CN109961307A - The appraisal procedure and device of object-oriented - Google Patents

The appraisal procedure and device of object-oriented Download PDF

Info

Publication number
CN109961307A
CN109961307A CN201711418565.1A CN201711418565A CN109961307A CN 109961307 A CN109961307 A CN 109961307A CN 201711418565 A CN201711418565 A CN 201711418565A CN 109961307 A CN109961307 A CN 109961307A
Authority
CN
China
Prior art keywords
article
indicate
order
user
goods attribute
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
Application number
CN201711418565.1A
Other languages
Chinese (zh)
Inventor
聂瑞
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Jingdong Century Trading Co Ltd
Beijing Jingdong Shangke Information Technology Co Ltd
Original Assignee
Beijing Jingdong Century Trading Co Ltd
Beijing Jingdong Shangke Information Technology Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Beijing Jingdong Century Trading Co Ltd, Beijing Jingdong Shangke Information Technology Co Ltd filed Critical Beijing Jingdong Century Trading Co Ltd
Priority to CN201711418565.1A priority Critical patent/CN109961307A/en
Publication of CN109961307A publication Critical patent/CN109961307A/en
Pending legal-status Critical Current

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
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0201—Market modelling; Market analysis; Collecting market data
    • 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]
    • G06Q30/0623—Item investigation
    • G06Q30/0625—Directed, with specific intent or strategy
    • 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]
    • G06Q30/0633—Lists, e.g. purchase orders, compilation or processing
    • G06Q30/0635—Processing of requisition or of purchase orders

Landscapes

  • Business, Economics & Management (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Strategic Management (AREA)
  • Development Economics (AREA)
  • Engineering & Computer Science (AREA)
  • General Business, Economics & Management (AREA)
  • Physics & Mathematics (AREA)
  • Marketing (AREA)
  • General Physics & Mathematics (AREA)
  • Economics (AREA)
  • Theoretical Computer Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Data Mining & Analysis (AREA)
  • Game Theory and Decision Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses a kind of appraisal procedure of object-oriented and devices, are related to field of computer technology.This method comprises: statistics is with category various article by proxy indicator data；According to the training set training machine learning model by proxy indicator data and goods attribute data for including the same category various article, learn to by proxy indicator, to obtain goods attribute to the impact factor of user's decision behavior；Item Value is calculated according to impact factor of the goods attribute to user's decision behavior.By above step, article can be quantified to the value of businessman, article is accurately reflected by influencing user's decision behavior bring shop income, consequently facilitating businessman formulates reasonable selection scheme.

