CN107730336A - Commodity method for pushing and device in a kind of online transaction - Google Patents
Commodity method for pushing and device in a kind of online transaction Download PDFInfo
- Publication number
- CN107730336A CN107730336A CN201610662922.8A CN201610662922A CN107730336A CN 107730336 A CN107730336 A CN 107730336A CN 201610662922 A CN201610662922 A CN 201610662922A CN 107730336 A CN107730336 A CN 107730336A
- Authority
- CN
- China
- Prior art keywords
- commodity
- correlation rule
- pointed
- rule
- degree
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
Abstract
The embodiment of the invention discloses the commodity method for pushing in a kind of online transaction and device, it is related to Internet technical field, it is possible to increase recommend the accuracy rate of the commodity of user.The present invention includes:The order data of commodity pointed by pending correlation rule is read, correlation rule is used to represent at least two commodity that incidence relation mutually be present;According to the order data of pointed commodity, the support and confidence level of correlation rule are obtained, for representing conditional probability of the commodity pointed by correlation rule in date granularity, confidence level is used to represent conditional probability of the commodity pointed by correlation rule on sales volume support;Invalid rule is filtered from correlation rule according to the support of correlation rule and confidence level;According to the correlation rule by filtering, commodity to be recommended are determined, and commodity to be recommended are pushed to user equipment.The commodity that the present invention is applied to related online transaction of user when doing shopping push.
Description
Technical field
The present invention relates to the commodity method for pushing in Internet technical field, more particularly to a kind of online transaction and device.
Background technology
Current major online shopping platform all employs some recommendation carried out around merchandise sales, promotion plans, with
Just dependent merchandise is recommended when user does shopping, so as to increase trading volume.Mainly phase between commodity is bought by analyzing user
Closing property after user completes single purchase behavior, actively recommends dependent merchandise according to recommendation rules to obtain recommendation rules, from
And add the cross-selling chance of commodity.
But rule analysis is associated at present or based on the quantity on order granularity actually accumulated, once run into promotion
During the non-daily sales volume fluctuation such as activity, much-sought-after item, it is too high make it that ratio occur in correlation rule in this kind of commodity, but
Often the market demand is larger for these classification commodity, can not be shielded during association rule mining.And ultimately result in, due to
The sales volume very different of each commodity, and the external factor for influenceing Sales Volume of Commodity in practice is sufficiently complex so that recommend use
The accuracy rate of the commodity at family is relatively low, it is difficult to meets the real demand of user.
The content of the invention
Embodiments of the invention provide the commodity method for pushing and device in a kind of online transaction, it is possible to increase recommend use
The accuracy rate of the commodity at family.
To reach above-mentioned purpose, embodiments of the invention adopt the following technical scheme that:
In a first aspect, the method that embodiments of the invention provide, including:Read commodity pointed by pending correlation rule
Order data, the correlation rule is used to represent at least two commodity that incidence relation mutually be present;According to described pointed
The order data of commodity, obtains the support and confidence level of the correlation rule, and the support is used to represent the association rule
Then conditional probability of the pointed commodity in date granularity, the confidence level are used to represent that commodity pointed by the correlation rule exist
Conditional probability on sales volume;According to the support and confidence level of the correlation rule, invalid rule are filtered from the correlation rule
Then;According to the correlation rule by filtering, commodity to be recommended are determined, and the commodity to be recommended are pushed to user equipment.
With reference in a first aspect, in the first possible implementation of first aspect, the pass according to by filtering
Connection rule, determines commodity to be recommended, including:Default recommendation bits number is determined, and according to the correlation rule by filtering, is obtained
The corresponding default commodity to be recommended for recommending bits number;The default weighting is than being more than or equal to 40:1.
It is described according to the pointed business with reference in a first aspect, in second of possible implementation of first aspect
The order data of product, the support of the correlation rule is obtained, including:According to the order data of the pointed commodity, it is determined that
The aggregate-value for the place period that the pointed commodity are bought simultaneously, and, the pointed commodity are bought simultaneously first
At the time of to current time period total value;According to the aggregate-value and the total value, the branch of the correlation rule is obtained
Degree of holding.
With reference in a first aspect, in the third possible implementation of first aspect, in addition to:According to the order numbers
According to determining the sales situations of commodity pointed by the correlation rule, the sales situation includes:The association rule are have purchased simultaneously
The then order number of pointed commodity, the order number of commodity pointed by the correlation rule is not bought simultaneously, is not all purchased
Buy the order number of commodity pointed by the correlation rule;Interest-degree is obtained according to the sales situation, and according to the interest
Degree whether there is association between detecting commodity pointed by the correlation rule, and the pass of association between commodity pointed by reservation be present
Connection rule.
With reference to the third possible implementation of first aspect, in the 4th kind of possible implementation, the basis
The sales situation obtains interest-degree, and whether there is between the commodity according to pointed by the interest-degree detects the correlation rule
Association, including:Obtain interest-degreeWherein, institute
Two kinds of commodity of A, B pointed by correlation rule are stated, a is represented to have purchased A and B order number simultaneously, and b is represented to have bought A but do not bought B
Order number, c is represented to have bought B but is not bought A order number, and d is represented without the order number for buying A and B;According to institute
State interest-degree and determine degree of correlation situation between commodity pointed by the correlation rule, and judged according to the degree of correlation situation be
No association to be present, the degree of correlation situation includes:When the interest-degree is more than zero, B and A sales situation positive correlation is represented;Institute
When stating interest-degree less than zero, represent that B and A sales situation is negatively correlated;When the interest-degree is equal to zero, B and A is not deposited independently of each other
In correlation.
With reference in a first aspect, in the 5th kind of possible implementation of first aspect, including:Determine the correlation rule
The period of appearance, and determine the distribution situation in time of the correlation rule;According to the temporal distribution situation,
Filter the correlation rule that concentration degree is more than threshold value.
With reference to the 5th kind of possible implementation of first aspect, in the 6th kind of possible implementation, the determination
The distribution situation in time of the correlation rule, including:The period occurred according to the correlation rule, obtain every pass
Join rule time span and number of days support, wherein, time span represent correlation rule occur first to finally go out current moment
Total duration, number of days support represent correlation rule occur total number of days and time span ratio;If correlation rule
Time span is more than first threshold and number of days support is more than Second Threshold, then retains this correlation rule.
With reference in a first aspect, in the 7th kind of possible implementation of first aspect, the pass according to by filtering
Connection rule, commodity to be recommended are determined, and the commodity to be recommended are pushed to user equipment, including:According to described by filtering
The first correlation rule, it is determined that corresponding with main commodity to be recommended first from commodity, and obtain it is described by filtering first
The interest-degree of correlation rule;The similar commodity set corresponding with the main commodity pointed by the second correlation rule is determined, and is obtained
Take the interest-degree of second correlation rule;Closed according to the interest-degree of first correlation rule by filtering and described second
Join the interest-degree of rule obtains the interest-degree of the 3rd correlation rule, and is advised according to described by the 3rd association of filtering
Then, it is determined that corresponding with the similar commodity set to be recommended second from commodity;By described first from commodity and described
Two fill recommendation position from commodity, and the commodity in the recommendation position are pushed to the user equipment.
