unique user in a certain business；The initial user property value is used to represent all unique user between a variety of Digital Medias or electronic commerce affair；

Second coupling subelement, for according to the initial item identification in the item attribute data, initial project attribute and initial project property value, matching to obtain unique actual items between a variety of Digital Medias or electronic commerce affair；The initial item identification represents unique project in a certain business；

First obtains subelement, for the minimum value of the user according to initial user-project score data of multiple digital media or electronic commerce affair, a variety of Digital Medias or electronic commerce affair-project scoring score range and the score range, the standardization result of initial user-project score data of a variety of Digital Medias or electronic commerce affair is obtained；

Integrate subelement, for according to the actual user, actual items and standardization result, integrate user-project score data of a variety of Digital Medias or electronic commerce affair, unified user-project the score data of generation, the unified user-project score data includes user-project score data after the integration of user and project in a variety of Digital Medias or electronic commerce affair.

15th, system as claimed in claim 14, it is characterised in that second coupling subelement includes：

3rd coupling subelement, for utilizing the initial of a variety of Digital Medias or electronic commerce affair Item attribute matches actual items attributes different between a variety of Digital Medias or electronic commerce affair；Second obtains subelement, for obtaining item attribute collection registration and each business and the item attribute collection registration average of other business between a variety of Digital Medias or electronic commerce affair according to the different actual items attribute；

Sorted subelement, and a variety of Digital Medias or electronic commerce affair are ranked up for the size according to the item attribute collection registration average；

Business coupling subelement, for the sequencing according to a variety of Digital Medias or electronic commerce affair after sequence, business matching flow is performed by current business of the first business, the business matching flow includes：The matching entries of the current business and other business are determined, and, delete the current business；

3rd obtains subelement, for performing business matching flow by the current business of second of business, until the business in the sequence is space-time, all unique actual items between a variety of Digital Medias or electronic commerce affair are obtained according to the matching entries and project matching relationship.

16th, system as claimed in claim 15, it is characterised in that the business coupling subelement concrete configuration is：

Project coupling subelement, for the initial item identification order included according to current business, first project of selection matches flow as current project project implementation；The project matching flow includes：Calculate current project and the project matching degree of each project in other business；For each other business, suitable project matching degree is chosen according to default threshold condition, to form multiple project matching degree set；Matching degree highest project is selected in each project matching degree set as the matching entries of current project；The matching relationship of the matched project of the current project is recorded, and deletes the matching entries that the multiple business includes；Delete the current project；

Subelement is circulated, for using second project in the current business as current project, performing the project matching flow, until the project included in the current business is sky.

17th, system as claimed in claim 13, it is characterised in that the generation score data unit concrete configuration is：

For every kind of business, all projects and/or the corresponding all users-project score data of total user that every kind of business is included are extracted in unified user-project score data, user-project score data after the mapping of a variety of Digital Medias or electronic commerce affair is constituted. 18th, system as claimed in claim 13, it is characterised in that the acquisition similarity unit concrete configuration is：

The item destination aggregation (mda) that the item destination aggregation (mda) and described two different users scored jointly according to two different users in the user after the mapping-project score data, same business each scores calculates the similarity between described two different users；And/or,

The similarity between described two disparity items is calculated according to the set of the user scored jointly two disparity items in the user after the mapping-project score data, same business and the set of the user each scored two disparity items.