Description

The appraisal procedure and device of object-oriented
Technical field
The present invention relates to field of computer technology more particularly to the appraisal procedures and device of a kind of object-oriented.
Background technique
As type of merchandize is increasingly various, how to design and not only met target consumer's demand but also kept self benefits maximum The commodity scheme on sale changed is to perplex the long standing difficulty of businessman.
In the prior art, commodity Zhen mainly is carried out using the intuition of traditional data statistical approach and business personnel Choosing.In general, business personnel can directly judge the quality of commodity, Yi Jishi in conjunction with the data such as sales volume, profit and business intuition No upper undercarriage.
In realizing process of the present invention, at least there are the following problems in the prior art for inventor's discovery: in the prior art, industry Business personnel by business intuition select commodity method it is more coarse, can not scientific, visual quantization commodity for businessman's Value can not also accurately reflect commodity by influencing user's decision behavior bring shop income.
Summary of the invention
In view of this, the present invention provides the appraisal procedure and device of a kind of object-oriented, article can be quantified to businessman's Value accurately reflects article by influencing user's decision behavior bring shop income, consequently facilitating businessman formulates reasonable choosing Product scheme.
To achieve the above object, according to an aspect of the invention, there is provided a kind of appraisal procedure of object-oriented.
The appraisal procedure of object-oriented of the invention includes: to count with category various article by proxy indicator data； According to the training set training machine by proxy indicator data and goods attribute data for including the same category various article Learning model learns to by proxy indicator, to obtain goods attribute to the impact factor of user's decision behavior；According to institute It states goods attribute and Item Value is calculated to the impact factor of user's decision behavior.
Optionally, the machine learning model includes: Random Forest model.
Optionally, the basis includes the same category various article by proxy indicator data and goods attribute number According to training set training machine learning model, learn to by proxy indicator, to obtain goods attribute to user's decision row For impact factor the step of include: that more decision trees are constructed based on least square regression tree algorithm, include described more to obtain The prediction model of decision tree；Randomization is carried out to an attribute dimensions of goods attribute sample point, and will be at randomization Data after reason input the prediction model, to obtain by proxy indicator predicted value；It calculates by proxy indicator predicted value and quilt Offset distance between proxy indicator true value, and using the offset distance as the goods attribute to the shadow of user's decision behavior Ring the factor.
Optionally, the method also includes: based on least square regression tree algorithm building decision tree when, according to anti-distance The output valve of weighting function calculating decision tree；
Wherein, cmIndicate output valve；J indicates goods attribute used in cutting；S indicates cut-off；It indicates new The article attribute data to be measured and region R of inputmThe inverse of the Euclidean distance between goods attribute sample point in (j, s)；xi、 xkIndicate region RmGoods attribute sample point in (j, s)；yiIt indicates and goods attribute sample point xiIt is corresponding by alternative finger Mark sample value；Expression pairIt sums.
Optionally, the Item Value is calculated according to the following formula；
Wherein, PskuIndicate Item Value, KeyskuIndicate the key index of the article, γiIndicate i-th of goods attribute pair By the impact factor of proxy indicator, WiIt indicates with all items in category with i-th of goods attribute in the category Key index accounting, n indicate the attribute number of the article.
Optionally, calculate according to the following formula the article by proxy indicator:
SA=∑All commodity∑All orders paira*S(A,B)；
Wherein, SAIt is article A by proxy indicator；S (A, B) indicates an order based on same user to calculating To the index that is substituted by article B of article A, ∑All orders pairA*S (A, B) indicate to based on all orders to obtained S (A, B) into Row weighted sum, a are weight factor, ∑All commodityIndicate that article A is summed by the index that the every other article of same category substitutes.
Optionally, S (A, B) is calculated according to the following formula:
Wherein, Sales1AIndicate the sales volume of article A in order 1；GAIndicate the sale of article A from order 1 to order 2 Volume accounting penalty values；GBIt indicates from order 1 to order 2, the sales volume accounting value added of article B；It indicates from order 1 to ordering Single 2, with article A, B with the sales volume accounting value added of other articles of category；N is the same category described in order 1, order 2 Other articles species number；S1For the sales volume of order 1；S2For the sales volume of order 2.
Optionally, the method also includes: the theil indexes of user are calculated according to the following formula, and according to the user's Theil indexes optimize S (A, B), with according to the S (A, B) after optimization calculate the article by proxy indicator:
Wherein, TIC,DIndicate that the theil indexes of user C, K indicate alternative object that user C is bought, article p in category D The species number of product, salesp,CIndicate the sales volume of the article p of user C purchase, avg salesD,CIndicate the category of user C purchase The sales volume mean value of all items, S'(A, B in D) indicate the result optimized to S (A, B).
To achieve the above object, according to another aspect of the present invention, a kind of assessment device of object-oriented is provided.
The assessment device of object-oriented of the invention includes: statistical module, for counting being replaced with category various article For property achievement data；Study module includes the same category various article by proxy indicator data and object for basis The training set training machine learning model of product attribute data, learns to by proxy indicator, to obtain goods attribute to quilt The impact factor of proxy indicator；Computing module, by according to the goods attribute to the impact factor by proxy indicator based on Calculate Item Value.
Optionally, the machine learning model in the study module includes: Random Forest model.
Optionally, the study module is according to including the same category various article by proxy indicator data and object The training set training machine learning model of product attribute data, learns to by proxy indicator, with obtain goods attribute to The operation of the impact factor of family decision behavior include: the study module be based on least square regression tree algorithm construct more decisions Tree, with obtain include the more decision trees prediction model；An attribute of the study module to goods attribute sample point Dimension carries out randomization, and the data after randomization are inputted the prediction model, pre- by proxy indicator to obtain Measured value；The study module is calculated by proxy indicator predicted value and by the offset distance between proxy indicator true value, and will The offset distance is as the goods attribute to the impact factor of user's decision behavior.
Optionally, the study module is added when based on least square regression tree algorithm building decision tree according to anti-distance The output valve of weight function calculating decision tree；
Wherein, cmIndicate output valve；J indicates goods attribute used in cutting；S indicates cut-off；It indicates new The article attribute data to be measured and region R of inputmThe inverse of the Euclidean distance between goods attribute sample point in (j, s)；xi、 xkIndicate region RmGoods attribute sample point in (j, s)；yiIt indicates and goods attribute sample point xiIt is corresponding by alternative finger Mark sample value；Expression pairIt sums.
Optionally, the computing module calculates the Item Value according to the following formula；
Wherein, PskuIndicate Item Value, KeyskuIndicate the key index of the article, γiIndicate i-th of goods attribute pair By the impact factor of proxy indicator, WiIt indicates with all items in category with i-th of goods attribute in the category Key index accounting, n indicate the attribute number of the article.
Optionally, the statistical module calculate according to the following formula the article by proxy indicator:
SA=∑All commodity∑All orders paira*S(A,B)；
Wherein, SAIt is article A by proxy indicator；S (A, B) indicates an order based on same user to calculating To the index that is substituted by article B of article A, ∑All orders pairA*S (A, B) indicate to based on all orders to obtained S (A, B) into Row weighted sum, a are weight factor, ∑All commodityIndicate that article A is summed by the index that the every other article of same category substitutes.
Optionally, the statistical module calculates S (A, B) according to the following formula:
Wherein, Sales1AIndicate the sales volume of article A in order 1；GAIndicate the sale of article A from order 1 to order 2 Volume accounting penalty values；GBIt indicates from order 1 to order 2, the sales volume accounting value added of article B；It indicates from order 1 to ordering Single 2, with article A, B with the sales volume accounting value added of other articles of category；N is the same category described in order 1, order 2 Other articles species number；S1For the sales volume of order 1；S2For the sales volume of order 2.
Optionally, described device further include: optimization module, for calculating the theil indexes of user according to the following formula, and S (A, B) is optimized according to the theil indexes of the user, so that the statistical module is calculated according to the S (A, B) after optimization The article by proxy indicator:
Wherein, TIC,DIndicate that the theil indexes of user C, K indicate alternative object that user C is bought, article p in category D The species number of product, salesp,CIndicate the sales volume of the article p of user C purchase, avg salesD,CIndicate the category of user C purchase The sales volume mean value of all items, S'(A, B in D) indicate the result optimized to S (A, B).
To achieve the above object, according to a further aspect of the invention, a kind of server is provided.
Server of the invention, comprising: one or more processors；And storage device, for storing one or more Program；When one or more of programs are executed by one or more of processors, so that one or more of processors Realize the appraisal procedure of object-oriented of the invention.
To achieve the above object, according to a further aspect of the invention, a kind of computer-readable medium is provided.
Computer-readable medium of the invention is stored thereon with computer program, real when described program is executed by processor The appraisal procedure of existing object-oriented of the invention.
One embodiment in foregoing invention has the following advantages that or the utility model has the advantages that by statistics article by alternative finger Data are marked, are assembled for training according to the training by proxy indicator data and goods attribute data for including the same category various article Practice machine learning model, learns to by proxy indicator, to obtain goods attribute to the impact factor of user's decision behavior, And Item Value is calculated according to impact factor of the goods attribute to user's decision behavior, article can be quantified to the valence of businessman Value accurately reflects article by influencing user's decision behavior bring shop income, consequently facilitating businessman's rapid development is reasonable Selection scheme.
Further effect possessed by above-mentioned non-usual optional way adds hereinafter in conjunction with specific embodiment With explanation.
Detailed description of the invention
Attached drawing for a better understanding of the present invention, does not constitute an undue limitation on the present invention.Wherein:
Fig. 1 is the key step schematic diagram of the appraisal procedure of object-oriented according to an embodiment of the invention；
Fig. 2 is the key step schematic diagram of the appraisal procedure of object-oriented according to another embodiment of the present invention；
Fig. 3 is the main modular schematic diagram of the assessment device of object-oriented according to an embodiment of the invention；
Fig. 4 is the main modular schematic diagram of the assessment device of object-oriented according to another embodiment of the present invention；
Fig. 5 is the main modular schematic diagram of selection system according to an embodiment of the present invention；
Fig. 6 is that the assessment device of the object-oriented of the embodiment of the present invention can be applied to exemplary system architecture therein Figure；
Fig. 7 is adapted for the structural schematic diagram for the computer system for realizing the server of the embodiment of the present invention.
Specific embodiment
Below in conjunction with attached drawing, an exemplary embodiment of the present invention will be described, including the various of the embodiment of the present invention Details should think them only exemplary to help understanding.Therefore, those of ordinary skill in the art should recognize It arrives, it can be with various changes and modifications are made to the embodiments described herein, without departing from scope and spirit of the present invention.Together Sample, for clarity and conciseness, descriptions of well-known functions and structures are omitted from the following description.
It should be pointed out that in the absence of conflict, the feature in embodiment and embodiment in the present invention can be with It is combined with each other.
Fig. 1 is the key step schematic diagram of the appraisal procedure of object-oriented according to an embodiment of the invention.Such as Fig. 1 institute Show, the appraisal procedure of the object-oriented of the embodiment of the present invention further include:
Step S101, statistics is with category various article by proxy indicator data.
Wherein, the article includes commodity, described to be used to measure article to the shadow of user's decision behavior by proxy indicator Ring power.Illustratively, for this category of soda, can count respectively all commodity under the category (for example, Coca-Cola, Pepsi Cola etc.) by proxy indicator.
In general, each article usually has multiple goods attributes, each goods attribute can to the influence power of user's decision behavior It can be different.For example, for the article under this category of soda, with article packaging, odor type, taste, specification, net content And other items attribute, the influence power that each goods attribute buys this decision behavior of soda for user may be different.In order to measure Change each goods attribute to the influence power of user's decision behavior, recurrence can be carried out according to proxy indicator of the step S102 to article It practises.
Step S102, basis includes the same category various article by proxy indicator data and goods attribute data Training set training machine learning model, learn to by proxy indicator, to obtain goods attribute to user's decision behavior Impact factor.
Wherein, the goods attribute data are made of the goods attribute value of multiple dimensions.The machine learning model can wrap It includes: Random Forest model or GBDT (gradient promotes decision tree) model.
Further, before step S102, the method for the embodiment of the present invention may also include that locates goods attribute value in advance Reason, to obtain goods attribute data.Specifically, when goods attribute value is continuous data, minimum value-maximum value can be used Normalization (Min-Max scaling) method is handled；When goods attribute value is discrete data, feature two-value can be used Change method is handled.
Step S103, Item Value is calculated according to impact factor of the goods attribute to user's decision behavior.
Specifically, Item Value is calculated in combination with the key index of the impact factor of each attribute of article and article. Wherein, the key index of the article can include: sales volume or profit.
In embodiments of the present invention, article can be quantified to the value of businessman by above step, it is logical accurately reflects article Influence user's decision behavior bring shop income is crossed, consequently facilitating the reasonable selection scheme of businessman's rapid development.
Fig. 2 is the key step schematic diagram of the appraisal procedure of object-oriented according to another embodiment of the present invention.Such as Fig. 2 institute Show, the appraisal procedure of the object-oriented of the embodiment of the present invention includes:
Step S201, the index S (A, B) that calculating article A is substituted by article B based on the same order of same user.
For example, it is assumed that same user there are four order, is order 1, order 2, order 3, order 4 respectively, then may make up as It places an order pair: (order 1, order 2), (order 1, order 3), (order 1, order 4), (order 2, order 3), (order 2, order 4), (order 3, order 4).For each order pair of the user, the index S (A, B) that article A is substituted by article B is calculated separately. Wherein, article A, article B are the article that any two product number (sku) is different in same category, such as in soda Pepsi Cola and Coca-Cola.
Illustratively, it can calculate according to the following formula S (A, B):
Wherein, Sales1AIndicate the sales volume of article A in order 1；GAIndicate the sale of article A from order 1 to order 2 Volume accounting penalty values；GBIt indicates from order 1 to order 2, the sales volume accounting value added of article B；It indicates from order 1 to ordering Single 2, with article A, B with the sales volume accounting value added of other articles of category；N is the same category described in order 1, order 2 Other articles species number；S1For the sales volume of order 1；S2For the sales volume of order 2.
Also, GAMeet GA=P1A-P2A, GBMeet GB=P2B-P1B.Wherein, P1AFor sales volume of the article A in order 1 Accounting, P2AThe sales volume accounting for being article A in order 2, P1BThe sales volume accounting for being article B in order 1, P2BFor article B Sales volume accounting in order 2.
Step S202, the theil indexes of user are calculated, and S (A, B) is optimized according to the theil indexes of the user, To obtain optimum results S'(A, B).
In embodiments of the present invention, it is contemplated that different performances can be presented when buying article in the consumer of different characters. In order to more accurately measure article to the influence power of user's decision behavior, therefore when calculating article is by proxy indicator, The influence of Consumer Characteristics can be removed according to step S202.
Specifically, in this step, the theil indexes and S'(A of user, B can be calculated based on following formula):
Wherein, TIC,DIndicate that the theil indexes of user C, K indicate alternative object that user C is bought, article p in category D The species number of product, salesp,CIndicate the sales volume of the article p of user C purchase, avg salesD,CIndicate the category of user C purchase The sales volume mean value of all items, S'(A, B in D) indicate the result optimized to S (A, B).
Step S203, based on all orders to and article A all alternative articles to S'(A, B) summation, to obtain Article A by proxy indicator SA。
Specifically, S can be calculated according to the following formulaA:
SA=∑All commodity∑All orders paira*S'(A,B)；
Wherein, SAIt is article A by proxy indicator；S'(A, B) indicate the result after optimizing to S (A, B)； ∑All orders pairA*S (A, B) expression is weighted summation to obtained S (A, B) to based on all orders, and a is weight factor, ∑All commodityIndicate that article A is summed by the index that the every other article of same category substitutes.
It should be pointed out that step S202 is optional step.In another embodiment, object can also be calculated based on S (A, B) Product A by proxy indicator SA。