Second aspect, the device that embodiments of the invention provide, including:Including:Data read module, wait to locate for reading
The order data of commodity pointed by the correlation rule of reason, the correlation rule are used to represent that at least two mutually have incidence relation
Commodity;Analysis module, for the order data according to the pointed commodity, obtain the support of the correlation rule and put
Reliability, the support are used to represent conditional probability of the commodity in date granularity pointed by the correlation rule, the confidence
Spend for representing conditional probability of the commodity on sales volume pointed by the correlation rule;And according to the support of the correlation rule
And confidence level, invalid rule is filtered from the correlation rule;Pushing module, for according to by filtering correlation rule, really
Fixed commodity to be recommended, and the commodity to be recommended are pushed to user equipment.
With reference to second aspect, in the first possible implementation of second aspect, the analysis module, it is specifically used for
According to the order data of the pointed commodity, the accumulative of the place period that the pointed commodity are bought simultaneously is determined
Value, and, at the time of the pointed commodity are bought first simultaneously to current time period total value;And according to described tired
Evaluation and the total value, obtain the support of the correlation rule;The analysis module, it is additionally operable to according to the order data,
The sales situation of commodity pointed by the correlation rule is determined, the sales situation includes:It has purchased the correlation rule simultaneously
The order number of pointed commodity, the order number of commodity pointed by the correlation rule is not bought simultaneously, is not all bought
The order number of commodity pointed by the correlation rule;And interest-degree is obtained according to the sales situation, and according to the interest
Degree whether there is association between detecting commodity pointed by the correlation rule, and the pass of association between commodity pointed by reservation be present
Connection rule;The pushing module, specifically for determining default recommendation bits number, and according to the correlation rule by filtering, obtain
Take the corresponding default commodity to be recommended for recommending bits number;Wherein, the default weighting is than being more than or equal to 40:1;It is described
Pushing module, it is additionally operable to according to first correlation rule by filtering, it is determined that corresponding with main commodity to be recommended first
From commodity, and obtain the interest-degree of first correlation rule by filtering;And determine the second correlation rule pointed by with
Similar commodity set corresponding to the main commodity, and obtain the interest-degree of second correlation rule;Passed through further according to described
The interest-degree of first correlation rule of filter and the interest-degree of second correlation rule obtain the interest-degree of the 3rd correlation rule,
And according to the 3rd correlation rule by filtering, it is determined that corresponding with the similar commodity set to be recommended second
From commodity;Afterwards by described first from commodity and described second from commodity fill recommend position, and by it is described recommendation position in business
Product push to the user equipment.
The quantity on order granularity actually accumulated relative to being currently based on is associated the mode of rule analysis, and the present invention is implemented
Commodity method for pushing and device in the online transaction that example provides, for the screening and filtering of correlation rule, based on correlation rule
The confidence level of support and correlation rule, the F1 values that the mediation for being weighted calculating by support and confidence level obtains filter nothing
Effect rule, by optimize support simultaneously and confidence level this biobjective scheduling problem be converted into single optimization F1 values single goal it is excellent
Change problem.And from rule occur number of days granularity on come consider rule time stability, rule are defined in number of days granularity
Support then, the support and confidence level of the date granularity of correlation rule are specifically employed, filtered out because promotion causes
Invalid rule, so as to improve the hits for recommending the commodity of user, accuracy rate and recall rate.
Brief description of the drawings
Technical scheme in order to illustrate the embodiments of the present invention more clearly, it will use below required in embodiment
Accompanying drawing is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the present invention, for ability
For the those of ordinary skill of domain, on the premise of not paying creative work, it can also be obtained according to these accompanying drawings other attached
Figure.
Fig. 1 is system architecture schematic diagram provided in an embodiment of the present invention;
Fig. 2 is the schematic flow sheet of the commodity method for pushing in online transaction provided in an embodiment of the present invention;
Fig. 3 is the commodity pusher structural representation in online transaction provided in an embodiment of the present invention.
Embodiment
To make those skilled in the art more fully understand technical scheme, below in conjunction with the accompanying drawings and specific embodiment party
Formula is described in further detail to the present invention.Embodiments of the present invention are described in more detail below, the embodiment is shown
Example is shown in the drawings, wherein same or similar label represents same or similar element or has identical or class from beginning to end
Like the element of function.Embodiment below with reference to accompanying drawing description is exemplary, is only used for explaining the present invention, and can not
It is construed to limitation of the present invention.Those skilled in the art of the present technique are appreciated that unless expressly stated, odd number shape used herein
Formula " one ", "one", " described " and "the" may also comprise plural form.It is to be further understood that the specification of the present invention
The middle wording " comprising " used refers to the feature, integer, step, operation, element and/or component be present, but it is not excluded that
In the presence of or other one or more features of addition, integer, step, operation, element, component and/or their groups.It should be understood that
When we claim element to be " connected " or during " coupled " to another element, it can be directly connected or coupled to other elements, or
There may also be intermediary element.In addition, " connection " used herein or " coupling " can include wireless connection or coupling.Here make
Wording "and/or" includes any cell of one or more associated list items and all combined.The art
Technical staff is appreciated that unless otherwise defined all terms (including technical term and scientific terminology) used herein have
With the general understanding identical meaning of the those of ordinary skill in art of the present invention.It is it should also be understood that such as general
Those terms defined in dictionary, which should be understood that, has the meaning consistent with the meaning in the context of prior art, and
Unless being defined as here, will not be explained with the implication of idealization or overly formal.
The method flow that present example is provided, can apply includes in a kind of system as shown in Figure 1, the system:
Online plateform system, management server and user equipment, wherein, management server is connected with online plateform system, or management
Server is some server apparatus being integrated in online plateform system.User equipment can pass through with online plateform system
Internet or wireless communication.
Online plateform system disclosed in the present embodiment can be:Online shopping platform, e-commerce platform etc., specifically
It is a kind of database for being used to carry out data interaction with user equipment for including multiple data servers and storage server composition
Server cluster system.The order datas of commodity, user browse the online plateform system such as record during day-to-day operation
Caused data are storable in the storage server of line platform system, or storage is that online plateform system establishes communication
Database Systems.It should be noted that the storage server cluster of online plateform system can also be special in actual applications
Carry out networking and form database, i.e., be integrated with special database in online plateform system.
Management server disclosed in the present embodiment, specifically for safeguarding and managing online plateform system used by close
Connection rule.
User equipment disclosed in the present embodiment can specifically make an independent table apparatus in fact, or be integrated in various differences
Media data playing device in, such as set top box, mobile phone, tablet personal computer (Tablet Personal Computer),
Laptop computer (Laptop Computer), multimedia player, digital camera, personal digital assistant (personal
Digital assistant, abbreviation PDA), guider, mobile Internet access device (Mobile Internet Device, MID)
Or wearable device (Wearable Device) etc..