In embodiments of the present invention, it by step S201 to step S203, realizes to article by proxy indicator It calculates, capable of objectively responding consumer, most intuitively consumption is reacted to the article, so that businessman be helped preferably to carry out to article Assessment.
Step S204, building includes same category various article by proxy indicator data, the instruction of goods attribute data Practice collection.
Wherein, the goods attribute data are made of the goods attribute value of multiple dimensions.For example, a certain goods attribute data It include: the attribute value of the attributes such as article packaging, odor type, taste, specification and net content.
Specifically, which includes: to pre-process to goods attribute value, to obtain goods attribute data, then, knot That closes the same category various article that statistics obtains obtains training set by proxy indicator data, with the goods attribute data.Tool When body is implemented, when goods attribute value is continuous data, minimum value-maximum value normalization (Min-Max can be used Scaling) method is handled；When goods attribute value is discrete data, feature binarization method can be used and handled. Wherein, the normalized calculation formula of minimum value-maximum value is as follows:
In formula, x' indicates after normalizing as a result, x indicates pending data, xmaxIndicate the maximum value in data, xminTable Minimum value of the registration in.
In embodiments of the present invention, by step S204, the training set of following form: { (x can be obtained1,y1),(x2, y2),…,(xi,yi)…,(xN,yN)}.Wherein, which includes N number of sample point；(xi,yi) indicate that i-th of article is corresponding Sample point；xiThe goods attribute data for indicating i-th of article, are made of the attribute value of multiple dimensions；yiIndicate i-th of article By proxy indicator sample value.
Step S205, basis includes the same category various article by proxy indicator data and goods attribute data Training set training Random Forest model, learn to by proxy indicator, to obtain goods attribute to user's decision behavior Impact factor.
In embodiments of the present invention, which can be divided into step 1 to step 3 again, specific as follows:
It include the more decision trees to obtain Step 1: constructing more decision trees based on least square regression tree algorithm Prediction model, specifically include:
A), criterion is minimized according to square error, selects optimal cutting vector x in dependence data.Find cutting to The cut-off s of the jth dimension attribute of x is measured, so that following formula obtains minimum value:
Wherein, c1For cutting region R1The output valve of (j, s), c2For cutting region R2The output valve of (j, s),To fall in region R1Sample point in (j, s) it is all by proxy indicator sample value and c1Square mistake The sum of difference,To fall in region R2Sample point in (j, s) by proxy indicator sample value and c2It is flat The sum of square error.
B), after obtaining optimal cutting vector x, calculate separately and each cut subregional output valve；
C), circulation executes step a and step b, until meeting circulation stop condition；
D), the input space is divided into M region, generates decision tree；
Wherein, cmIndicate the output valve of decision tree.
Further, in order to which the output valve to decision tree optimizes, decision tree can be calculated by inverse distance-weighting function Output valve；
Wherein, cmIndicate output valve；J indicates goods attribute used in cutting；S indicates cut-off；It indicates new The article attribute data to be measured and region R of inputmThe inverse of the Euclidean distance between goods attribute sample point in (j, s)；xi、 xkIndicate region RmGoods attribute sample point in (j, s)；yiIt indicates and goods attribute sample point xiIt is corresponding by alternative finger Mark sample value；Expression pairIt sums.
Step 2: an attribute dimensions to goods attribute sample point carry out randomization, and will be after randomization Data input the prediction model, to obtain by proxy indicator predicted value.
Illustratively, institute can be realized by way of adding noise jamming to goods attribute sample point a attribute dimensions State randomization.
Step 3: calculating by proxy indicator predicted value and by the offset distance between proxy indicator true value, and by institute State impact factor of the offset distance as the goods attribute to user's decision behavior.
In this step, the offset distance is bigger, indicates influence of the goods attribute of the dimension to user's decision behavior It is bigger.In embodiments of the present invention, by step S204 to step S205, the calculating to impact factor is realized, so that businessman Influence of the goods attribute for user's decision behavior can be intuitively understood, convenient for instructing businessman to carry out category management.
Step S206, Item Value is calculated according to impact factor of the goods attribute to user's decision behavior.
In embodiments of the present invention, the Item Value can be calculated according to the following formula；
Wherein, PskuIndicate Item Value, KeyskuIndicate the key index of the article, γiIndicate i-th of goods attribute pair By the impact factor of proxy indicator, WiIt indicates with all items in category with i-th of goods attribute in the category Key index accounting, n indicate the attribute number of the article.
Further, the key index of the article includes: the sales volume of article or the profit of article.Work as KeyskuFor article Sales volume when, WiFor sales volume accounting of all items in same category with i-th of goods attribute in the category.When KeyskuFor article profit when, WiIt is accounted for for profit of all items in same category with i-th of goods attribute in the category Than.
In embodiments of the present invention, article can be quantified to the value of businessman by above step more accurately, objective, Intuitively reflection article is by influencing user's decision behavior bring shop income, consequently facilitating businessman's rapid development reasonably selects Product scheme.
Fig. 3 is the main modular schematic diagram of the assessment device of object-oriented according to an embodiment of the invention.Such as Fig. 3 institute Show, the assessment device 300 of the object-oriented of the embodiment of the present invention includes: statistical module 301, study module 302, computing module 303。
Statistical module 301, for counting with category various article by proxy indicator data.
Wherein, the article includes commodity, described to be used to measure article to the shadow of user's decision behavior by proxy indicator Ring power.Illustratively, for this category of soda, can count respectively all items under the category (for example, Coca-Cola, Pepsi Cola etc.) by proxy indicator.
In general, each article usually has multiple goods attributes, each goods attribute can to the influence power of user's decision behavior It can be different.For example, for the article under this category of soda, with article packaging, odor type, taste, specification, net content And other items attribute, the influence power that each goods attribute buys soda for user may be different.In order to quantify each goods attribute To the influence power of user's decision behavior, recurrence learning can be carried out by proxy indicator of the study module 302 to article.
Study module 302 includes the same category various article by proxy indicator data and article for basis The training set training machine learning model of attribute data, learns to by proxy indicator, to obtain goods attribute to user The impact factor of decision behavior.
Wherein, the goods attribute data are made of the goods attribute value of multiple dimensions.The machine learning model can wrap It includes: Random Forest model or GBDT (gradient promotes decision tree) model.
Computing module 303, for calculating Item Value according to impact factor of the goods attribute to user's decision behavior.
Specifically, computing module 303 is in combination with the impact factor of each attribute of article and the key index meter of article Calculate Item Value.Wherein, the key index of the article can include: sales volume or profit.
In embodiments of the present invention, study module is passed through by proxy indicator data by statistical module counts article Goods attribute is obtained to the impact factor of user's decision behavior, and passes through computing module according to the goods attribute to user's decision The impact factor of behavior calculates Item Value, can quantify article to the value of businessman, accurately reflect article by influencing user Decision behavior bring shop income, consequently facilitating the reasonable selection scheme of businessman's rapid development.
Fig. 4 is the main modular schematic diagram of the assessment device of object-oriented according to another embodiment of the present invention.Such as Fig. 4 institute Show, the assessment device 400 of the object-oriented of the embodiment of the present invention includes: statistical module 401, optimization module 402, study module 403, computing module 404.
Statistical module 401, the index S that calculating article A is substituted by article B for the same order based on same user (A,B)。
For example, it is assumed that same user there are four order, is order 1, order 2, order 3, order 4 respectively, then may make up as It places an order pair: (order 1, order 2), (order 1, order 3), (order 1, order 4), (order 2, order 3), (order 2, order 4), (order 3, order 4).For each order pair of the user, the index S (A, B) that article A is substituted by article B is calculated separately. Wherein, article A, article B are the article that any two product number (sku) is different in same category, such as in soda Pepsi Cola and Coca-Cola.
Illustratively, statistical module 401 can calculate S (A, B) according to the following formula:
Wherein, Sales1AIndicate the sales volume of article A in order 1；GAIndicate the sale of article A from order 1 to order 2 Volume accounting penalty values；GBIt indicates from order 1 to order 2, the sales volume accounting value added of article B；It indicates from order 1 to ordering Single 2, with article A, B with the sales volume accounting value added of other articles of category；N is the same category described in order 1, order 2 Other articles species number；S1For the sales volume of order 1；S2For the sales volume of order 2.Also, GAMeet GA=P1A-P2A, GB Meet GB=P2B-P1B.Wherein, P1AThe sales volume accounting for being article A in order 1, P2AFor sales volume of the article A in order 2 Accounting, P1BThe sales volume accounting for being article B in order 1, P2BThe sales volume accounting for being article B in order 2.