The embodiment of the present invention provides the commodity method for pushing in a kind of online transaction, as shown in Fig. 2 including:
S1, the order data for reading commodity pointed by pending correlation rule.
Wherein, the order data of commodity pointed by the pending correlation rule read by management server, at least
Including merchandise news, transaction record, sales figure etc., promotion record, logistics record etc. may also comprise.The correlation rule is used for
Represent at least two commodity that incidence relation mutually be present.Such as:Correlation rule can be expressed as<g,f,bab_gf>, wherein g
It is main commodity g and from commodity f respectively with f, bab_gf is the interest-degree of this rule, wherein, regular interest-degree is used to reflect
User have received pushed merchandise news by the rule, and have purchased the situation of pushed commodity, the business pushed
The numerical value of the quantity that product are successfully bought more at most interest-degree is higher.Again for example:Correlation rule can be expressed as<g,h,vab_gh
>, wherein g and h are main commodity g and from commodity h respectively, and vab_gh is the interest-degree of this rule, wherein the quantity from commodity h can
Think it is multiple, such as:It is specially the acquired commodity set similar to main commodity from commodity h, g and h are with the same commodity of brand
Group, and h commodity are at most taken the commodity of specified quantity by vab_gh.
S2, the order data according to the pointed commodity, obtain the support and confidence level of the correlation rule.
Wherein, the support is used to represent conditional probability of the commodity in date granularity pointed by the correlation rule,
The confidence level is used to represent conditional probability of the commodity on sales volume pointed by the correlation rule.
S3, support and confidence level according to the correlation rule, filter invalid rule from the correlation rule.
In the present embodiment, management server for correlation rule screening and filtering, support based on correlation rule and
The confidence level of correlation rule, wherein, retain the correlation rule that the two indexs are both greater than respective threshold value, or can according to this two
The size of individual index retains the correlation rule of specified quantity.F1 values can be the mediation that support and confidence level are weighted calculating
Average value.So as to which support will be optimized simultaneously and confidence level this biobjective scheduling problem is converted into the monocular of single optimization F1 values
Mark optimization problem.
S4, according to the correlation rule by filtering, determine commodity to be recommended, and by the commodity to be recommended to user equipment
Push.
For example:In actual order, often occur after buying certain mobile phone, buy the confidence level of soy bean milk making machine than purchase
The confidence level of portable power source is also high, but in a past season, the order for buying soy bean milk making machine but concentrates on continuous 3 entirely
My god, it therefore can substantially conclude, buy this mobile phone and repurchased caused by soy bean milk making machine is due to promotion, that is, after having bought this mobile phone,
The commodity that with extremely low price purchase soy bean milk making machine this promotion strategy this two pieces correlation can be caused less strong show very high
Correlation.Comparatively speaking, after having bought this mobile phone, the rule of portable power source is bought, although its confidence level does not buy soya-bean milk
Machine is high, but the distribution of such order in time is very smoothly, not occur sudden, therefore is purchased again after buying this mobile phone
The real behavior that portable power source is user is bought, the rule has real value.
The quantity on order granularity actually accumulated relative to being currently based on is associated the mode of rule analysis, and the present invention is implemented
Commodity method for pushing in the online transaction that example provides, for the screening and filtering of correlation rule, the support based on correlation rule
With the confidence level of correlation rule, the F1 values that the mediation for being weighted calculating by support and confidence level obtains filter invalid rule
Then, support will be optimized simultaneously and confidence level this biobjective scheduling problem is converted into the single object optimizations of single optimization F1 values and asked
Topic.And from rule occur number of days granularity on come consider rule time stability, defined in number of days granularity rule
Support, the support and confidence level of the date granularity of correlation rule are specifically employed, has been filtered out due to nothing caused by promotion
Effect rule, so as to improve the hits for recommending the commodity of user, accuracy rate and recall rate.
In the present embodiment, according to the order data of the pointed commodity, the support of the correlation rule is obtained
Concrete mode, including:Management server determines the pointed commodity by simultaneously according to the order data of the pointed commodity
The aggregate-value of the place period of purchase, and, at the time of the pointed commodity are bought first simultaneously to current time when
Between section total value.Further according to the aggregate-value and the total value, the support of the correlation rule is obtained.
Wherein, the support of correlation rule can be expressed asWherein, described in day (AB) is represented
The aggregate-value for the place period that pointed commodity are bought simultaneously, days represent that the pointed commodity are bought simultaneously first
At the time of to current time period total value.So that two commodity of AB are bought simultaneously as an example, it is assumed that find AB in day.begin
Day occurs by user while bought first, then is the accumulative of regular A-B appearance in the regular grid DEM as day before yesterday day.now
The ratio between number of days and total number of days day.now-day.begin.In molecule for cut-off current date, AB
The characteristics bought simultaneously, denominator are total number of days from the date that discovery AB is bought first simultaneously to current date.
The actual time stability for having weighed rule of time support so defined.Obviously, good correlation rule should have higher
Time stability, i.e. the rule is distributed than more uniform in the granularity of date-time.
In traditional correlation rule, weigh whether the rule has statistics meaning using regular grid DEM threshold value
Justice.Support is weighed using the order numbers comprising principal and subordinate's commodity simultaneously divided by total orders, and the selection of its minimum threshold has
Certain subjectivity.For comprehensive Retail e-commerce enterprise, its daily order total amount changes very greatly, if individually considering two
The total orders of individual commodity, then the quantity on order between different principal and subordinate's commodity great difference occurs, this causes support minimum
The selection of threshold value is difficult to take into account this species diversity.Current scheme is main or the support for the restrictive rule that gets on from quantity on order granularity
Spend threshold value.But for comprehensive Retail e-commerce enterprise, the period is promoted at it, because user behavior is by promotional impact, user
Consumer behavior and do not have very strong interpretation.For example mobile phone and soy bean milk making machine in itself and do not have very strong correlation, but
It is possible to the promotion strategy that bull's machine send soy bean milk making machine occur, so that the order for buying mobile phone and soy bean milk making machine simultaneously significantly increases
Add, ultimately result in bull's machine and recommend the confidence level of this rule of soy bean milk making machine very high.But due to the influence of promotion, it is clear that this
Correlation rule actually when it is invalid.Such as:In the correlation rule for the existing standard mentioned in the present embodiment, for commodity
A, B confidence level is defined as follows:Wherein, molecule represents AB while is purchased order numbers, and denominator
Purchase A order numbers are represented, confidence level is a conditional probability, under conditions of purchase A, then buys B probability.
In the present embodiment, management server filters invalid rule by the F1 values, not from quantity on order granularity
The support threshold of restrictive rule is removed, but the time stability of rule is considered in the granularity of the number of days occurred from rule,
Regular grid DEM is defined in number of days granularity, specifically employs the support and confidence level of the date granularity of correlation rule, mistake
Filter due to invalid rule caused by promotion, such as:Order data based on mobile phone and soy bean milk making machine, pass through the offline survey of reality
Test result shows:Compared with the scheme before improvement, the accuracy rate lifting about 10% of scheme after improvement, recall rate lifting about 10-
15%, recommend hits increase about 10-15% so that the hits of recommendation, accuracy rate, recall rate these three key indexs obtain
It is obviously improved.
In the present embodiment, the concrete mode of commodity to be recommended is determined according to the correlation rule by filtering, including:It is determined that
Default recommendation bits number, and wait to push away according to the correlation rule by filtering, the corresponding default recommendation bits number of acquisition
Recommend commodity.Wherein, the support of same correlation rule is typically much deeper than its confidence level, in order to avoid F1 values are led by confidence level
Lead, when calculating F1 values, can give support larger weight coefficient.In a preferred approach, the default weighting ratio is more than
Equal to 40:1.
Specifically, default recommendation bits number could be arranged to 5,10 or 20, management server can be recommended by adjusting
The number of position improves the accuracy rate of commodity to be recommended and recall rate, such as:In default weighting than being equal to 40:1 weighting F1
In the case of value, test of the correlation rule from standard association rule under different recommendation bits numbers after the present embodiment screens
Comparative result.
Wherein, in test process, 10 are taken as to the minimum threshold of regular characteristics in support, in confidence level points
The minimum threshold of son is taken as 15.The minimum threshold of its molecule is expressed as rule_count during confidence calculations.Relative to not considering
The correlation rule of support, the scheme accuracy rate based on weighting F1 values lift about 20-23%, recall rate lifting about 2.6-5.6%.
Because weighting F1 values only change the sequence from commodity, therefore it does not influence on coverage rate.Also, bits number is fewer when recommending
When, the lifting for weighting the accuracy rate and recall rate of F1 value schemes is more notable.
In the present embodiment, also it is between a kind of further commodity according to pointed by interest-degree detects the correlation rule of offer
The no rule-based filtering mode that association be present, further to weigh the practical significance of correlation rule, filters invalid correlation rule.
Including:
According to the order data, the sales situation of commodity pointed by the correlation rule is determined.And according to the sale
Situation obtains interest-degree, and with the presence or absence of association between the commodity according to pointed by the interest-degree detects the correlation rule, and
The correlation rule of association between commodity pointed by reservation be present.
Wherein, the sales situation includes:The order number of commodity pointed by the correlation rule is have purchased simultaneously, is not had
The order number of commodity pointed by the correlation rule, all ordering without commodity pointed by the purchase correlation rule are bought simultaneously
Odd number mesh.
Idiographic flow therein includes:Management server obtains interest-degreeWherein, A, B pointed by the correlation rule
Two kinds of commodity, a is represented while be have purchased A and B order number, and b is represented to have bought A but do not bought B order number, and c represents to buy
B but do not buy A order number, d is represented without the order number for buying A and B.And according to determining the interest-degree
Degree of correlation situation between commodity pointed by correlation rule, and determine whether association be present according to the degree of correlation situation, it is described
Degree of correlation situation includes:When the interest-degree is more than zero, B purchase contributes positively to A sale, represents B and A sales situation
Positive correlation.When the interest-degree is less than zero, the B sale bought on the contrary to A has detrimental effect, represents B and A sales situation
It is negatively correlated.When the interest-degree is equal to zero, B purchase and A sale are completely independent, and correlation is not present in B and A independently of each other,
It is independent of each other.
The scheme of the practical significance of correlation rule is weighed based on lifting degree according to traditional, wherein:Lifting degree is determined
Justice isIn practical application, rule of the usual lifting degree more than 3 just has actual application value.But
It is that to calculate lifting degree, also to have sales volume difference between different commodity pair greatly difficult based on actual order data.Such as:Make
When being excavated with traditional correlation rule, the scheme based on support threshold and confidence threshold value is although some rule can be excavated
Then, but the rule excavated may not be of practical significance.For example:The probability that user buys milk is 0.3, excavates user's purchase
It is 0.2 to buy and buy the probability of milk after bread again, and 0.2 is a higher value for confidence level, it appears that to having bought bread
It is rational that user, which recommends milk,.But the unconditional probability 0.3 that milk is bought with user is compared, it is follow-up that user buys bread
Probability of the probability of continuous purchase milk on the contrary than unconditional purchase milk is also low, i.e. this behavior of purchase bread actually results in
The probability of purchase milk reduces.Therefore, having bought bread recommends this rule of milk to have no practical significance.
The practical significance of correlation rule is weighed in the present embodiment using interest-degree, is only used in the correlation rule of comparison with standard
A and b two dimensions of information has been arrived, has been weighed using interest-degree, make use of a, b, c, d four-dimensional information altogether, therefore interest-degree index can be more
Add and subtly portray the degree of correlation that A and B are bought simultaneously.Dug by taking hot item paper extraction as an example, using traditional correlation rule
During pick, paper extraction is appeared in many mining rules, and this is due in actual order, and paper extraction and many commodity really can be by simultaneously
Purchase, but in the index based on interest-degree, paper extraction is bought, made on the contrary as hot item, itself and other many commodity simultaneously
Paper extraction " selectivity " that shows is very poor, so as to cause paper extraction not high or even small as the regular interest-degree from commodity
In zero.Draw:Bought A repurchase B interest-degree it is high, the order numbers that not only to buy A and B simultaneously in actual order are more,
And the order numbers that A and other commodity are bought can not be too many simultaneously, limiting case is exactly that B was only purchased together with A, always
Bought simultaneously without and except commodity in addition to A, so, saying of could firmly believing very much bought repurchase after A B interest-degree it is certain
It is high.
In the present embodiment, a kind of scheme that classification analysis is carried out to correlation rule is also provided, including:
The period that the correlation rule occurs is determined, and determines the distribution situation in time of the correlation rule.
And the correlation rule of threshold value is more than according to the temporal distribution situation, filtering concentration degree.
Specifically, the distribution situation in time for determining the correlation rule, including:According to the correlation rule
The period of appearance, obtain the time span and number of days support of every correlation rule.Wherein, time span represents correlation rule
Occur first to the total duration for finally going out current moment, number of days support represents total number of days and time span that correlation rule occurs
Ratio.Such as:Specifically in rule digging, to appearing in each rule in training set, there is following three parameter:The rule
The first appearance phase, the last of the rule there is the date, total number of days of rule appearance.Relatively reasonable correlation rule is divided into
Two classes:The first kind is more steady in time, such as bull's casing after bull's machine, no time and season sensitiveness.Second class is then
Chocolate is bought with season and time sensitivity, such as fresh flower of having bought before and after Valentine's Day.