In embodiments of the present invention, it is contemplated that different performances can be presented when buying article in the consumer of different characters. In order to more accurately measure article to the influence power of user's decision behavior, therefore when calculating article is by proxy indicator, The influence of Consumer Characteristics can be removed according to optimization module 402.
Optimization module 402, for calculating the theil indexes of user, and according to the theil indexes of the user to S (A, B) into Row optimization, to obtain optimum results S'(A, B).
Specifically, optimization module 402 can calculate the theil indexes and S'(A of user, B based on following formula):
Wherein, TIC,DIndicate that the theil indexes of user C, K indicate alternative object that user C is bought, article p in category D The species number of product, salesp,CIndicate the sales volume of the article p of user C purchase, avg salesD,CIndicate the category of user C purchase The sales volume mean value of all items, S'(A, B in D) indicate the result optimized to S (A, B).
Statistical module 401 be also used to based on all orders to and article A all alternative articles to S'(A, B) ask With, with obtain article A by proxy indicator SA。
Specifically, statistical module 401 can calculate S according to the following formulaA:
SA=∑All commodity∑All orders paira*S'(A,B)；
Wherein, SAIt is article A by proxy indicator；S'(A, B) indicate the result after optimizing to S (A, B)； ∑All orders pairA*S (A, B) expression is weighted summation to obtained S (A, B) to based on all orders, and a is weight factor, ∑All commodityIndicate that article A is summed by the index that the every other article of same category substitutes.
It should be pointed out that optimization module is optional module.In another embodiment, it can also be not provided with optimization module, But by statistical module 401 be directly based upon S (A, B) calculate article A by proxy indicator SA。
In embodiments of the present invention, by statistical module, the above-mentioned function of optimization module, being substituted to article is realized The calculating of property index.The obtained consumer that can be objectively responded by proxy indicator to the article, most intuitively react by consumption, from And businessman is helped preferably to assess article.
Study module 403 includes the same category various article by proxy indicator data and article for basis The training set training Random Forest model of attribute data, learns to by proxy indicator, to obtain goods attribute to user The impact factor of decision behavior.
Wherein, the goods attribute data are made of the goods attribute value of multiple dimensions.For example, a certain goods attribute data It include: the attribute value of the attributes such as article packaging, odor type, taste, specification and net content.
In embodiments of the present invention, study module 403 can be realized according to following below scheme and learn to by proxy indicator Operation:
One, study module 403 is based on least square regression tree algorithm and constructs more decision trees, includes described more to obtain The prediction model of decision tree, specifically includes:
A), study module 403 minimizes criterion according to square error, selects optimal cutting vector x in dependence data.I.e. The cut-off s of the jth dimension attribute of cutting vector x is found, so that following formula obtains minimum value:
Wherein, c1For cutting region R1The output valve of (j, s), c2For cutting region R2The output valve of (j, s),To fall in region R1Sample point in (j, s) it is all by proxy indicator sample value and c1Square mistake The sum of difference,To fall in region R2Sample point in (j, s) by proxy indicator sample value and c2It is flat The sum of square error.
B), for study module 403 after obtaining optimal cutting vector x, subregional output valve is each cut in calculating；
C), the circulation of study module 403 executes step a and step b, until meeting circulation stop condition；
D), the input space is divided into M region by study module 403, generates decision tree；
Wherein, cmSubregional output valve is each cut in expression.
Further, in order to which the output valve to decision tree optimizes, study module 403 can pass through inverse distance-weighting function meter Calculate the output valve of decision tree；
Wherein, cmIndicate output valve；J indicates goods attribute used in cutting；S indicates cut-off；It indicates new The article attribute data to be measured and region R of inputmThe inverse of the Euclidean distance between goods attribute sample point in (j, s)；xi、 xkIndicate region RmGoods attribute sample point in (j, s)；yiIt indicates and goods attribute sample point xiIt is corresponding by alternative finger Mark sample value；Expression pairIt sums.
Two, study module 403 carries out randomization to an attribute dimensions of goods attribute sample point, and will randomization Data that treated input the prediction model, to obtain by proxy indicator predicted value.
Illustratively, study module 403 can be interfered by an attribute dimensions plus noise to goods attribute sample point Mode realizes the randomization.
Three, study module 403 is calculated by proxy indicator predicted value and by the offset distance between proxy indicator true value, And using the offset distance as the goods attribute to the impact factor of user's decision behavior.Wherein, the offset distance is bigger, Indicate that influence of the goods attribute of the dimension to user's decision behavior is bigger.
In embodiments of the present invention, the calculating to impact factor is realized by the above-mentioned function of study module, so that quotient Family can intuitively understand influence of the goods attribute for user's decision behavior, convenient for instructing businessman to carry out category management.
Computing module 404, for calculating Item Value according to impact factor of the goods attribute to user's decision behavior.
Illustratively, computing module 404 can calculate the Item Value according to the following formula；
Wherein, PskuIndicate Item Value, KeyskuIndicate the key index of the article, γiIndicate i-th of goods attribute pair By the impact factor of proxy indicator, WiIt indicates with all items in category with i-th of goods attribute in the category Key index accounting, n indicate the attribute number of the article.
Wherein, the key index of the article includes: the sales volume of article or the profit of article.Work as KeyskuFor article When sales volume, WiFor sales volume accounting of all items in same category with i-th of goods attribute in the category.When KeyskuFor article profit when, WiIt is accounted for for profit of all items in same category with i-th of goods attribute in the category Than.
Further, the device of the embodiment of the present invention may also include that building module, include a variety of objects of same category for constructing The training set by proxy indicator data, goods attribute data of product.
Specifically, building module constructs the operation of the training set can include: building module carries out goods attribute value Pretreatment, to obtain goods attribute data；Then, the same category various article obtained in conjunction with statistical module by proxy indicator Data obtain training set with the goods attribute data.When it is implemented, can be adopted when goods attribute value is continuous data It is handled with minimum value-maximum value normalization (Min-Max scaling) method；When goods attribute value is discrete data When, feature binarization method can be used and handled.
The device of the embodiment of the present invention can quantify article to the value of businessman more accurately, objective, intuitively reflect Article is by influencing user's decision behavior bring shop income, consequently facilitating the reasonable selection scheme of businessman's rapid development.
In addition, the assessment device of the object-oriented based on the embodiment of the present invention, has further obtained selection system shown in fig. 5 System.As shown in figure 5, the selection system 500 of the embodiment of the present invention includes: data module 501, algoritic module 502, interface module 503。
Data module 501, comprising: basic data module and interim data module.Wherein, basic data module, it is main to use In storage order data, location data, User ID (mark) data and goods attribute data etc.；Interim data module, Be mainly used for store algoritic module generate intermediate data, such as: article by proxy indicator, goods attribute to user's decision The impact factor etc. of behavior.
Algoritic module 501 specifically includes that statistical module, study module, computing module, selection module.
Statistical module, be mainly used for calculating same category various article by proxy indicator.Wherein, the article can be Commodity.
Study module, be mainly used for calculating based on Random Forest model influence of the goods attribute to user's decision behavior because Son.
Computing module is mainly used for calculating article valence according to impact factor of the goods attribute to user's decision behavior Value.
Selection module is mainly used for according to the Item Value and consumer demand structure and category being calculated Management rule formulates selection scheme.
Illustratively, the selection module according to the Item Value and consumer demand structure being calculated and Category management rule, formulates selection scheme can include: article on sale is divided into four categories using profit and sales volume, specifically Are as follows: emphasis category, profit protection category, passenger flow pull category, fillibility category.For first three category, needed using consumer The each capacity that segments market of structure determination is sought, capacity is then segmented market according to this and the Item Value that is calculated carries out Selection.For fillibility category, the selection quantity of the category is calculated using market capacity accounting, then according to determining selection Quantity and the Item Value being calculated carry out selection.