For previous rule-like, for a long time from the point of view of, its generation is distributed than more uniform in time.It is long to the second rule-like
From the point of view of time, the generation of the rule-like is more concentrated on the date.According to this feature, for every rule, can set as follows
Filtering rule:If the time span of a correlation rule is more than first threshold and number of days support is more than Second Threshold, protect
Stay this correlation rule.Such as:Management server calculates the time span per rule, i.e. the rule occurs to last first
There is total number of days day_diff during day, if the day_diff>30 days (first threshold), then the rule be likely to belong to
One rule-like.But day_diff is used only>30 judge not enough, also need to enter row constraint to the time stationarity of rule, define
The number of days support of rule for rule generation total number of days/day_diff, set here the minimum threshold of number of days support as
0.2 (Second Threshold).That is day_diff>30 and regular number of days support belong to the first rule-like more than 0.2 rule.To second
Rule-like, its day_diff value can not be constrained, then management server judges according to rule_days, wherein, rule_days tables
Show that rule from the appearance, the regular characteristics continuously occurs, because in view of filtering due to caused by promotion
Unreasonable rule, if the number of days that some rule occurs is big by 5, it is determined as the second rule-like.
In the practical application of the present embodiment, the business to be recommended according to determined by the correlation rule by filtering also occurs
Product number can not be fully filled with the situation of available recommendation position.Recommend position to make full use of, there is provided a kind of according to by filtering
Correlation rule, commodity to be recommended, and the specific method that the commodity to be recommended are pushed to user equipment are determined, including:
According to first correlation rule by filtering, it is determined that corresponding with main commodity to be recommended first from commodity,
And obtain the interest-degree of first correlation rule by filtering.And determine the second correlation rule pointed by with the main business
Similar commodity set corresponding to product, and obtain the interest-degree of second correlation rule.Pass through the first of filtering further according to described
The interest-degree of the interest-degree of correlation rule and second correlation rule obtains the interest-degree of the 3rd correlation rule, and according to institute
The 3rd correlation rule by filtering is stated, it is determined that corresponding with the similar commodity set to be recommended second from commodity.
By described first from commodity and described second from commodity fill recommend position, and by it is described recommendation position in commodity to
The user equipment push.Such as:For the first correlation rule<g,f,bab_gf>, wherein g and f are main commodity and from business respectively
Product, bab_gf are the interest-degrees of this rule;By g commodity, the second correlation rule is associated<g,h,vab_gh>, wherein g and h divide
It is not main commodity and from commodity h, vab_gh is the interest-degree of this rule.Similar commodity h is obtained, limitation g and h is same with brand
Commodity group, and h commodity are at most taken by vab_gh 20 before ranking;By similar commodity h, the 3rd correlation rule is obtained<h,m,bab_
hm>, dependent merchandise m is obtained, then<g,m,vab_gh*bab_hm>It is regular as completion has been bought.If paying attention to given g and m,
Multiple h be present, then h different in vab_gh*bab_hm is summed to obtain completion rule<h,m,bab_hm>Confidence
Spend bab_hm.It is achieved thereby that for the mutually cascade between different correlation rules so that recommend commodity and the master of position for completion
Correlation between commodity is guaranteed, and commodity number and recall rate are hit after further improving completion.
The embodiment of the present invention also provides the commodity pusher in a kind of online transaction, specifically may operate in such as Fig. 1 institutes
In the management server shown, the device includes as shown in Figure 3:
Data read module, for reading the order data of commodity pointed by pending correlation rule, the association rule
Then it is used to represent at least two commodity that incidence relation mutually be present.
Analysis module, for the order data according to the pointed commodity, obtain the correlation rule support and
Confidence level, the support is used to represent conditional probability of the commodity in date granularity pointed by the correlation rule, described to put
Reliability is used to represent conditional probability of the commodity on sales volume pointed by the correlation rule.And according to the support of the correlation rule
Degree and confidence level, filter invalid rule from the correlation rule.
Pushing module, for according to the correlation rule by filtering, determining commodity to be recommended, and by the commodity to be recommended
Pushed to user equipment.
In the present embodiment, the analysis module, specifically for the order data according to the pointed commodity, institute is determined
The aggregate-value for the place period that pointed commodity are bought simultaneously is stated, and, what the pointed commodity were bought simultaneously first
The total value of the period at moment to current time.And according to the aggregate-value and the total value, obtain the branch of the correlation rule
Degree of holding.
The analysis module, it is additionally operable to, according to the order data, determine the sale of commodity pointed by the correlation rule
Situation, the sales situation include:The order number of commodity pointed by the correlation rule is have purchased simultaneously, is not bought simultaneously
The order number of commodity pointed by the correlation rule, all without the order number for buying commodity pointed by the correlation rule.
And interest-degree is obtained according to the sales situation, and be between the commodity according to pointed by the interest-degree detects the correlation rule
It is no association to be present, and the correlation rule of association between commodity pointed by reservation be present.
The pushing module, specifically for determining default recommendation bits number, and according to the correlation rule by filtering, obtain
Take the corresponding default commodity to be recommended for recommending bits number.Wherein, the default weighting is than being more than or equal to 40:1.
The pushing module, it is additionally operable to according to first correlation rule by filtering, it is determined that corresponding with main commodity
To be recommended first obtains the interest-degree of first correlation rule by filtering from commodity.And determine the second association rule
Then pointed similar commodity set corresponding with the main commodity, and obtain the interest-degree of second correlation rule.Root again
Obtain the 3rd associating according to interest-degree and the interest-degree of second correlation rule of first correlation rule by filtering
The interest-degree of rule, and according to the 3rd correlation rule by filtering, it is determined that corresponding with the similar commodity set
To be recommended second from commodity.Filled afterwards by described first from commodity and described second from commodity and recommend position, and by institute
State and recommend the commodity in position to be pushed to the user equipment.
The quantity on order granularity actually accumulated relative to being currently based on is associated the mode of rule analysis, and the present invention is implemented
Commodity pusher in the online transaction that example provides, for the screening and filtering of correlation rule, the support based on correlation rule
With the confidence level of correlation rule, the F1 values that the mediation for being weighted calculating by support and confidence level obtains filter invalid rule
Then, support will be optimized simultaneously and confidence level this biobjective scheduling problem is converted into the single object optimizations of single optimization F1 values and asked
Topic.And from rule occur number of days granularity on come consider rule time stability, defined in number of days granularity rule
Support, the support and confidence level of the date granularity of correlation rule are specifically employed, has been filtered out due to nothing caused by promotion
Effect rule, so as to improve the hits for recommending the commodity of user, accuracy rate and recall rate.
Each embodiment in this specification is described by the way of progressive, identical similar portion between each embodiment
Divide mutually referring to what each embodiment stressed is the difference with other embodiment.It is real especially for equipment
For applying example, because it is substantially similar to embodiment of the method, so describing fairly simple, related part is referring to embodiment of the method
Part explanation.The foregoing is only a specific embodiment of the invention, but protection scope of the present invention is not limited to
This, any one skilled in the art the invention discloses technical scope in, the change that can readily occur in or replace
Change, should all be included within the scope of the present invention.Therefore, protection scope of the present invention should be with the protection model of claim
Enclose and be defined.
Claims (10)
- A kind of 1. commodity method for pushing in online transaction, it is characterised in that including:The order data of commodity pointed by pending correlation rule is read, the correlation rule is used to represent that at least two is mutual The commodity of incidence relation be present;According to the order data of the pointed commodity, the support and confidence level of the correlation rule, the support are obtained For representing conditional probability of the commodity in date granularity pointed by the correlation rule, the confidence level is used to represent the pass Conditional probability of the commodity on sales volume pointed by connection rule;According to the support and confidence level of the correlation rule, invalid rule is filtered from the correlation rule;According to the correlation rule by filtering, commodity to be recommended are determined, and the commodity to be recommended are pushed to user equipment.
- 2. according to the method for claim 1, it is characterised in that the correlation rule according to by filtering, it is determined that waiting to push away Commodity are recommended, including:Default recommendation bits number is determined, and according to the correlation rule by filtering, is obtained corresponding described default Recommend the commodity to be recommended of bits number;The default weighting is than being more than or equal to 40:1.
- 3. according to the method for claim 1, it is characterised in that the order data according to the pointed commodity, obtain The support of the correlation rule is taken, including:According to the order data of the pointed commodity, the tired of the place period that the pointed commodity are bought simultaneously is determined Evaluation, and, at the time of the pointed commodity are bought first simultaneously to current time period total value;According to the aggregate-value and the total value, the support of the correlation rule is obtained.
- 4. according to the method for claim 1, it is characterised in that also include:According to the order data, the sales situation of commodity pointed by the correlation rule is determined, the sales situation includes:Together When have purchased the order numbers of commodity pointed by the correlation rule, do not buy commodity pointed by the correlation rule simultaneously Order number, all without the order number for buying commodity pointed by the correlation rule;Interest-degree is obtained according to the sales situation, and between the commodity according to pointed by the interest-degree detects the correlation rule The correlation rule of association with the presence or absence of association, and between commodity pointed by reservation be present.
- 5. according to the method for claim 4, it is characterised in that described that interest-degree, and root are obtained according to the sales situation It whether there is association between detecting commodity pointed by the correlation rule according to the interest-degree, including:Obtain interest-degreeWherein, the association Two kinds of commodity of A, B pointed by rule, a is represented while be have purchased A and B order number, and b is represented to have bought A but do not bought B order Number, c are represented to have bought B but are not bought A order number, and d is represented without purchase A and B order number;Degree of correlation situation between the commodity according to pointed by the interest-degree determines the correlation rule, and according to the degree of correlation Situation determines whether association be present, and the degree of correlation situation includes:When the interest-degree is more than zero, B and A sales situation is represented Positive correlation;When the interest-degree is less than zero, represent that B and A sales situation is negatively correlated;When the interest-degree is equal to zero, B and A phases It is mutually independent that correlation is not present.
- 6. according to the method for claim 1, it is characterised in that including:The period that the correlation rule occurs is determined, and determines the distribution situation in time of the correlation rule;According to the temporal distribution situation, filtering concentration degree is more than the correlation rule of threshold value.
- 7. according to the method for claim 6, it is characterised in that the distribution in time for determining the correlation rule Situation, including:The period occurred according to the correlation rule, the time span and number of days support of every correlation rule are obtained, wherein, Time span represents that correlation rule occurs representing correlation rule to the total duration for finally going out current moment, number of days support first Total number of days and time span ratio;If the time span of a correlation rule is more than first threshold and number of days support is more than Second Threshold, retain this Correlation rule.
- 8. according to the method for claim 1, it is characterised in that the correlation rule according to by filtering, it is determined that waiting to push away Commodity are recommended, and the commodity to be recommended are pushed to user equipment, including:According to first correlation rule by filtering, it is determined that corresponding with main commodity to be recommended first from commodity, and obtain Take the interest-degree of first correlation rule by filtering;The similar commodity set corresponding with the main commodity pointed by the second correlation rule is determined, and obtains second association The interest-degree of rule;The is obtained according to the interest-degree of the interest-degree of first correlation rule by filtering and second correlation rule The interest-degree of three correlation rules, and according to it is described by filtering the 3rd correlation rule, it is determined that to the similar commodity collection To be recommended second from commodity corresponding to conjunction;Filled by described first from commodity and described second from commodity and recommend position, and by the commodity in the recommendation position to described User equipment pushes.
- A kind of 9. commodity pusher in online transaction, it is characterised in that including:Data read module, for reading the order data of commodity pointed by pending correlation rule, the correlation rule is used The commodity of incidence relation mutually be present in expression at least two;Analysis module, for the order data according to the pointed commodity, obtain the support and confidence of the correlation rule Degree, the support are used to represent conditional probability of the commodity in date granularity pointed by the correlation rule, the confidence level For representing conditional probability of the commodity on sales volume pointed by the correlation rule;And according to the support of the correlation rule and Confidence level, invalid rule is filtered from the correlation rule;Pushing module, for according to the correlation rule by filtering, determining commodity to be recommended, and by the commodity to be recommended to Family equipment push.