Wherein, the consumer demand structure can be determined according to such as under type: the article based on each businessman in target area Sales data estimates consumer demand, specifically includes: carrying out category using the sales data of each businessman in target area It splits, obtains the items sold situation of corresponding category；Article is broken up into the set for attribute, calculates separately corresponding combinations of attributes Market capacity.
Interface module 503, for according to user request carry out by proxy indicator data, the Item Value being calculated, Or the displaying of selection scheme.
Fig. 6 is shown can be using the appraisal procedure of the object-oriented of the embodiment of the present invention or the assessment device of object-oriented Exemplary system architecture 600.
As shown in fig. 6, system architecture 600 may include terminal device 601,602,603, network 604 and server 605. Network 604 between terminal device 601,602,603 and server 605 to provide the medium of communication link.Network 604 can be with Including various connection types, such as wired, wireless communication link or fiber optic cables etc..
User can be used terminal device 601,602,603 and be interacted by network 604 with server 605, to receive or send out Send message etc..Various telecommunication customer end applications, such as the application of shopping class, net can be installed on terminal device 601,602,603 The application of page browsing device, searching class application, instant messaging tools, mailbox client, social platform software etc..
Terminal device 601,602,603 can be the various electronic equipments with display screen and supported web page browsing, packet Include but be not limited to smart phone, tablet computer, pocket computer on knee and desktop computer etc..
Server 605 can be to provide the server of various services, such as utilize terminal device 601,602,603 to user The selection system demonstration interface browsed provides the back-stage management server supported.Back-stage management server can be to receiving The data such as inquiry request carry out the processing such as analyzing, and processing result is fed back to terminal device.
It should be noted that the appraisal procedure of object-oriented provided by the embodiment of the present invention is generally held by server 605 Row, correspondingly, the assessment device of object-oriented is generally positioned in server 605.
It should be understood that the number of terminal device, network and server in Fig. 6 is only schematical.According to realization need It wants, can have any number of terminal device, network and server.
Fig. 7 shows the structural schematic diagram for being suitable for the computer system 700 for the server for being used to realize the embodiment of the present invention. Computer system shown in Fig. 7 is only an example, should not function to the embodiment of the present invention and use scope bring it is any Limitation.
As shown in fig. 7, computer system 700 includes central processing unit (CPU) 701, it can be read-only according to being stored in Program in memory (ROM) 702 or be loaded into the program in random access storage device (RAM) 703 from storage section 708 and Execute various movements appropriate and processing.In RAM 703, also it is stored with system 700 and operates required various programs and data. CPU 701, ROM 702 and RAM 703 are connected with each other by bus 704.Input/output (I/O) interface 705 is also connected to always Line 704.
I/O interface 705 is connected to lower component: the importation 706 including keyboard, mouse etc.；It is penetrated including such as cathode The output par, c 707 of spool (CRT), liquid crystal display (LCD) etc. and loudspeaker etc.；Storage section 708 including hard disk etc.； And the communications portion 709 of the network interface card including LAN card, modem etc..Communications portion 709 via such as because The network of spy's net executes communication process.Driver 710 is also connected to I/O interface 705 as needed.Detachable media 711, such as Disk, CD, magneto-optic disk, semiconductor memory etc. are mounted on as needed on driver 710, in order to read from thereon Computer program be mounted into storage section 708 as needed.
Particularly, disclosed embodiment, the process described above with reference to flow chart may be implemented as counting according to the present invention Calculation machine software program.For example, embodiment disclosed by the invention includes a kind of computer program product comprising be carried on computer Computer program on readable medium, the computer program include the program code for method shown in execution flow chart.? In such embodiment, which can be downloaded and installed from network by communications portion 709, and/or from can Medium 711 is dismantled to be mounted.When the computer program is executed by central processing unit (CPU) 701, system of the invention is executed The above-mentioned function of middle restriction.
It should be noted that computer-readable medium shown in the present invention can be computer-readable signal media or meter Calculation machine readable storage medium storing program for executing either the two any combination.Computer readable storage medium for example can be --- but not Be limited to --- electricity, magnetic, optical, electromagnetic, infrared ray or semiconductor system, device or device, or any above combination.Meter The more specific example of calculation machine readable storage medium storing program for executing can include but is not limited to: have the electrical connection, just of one or more conducting wires Taking formula computer disk, hard disk, random access storage device (RAM), read-only memory (ROM), erasable type may be programmed read-only storage Device (EPROM or flash memory), optical fiber, portable compact disc read-only memory (CD-ROM), light storage device, magnetic memory device, Or above-mentioned any appropriate combination.In the present invention, computer readable storage medium can be it is any include or storage journey The tangible medium of sequence, the program can be commanded execution system, device or device use or in connection.And at this In invention, computer-readable signal media may include in a base band or as carrier wave a part propagate data-signal, Wherein carry computer-readable program code.The data-signal of this propagation can take various forms, including but unlimited In electromagnetic signal, optical signal or above-mentioned any appropriate combination.Computer-readable signal media can also be that computer can Any computer-readable medium other than storage medium is read, which can send, propagates or transmit and be used for By the use of instruction execution system, device or device or program in connection.Include on computer-readable medium Program code can transmit with any suitable medium, including but not limited to: wireless, electric wire, optical cable, RF etc. are above-mentioned Any appropriate combination.
Flow chart and block diagram in attached drawing are illustrated according to the system of various embodiments of the invention, method and computer journey The architecture, function and operation in the cards of sequence product.In this regard, each box in flowchart or block diagram can generation A part of one module, program segment or code of table, a part of above-mentioned module, program segment or code include one or more Executable instruction for implementing the specified logical function.It should also be noted that in some implementations as replacements, institute in box The function of mark can also occur in a different order than that indicated in the drawings.For example, two boxes succeedingly indicated are practical On can be basically executed in parallel, they can also be executed in the opposite order sometimes, and this depends on the function involved.Also it wants It is noted that the combination of each box in block diagram or flow chart and the box in block diagram or flow chart, can use and execute rule The dedicated hardware based systems of fixed functions or operations is realized, or can use the group of specialized hardware and computer instruction It closes to realize.
Being described in module involved in the embodiment of the present invention can be realized by way of software, can also be by hard The mode of part is realized.Described module also can be set in the processor, for example, can be described as: a kind of processor packet Include statistical module, study module, computing module.Wherein, the title of these modules is not constituted under certain conditions to the module The restriction of itself, for example, statistical module is also described as " module by proxy indicator data of statistics article ".
As on the other hand, the present invention also provides a kind of computer-readable medium, which be can be Included in equipment described in above-described embodiment；It is also possible to individualism, and without in the supplying equipment.Above-mentioned calculating Machine readable medium carries one or more program, when said one or multiple programs are executed by the equipment, makes Obtain the equipment and execute following below scheme: statistics is with category various article by proxy indicator data；According to including the same product The training set training machine learning model by proxy indicator data and goods attribute data of class various article, to alternative Index is learnt, to obtain goods attribute to the impact factor of user's decision behavior；It is determined according to the goods attribute to user The impact factor of plan behavior calculates Item Value.
Above-mentioned specific embodiment, does not constitute a limitation on the scope of protection of the present invention.Those skilled in the art should be bright It is white, design requirement and other factors are depended on, various modifications, combination, sub-portfolio and substitution can occur.It is any Made modifications, equivalent substitutions and improvements etc. within the spirit and principles in the present invention, should be included in the scope of the present invention Within.