- 10. device according to claim 9, it is characterised in that the analysis module, specifically for according to described pointed The order data of commodity, the aggregate-value for the place period that the pointed commodity are bought simultaneously is determined, and, it is described pointed At the time of commodity are bought first simultaneously to current time period total value;And according to the aggregate-value and the total value, Obtain the support of the correlation rule;The analysis module, it is additionally operable to, according to the order data, determine the sales situation of commodity pointed by the correlation rule, The sales situation includes:The order number of commodity pointed by the correlation rule is have purchased simultaneously, without described in purchase simultaneously The order number of commodity pointed by correlation rule, all without the order number for buying commodity pointed by the correlation rule;And root Interest-degree is obtained according to the sales situation, and whether is deposited between the commodity according to pointed by the interest-degree detects the correlation rule Associating, and the correlation rule of association between commodity pointed by reservation be present;The pushing module, specifically for determining default recommendation bits number, and according to the correlation rule by filtering, acquisition pair Answer the default commodity to be recommended for recommending bits number;Wherein, the default weighting is than being more than or equal to 40:1;The pushing module, it is additionally operable to according to first correlation rule by filtering, it is determined that corresponding with main commodity wait to push away First recommended obtains the interest-degree of first correlation rule by filtering from commodity;And determine the second correlation rule institute The similar commodity set corresponding with the main commodity pointed to, and obtain the interest-degree of second correlation rule;Further according to institute That states the interest-degree of the first correlation rule by filtering and the interest-degree of second correlation rule obtains the 3rd correlation rule Interest-degree, and according to it is described by filtering the 3rd correlation rule, it is determined that corresponding with the similar commodity set treat Second recommended is from commodity;Filled afterwards by described first from commodity and described second from commodity and recommend position, and pushed away described The commodity recommended in position push to the user equipment.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610662922.8A CN107730336A (en) | 2016-08-12 | 2016-08-12 | Commodity method for pushing and device in a kind of online transaction |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610662922.8A CN107730336A (en) | 2016-08-12 | 2016-08-12 | Commodity method for pushing and device in a kind of online transaction |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107730336A true CN107730336A (en) | 2018-02-23 |
Family
ID=61200915
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610662922.8A Pending CN107730336A (en) | 2016-08-12 | 2016-08-12 | Commodity method for pushing and device in a kind of online transaction |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107730336A (en) |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108806767A (en) * | 2018-06-15 | 2018-11-13 | 中南大学 | Disease symptoms association analysis method based on electronic health record |
CN108898459A (en) * | 2018-06-25 | 2018-11-27 | 中国联合网络通信集团有限公司 | A kind of Method of Commodity Recommendation and device |
CN110197390A (en) * | 2019-04-09 | 2019-09-03 | 深圳市梦网百科信息技术有限公司 | A kind of recommended method and system based on the correlation rule degree of association and economic value |
CN110852863A (en) * | 2019-11-15 | 2020-02-28 | 安徽海汇金融投资集团有限公司 | Receivable account transfer recommendation method and system based on correlation analysis |
CN111652631A (en) * | 2020-06-09 | 2020-09-11 | 创新奇智(成都)科技有限公司 | Commodity recall method and device, electronic equipment and computer-readable storage medium |
CN111767466A (en) * | 2020-09-01 | 2020-10-13 | 腾讯科技(深圳)有限公司 | Recommendation information recommendation method and device based on artificial intelligence and electronic equipment |
CN112734511A (en) * | 2020-12-30 | 2021-04-30 | 昆山巨星行动电子商务有限公司 | Commodity search recommendation method based on e-commerce system and application |
CN115063187A (en) * | 2022-08-05 | 2022-09-16 | 广东省轻工业技师学院 | Electronic commerce data processing method, system, electronic device and medium |
CN115878903A (en) * | 2023-02-21 | 2023-03-31 | 万链指数(青岛)信息科技有限公司 | Intelligent information recommendation method based on big data |
CN116402438A (en) * | 2023-03-29 | 2023-07-07 | 乐麦信息技术(杭州)有限公司 | Warehouse management method and system |
CN116662673A (en) * | 2023-07-28 | 2023-08-29 | 西安银信博锐信息科技有限公司 | User preference data analysis method based on data monitoring |
Citations (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101082972A (en) * | 2007-05-30 | 2007-12-05 | 华为技术有限公司 | Method and device for forecasting user's interest to commercial product and method for publishing advertisement thereof |
CN101344902A (en) * | 2008-07-15 | 2009-01-14 | 北京科技大学 | Secondary protein structure forecasting technique based on association analysis and association classification |
CN101515360A (en) * | 2009-04-13 | 2009-08-26 | 阿里巴巴集团控股有限公司 | Method and server for recommending network object information to user |
US7720720B1 (en) * | 2004-08-05 | 2010-05-18 | Versata Development Group, Inc. | System and method for generating effective recommendations |
CN102402757A (en) * | 2010-09-15 | 2012-04-04 | 阿里巴巴集团控股有限公司 | Method and device for providing information, and method and device for determining comprehensive relevance |
CN103345698A (en) * | 2013-07-09 | 2013-10-09 | 焦点科技股份有限公司 | Personalized recommendation method based on cloud processing mode and applied in e-business environment |
CN103605815A (en) * | 2013-12-11 | 2014-02-26 | 焦点科技股份有限公司 | Automatic commodity information classifying and recommending method applicable to B2B (Business to Business) e-commerce platform |
CN103631801A (en) * | 2012-08-23 | 2014-03-12 | 阿里巴巴集团控股有限公司 | Method and device for providing commodity information |
CN103700005A (en) * | 2013-12-17 | 2014-04-02 | 南京信息工程大学 | Association-rule recommending method based on self-adaptive multiple minimum supports |
CN103823908A (en) * | 2014-03-21 | 2014-05-28 | 北京飞流九天科技有限公司 | Method and server for content recommendation on basis of user preferences |
CN103839167A (en) * | 2012-11-21 | 2014-06-04 | 大连灵动科技发展有限公司 | Commodity candidate set recommendation method |
CN104317945A (en) * | 2014-10-31 | 2015-01-28 | 亚信科技(南京)有限公司 | E-commerce website commodity recommending method on basis of search behaviors |
CN104573108A (en) * | 2015-01-30 | 2015-04-29 | 联想(北京)有限公司 | Information processing method and information processing unit |
CN104599160A (en) * | 2015-02-06 | 2015-05-06 | 腾讯科技(深圳)有限公司 | Commodity recommendation method and commodity recommendation device |
CN105321089A (en) * | 2014-07-16 | 2016-02-10 | 苏宁云商集团股份有限公司 | Method and system for e-commerce recommendation based on multi-algorithm fusion |
CN105608602A (en) * | 2015-12-18 | 2016-05-25 | 齐鲁工业大学 | Application of effective negative association rule selected through quadratic correlation judging method to analysis of customer buying behaviors |
CN105678578A (en) * | 2016-01-05 | 2016-06-15 | 重庆邮电大学 | Method for measuring user brand preference on the basis of online shopping behavior data |
-
2016
- 2016-08-12 CN CN201610662922.8A patent/CN107730336A/en active Pending
Patent Citations (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7720720B1 (en) * | 2004-08-05 | 2010-05-18 | Versata Development Group, Inc. | System and method for generating effective recommendations |
CN101082972A (en) * | 2007-05-30 | 2007-12-05 | 华为技术有限公司 | Method and device for forecasting user's interest to commercial product and method for publishing advertisement thereof |
CN101344902A (en) * | 2008-07-15 | 2009-01-14 | 北京科技大学 | Secondary protein structure forecasting technique based on association analysis and association classification |
CN101515360A (en) * | 2009-04-13 | 2009-08-26 | 阿里巴巴集团控股有限公司 | Method and server for recommending network object information to user |
CN102402757A (en) * | 2010-09-15 | 2012-04-04 | 阿里巴巴集团控股有限公司 | Method and device for providing information, and method and device for determining comprehensive relevance |
CN103631801A (en) * | 2012-08-23 | 2014-03-12 | 阿里巴巴集团控股有限公司 | Method and device for providing commodity information |
CN103839167A (en) * | 2012-11-21 | 2014-06-04 | 大连灵动科技发展有限公司 | Commodity candidate set recommendation method |
CN103345698A (en) * | 2013-07-09 | 2013-10-09 | 焦点科技股份有限公司 | Personalized recommendation method based on cloud processing mode and applied in e-business environment |
CN103605815A (en) * | 2013-12-11 | 2014-02-26 | 焦点科技股份有限公司 | Automatic commodity information classifying and recommending method applicable to B2B (Business to Business) e-commerce platform |
CN103700005A (en) * | 2013-12-17 | 2014-04-02 | 南京信息工程大学 | Association-rule recommending method based on self-adaptive multiple minimum supports |
CN103823908A (en) * | 2014-03-21 | 2014-05-28 | 北京飞流九天科技有限公司 | Method and server for content recommendation on basis of user preferences |
CN105321089A (en) * | 2014-07-16 | 2016-02-10 | 苏宁云商集团股份有限公司 | Method and system for e-commerce recommendation based on multi-algorithm fusion |
CN104317945A (en) * | 2014-10-31 | 2015-01-28 | 亚信科技(南京)有限公司 | E-commerce website commodity recommending method on basis of search behaviors |
CN104573108A (en) * | 2015-01-30 | 2015-04-29 | 联想(北京)有限公司 | Information processing method and information processing unit |
CN104599160A (en) * | 2015-02-06 | 2015-05-06 | 腾讯科技(深圳)有限公司 | Commodity recommendation method and commodity recommendation device |
CN105608602A (en) * | 2015-12-18 | 2016-05-25 | 齐鲁工业大学 | Application of effective negative association rule selected through quadratic correlation judging method to analysis of customer buying behaviors |
CN105678578A (en) * | 2016-01-05 | 2016-06-15 | 重庆邮电大学 | Method for measuring user brand preference on the basis of online shopping behavior data |
Non-Patent Citations (6)
Title |
---|
张同启: "基于关联规则和用户喜好程度的综合电子商务推荐系统的研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
张玉芳 等: "基于兴趣度含正负项目的关联规则挖掘方法", 《电子科技大学学报》 * |
王枭翔: "基于相关兴趣度的关联规则挖掘", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
王璇: "一种带兴趣度的正负关联规则挖掘算法", 《常州工学院学报》 * |
靳峰: "基于兴趣度的推荐系统研究", 《中国优秀博硕士学位论文全文数据库 (硕士)社会科学Ⅰ辑(经济政治与法律)》 * |
韦素云 等: "基于项目类别和兴趣度的协同过滤推荐算法", 《南京大学学报(自然科学)》 * |
Cited By (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108806767B (en) * | 2018-06-15 | 2021-10-22 | 中南大学 | Disease symptom correlation analysis method based on electronic medical record |
CN108806767A (en) * | 2018-06-15 | 2018-11-13 | 中南大学 | Disease symptoms association analysis method based on electronic health record |
CN108898459A (en) * | 2018-06-25 | 2018-11-27 | 中国联合网络通信集团有限公司 | A kind of Method of Commodity Recommendation and device |
CN108898459B (en) * | 2018-06-25 | 2020-12-01 | 中国联合网络通信集团有限公司 | Commodity recommendation method and device |
CN110197390A (en) * | 2019-04-09 | 2019-09-03 | 深圳市梦网百科信息技术有限公司 | A kind of recommended method and system based on the correlation rule degree of association and economic value |
CN110197390B (en) * | 2019-04-09 | 2024-01-05 | 深圳市梦网视讯有限公司 | Recommendation method and system based on association degree and economic value of association rule |
CN110852863A (en) * | 2019-11-15 | 2020-02-28 | 安徽海汇金融投资集团有限公司 | Receivable account transfer recommendation method and system based on correlation analysis |
CN111652631A (en) * | 2020-06-09 | 2020-09-11 | 创新奇智(成都)科技有限公司 | Commodity recall method and device, electronic equipment and computer-readable storage medium |
CN111652631B (en) * | 2020-06-09 | 2023-06-06 | 创新奇智(成都)科技有限公司 | Commodity recall method, commodity recall device, electronic equipment and computer readable storage medium |
CN111767466A (en) * | 2020-09-01 | 2020-10-13 | 腾讯科技(深圳)有限公司 | Recommendation information recommendation method and device based on artificial intelligence and electronic equipment |
CN112734511A (en) * | 2020-12-30 | 2021-04-30 | 昆山巨星行动电子商务有限公司 | Commodity search recommendation method based on e-commerce system and application |
CN115063187A (en) * | 2022-08-05 | 2022-09-16 | 广东省轻工业技师学院 | Electronic commerce data processing method, system, electronic device and medium |
CN115878903A (en) * | 2023-02-21 | 2023-03-31 | 万链指数(青岛)信息科技有限公司 | Intelligent information recommendation method based on big data |
CN115878903B (en) * | 2023-02-21 | 2023-05-09 | 万链指数(青岛)信息科技有限公司 | Information intelligent recommendation method based on big data |
CN116402438A (en) * | 2023-03-29 | 2023-07-07 | 乐麦信息技术(杭州)有限公司 | Warehouse management method and system |
CN116662673A (en) * | 2023-07-28 | 2023-08-29 | 西安银信博锐信息科技有限公司 | User preference data analysis method based on data monitoring |
CN116662673B (en) * | 2023-07-28 | 2023-11-03 | 西安银信博锐信息科技有限公司 | User preference data analysis method based on data monitoring |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107730336A (en) | Commodity method for pushing and device in a kind of online transaction | |
JP6356744B2 (en) | Method and system for displaying cross-website information | |
US10423999B1 (en) | Performing personalized category-based product sorting | |
CN101324948B (en) | A kind of method of information recommendation and device | |
CN109582857A (en) | Based on big data information-pushing method, device, computer equipment and storage medium | |
CN109658206A (en) | Information recommendation method and device | |
US20110010307A1 (en) | Method and system for recommending articles and products | |
CN109447713A (en) | A kind of recommended method and device of knowledge based map | |
CN103092861B (en) | A kind of choosing method of commodity representative picture and system | |
CN103886001A (en) | Personalized commodity recommendation system | |
US20190005572A1 (en) | Method and system of building store product finders | |
CN106294420B (en) | The method and device of business object collocation information is provided | |
US8738543B2 (en) | Business intelligence based social network with virtual data-visualization cards | |
KR20140056731A (en) | Purchase recommendation service system and method | |
CN103034660B (en) | Information providing method, Apparatus and system | |
CN110363604A (en) | Page generation method and device | |
CN107230136A (en) | A kind of shopping sequence method for pushing based on big data | |
CN102629354A (en) | System and method for automatically accounting consumption data (records) of consumer | |
CN108073667A (en) | Method for generating user browsing attributes, and non-transitory computer readable medium | |
CN106776697A (en) | Content recommendation method and device | |
CN107578180A (en) | Intellectualized analysis platform and its system based on ETL | |
CN115496566A (en) | Regional specialty recommendation method and system based on big data | |
CN109190027A (en) | Multi-source recommended method, terminal, server, computer equipment, readable medium | |
CN108537635A (en) | A kind of recommendation method and device of product | |
CN103577473A (en) | Classification and disambiguation method, classification and disambiguation device and system thereof |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180223 |
|
RJ01 | Rejection of invention patent application after publication |