Claims (18)

1. a kind of appraisal procedure of object-oriented, which is characterized in that the described method includes:
Statistics is with category various article by proxy indicator data；
According to include the same category various article by the training set of proxy indicator data and goods attribute data training Machine learning model learns to by proxy indicator, to obtain goods attribute to the impact factor of user's decision behavior；
Item Value is calculated according to impact factor of the goods attribute to user's decision behavior.
2. the method according to claim 1, wherein the machine learning model includes: Random Forest model.
3. according to the method described in claim 2, it is characterized in that, the basis includes the quilt of the same category various article The training set training machine learning model of proxy indicator data and goods attribute data, learns to by proxy indicator, To include: the step of obtaining impact factor of the goods attribute to user's decision behavior
Construct more decision trees based on least square regression tree algorithm, with obtain include the more decision trees prediction model；
Randomization is carried out to an attribute dimensions of goods attribute sample point, and should by the data input after randomization Prediction model, to obtain by proxy indicator predicted value；
It calculates by proxy indicator predicted value and by the offset distance between proxy indicator true value, and the offset distance is made It is the goods attribute to the impact factor of user's decision behavior.
4. according to the method described in claim 3, it is characterized in that, the method also includes:
When based on least square regression tree algorithm building decision tree, the output of decision tree is calculated according to inverse distance-weighting function Value；
Wherein, cmIndicate output valve；J indicates goods attribute used in cutting；S indicates cut-off；Indicate new input Goods attribute data and region RmThe inverse of the Euclidean distance between goods attribute sample point in (j, s)；xi、xkIndicate Region RmGoods attribute sample point in (j, s)；yiIt indicates and goods attribute sample point xiIt is corresponding by proxy indicator sample Value；Expression pairIt sums.
5. the method according to claim 1, wherein calculating the Item Value according to the following formula；
Wherein, PskuIndicate Item Value, KeyskuIndicate the key index of the article, γiIndicate i-th of goods attribute to being replaced For the impact factor of property index, WiIndicate the key with all items in category with i-th of goods attribute in the category Index accounting, n indicate the attribute number of the article.
6. the method according to claim 1, wherein calculate according to the following formula the article by alternative finger Mark:
SA=∑All commodity∑All orders paira*S(A,B)；
Wherein, SAIt is article A by proxy indicator；S (A, B) indicates an order based on same user to being calculated The index that article A is substituted by article B, ∑All orders pairA*S (A, B) expression adds obtained S (A, B) to based on all orders Power summation, a is weight factor, ∑All commodityIndicate that article A is summed by the index that the every other article of same category substitutes.
7. according to the method described in claim 6, it is characterized in that, calculating S (A, B) according to the following formula:
Wherein, Sales1AIndicate the sales volume of article A in order 1；GAIndicate that the sales volume of article A accounts for from order 1 to order 2 Compare penalty values；GBIt indicates from order 1 to order 2, the sales volume accounting value added of article B；It indicates from order 1 to order 2, With article A, B with the sales volume accounting value added of other articles of category；N be described in order 1, order 2 with category its The species number of his article；S1For the sales volume of order 1；S2For the sales volume of order 2.
8. the method according to the description of claim 7 is characterized in that the method also includes:
The theil indexes of user are calculated according to the following formula, and S (A, B) is optimized according to the theil indexes of the user, With according to the S (A, B) after optimization calculate the article by proxy indicator:
Wherein, TIC,DIndicate that the theil indexes of user C, K indicate user C purchase, the alternative article of article p in category D Species number, salesp,CIndicate the sales volume of the article p of user C purchase, avg salesD,CIn the category D for indicating user C purchase The sales volume mean value of all items, S'(A, B) indicate the result optimized to S (A, B).
9. a kind of assessment device of object-oriented, which is characterized in that described device includes:
Statistical module, for counting with category various article by proxy indicator data；
Study module includes the same category various article by proxy indicator data and goods attribute data for basis Training set training machine learning model, learn to by proxy indicator, to obtain goods attribute to by proxy indicator Impact factor；
Computing module, for calculating Item Value to by the impact factor of proxy indicator according to the goods attribute.
10. device according to claim 9, which is characterized in that the machine learning model in the study module include: with Machine forest model.
11. device according to claim 9, which is characterized in that the study module is according to including that the same category is more The training set training machine learning model by proxy indicator data and goods attribute data of kind article, to by proxy indicator Learnt, includes: to obtain goods attribute to the operation of the impact factor of user's decision behavior
The study module is based on least square regression tree algorithm and constructs more decision trees, includes the more decision trees to obtain Prediction model；
The study module carries out randomization to an attribute dimensions of goods attribute sample point, and will be after randomization Data input the prediction model, to obtain by proxy indicator predicted value；
The study module is calculated by proxy indicator predicted value and by the offset distance between proxy indicator true value, and by institute State impact factor of the offset distance as the goods attribute to user's decision behavior.
12. device according to claim 11, which is characterized in that
The study module is determined when based on least square regression tree algorithm building decision tree according to the calculating of inverse distance-weighting function The output valve of plan tree；
Wherein, cmIndicate output valve；J indicates goods attribute used in cutting；S indicates cut-off；Indicate new input Attribute data to be measured and region RmThe inverse of the Euclidean distance between goods attribute sample point in (j, s)；xi、xkIndicate Region RmGoods attribute sample point in (j, s)；yiIt indicates and goods attribute sample point xiIt is corresponding by proxy indicator sample Value；Expression pairIt sums.
13. device according to claim 9, which is characterized in that the computing module calculates the object according to the following formula Product value；
Wherein, PskuIndicate Item Value, KeyskuIndicate the key index of the article, γiIndicate i-th of goods attribute to being replaced For the impact factor of property index, WiIndicate the key with all items in category with i-th of goods attribute in the category Index accounting, n indicate the attribute number of the article.
14. device according to claim 9, which is characterized in that the statistical module calculates the object according to the following formula Product by proxy indicator:
SA=∑All commodity∑All orders paira*S(A,B)；
Wherein, SAIt is article A by proxy indicator；S (A, B) indicates an order based on same user to being calculated The index that article A is substituted by article B；∑All orders pairA*S (A, B) expression adds obtained S (A, B) to based on all orders Power summation, a is weight factor, ∑All commodityIndicate that article A is summed by the index that the every other article of same category substitutes.
15. device according to claim 14, which is characterized in that the statistical module calculate according to the following formula S (A, B):
Wherein, Sales1AIndicate the sales volume of article A in order 1；GAIndicate that the sales volume of article A accounts for from order 1 to order 2 Compare penalty values；GBIt indicates from order 1 to order 2, the sales volume accounting value added of article B；It indicates from order 1 to order 2, With article A, B with the sales volume accounting value added of other articles of category；N be described in order 1, order 2 with category its The species number of his article；S1For the sales volume of order 1；S2For the sales volume of order 2.
16. device according to claim 15, which is characterized in that described device further include:
Optimization module, for calculating the theil indexes of user according to the following formula, and according to the theil indexes of the user to S (A, B) is optimized, so that the statistical module is according to S (A, B) the calculating article after optimization by proxy indicator:
Wherein, TIC,DIndicate that the theil indexes of user C, K indicate user C purchase, the alternative article of article p in category D Species number, salesp,CIndicate the sales volume of the article p of user C purchase, avg salesD,CIn the category D for indicating user C purchase The sales volume mean value of all items, S'(A, B) indicate the result optimized to S (A, B).
17. a kind of server characterized by comprising
One or more processors；
Storage device, for storing one or more programs,
When one or more of programs are executed by one or more of processors, so that one or more of processors are real The now method as described in any in claim 1 to 8.
18. a kind of computer-readable medium, is stored thereon with computer program, which is characterized in that described program is held by processor The method as described in any in claim 1 to 8 is realized when row.
CN201711418565.1A 2017-12-25 2017-12-25 The appraisal procedure and device of object-oriented Pending CN109961307A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201711418565.1A CN109961307A (en) 2017-12-25 2017-12-25 The appraisal procedure and device of object-oriented

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201711418565.1A CN109961307A (en) 2017-12-25 2017-12-25 The appraisal procedure and device of object-oriented

Publications (1)

Publication Number Publication Date
CN109961307A true CN109961307A (en) 2019-07-02

Family

ID=67020902

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201711418565.1A Pending CN109961307A (en) 2017-12-25 2017-12-25 The appraisal procedure and device of object-oriented

Country Status (1)

Country Link
CN (1) CN109961307A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111695695A (en) * 2020-06-09 2020-09-22 北京百度网讯科技有限公司 Quantitative analysis method and device for user decision behaviors
CN112613953A (en) * 2020-12-29 2021-04-06 北京环球国广媒体科技有限公司 Commodity selection method, system and computer readable storage medium

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105844300A (en) * 2016-03-24 2016-08-10 河南师范大学 Optimized classification method and optimized classification device based on random forest algorithm
CN106611366A (en) * 2015-10-26 2017-05-03 阿里巴巴集团控股有限公司 Method and device for automatically selecting article
CN107358472A (en) * 2017-07-19 2017-11-17 北京京东尚科信息技术有限公司 Data processing method and device

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106611366A (en) * 2015-10-26 2017-05-03 阿里巴巴集团控股有限公司 Method and device for automatically selecting article
CN105844300A (en) * 2016-03-24 2016-08-10 河南师范大学 Optimized classification method and optimized classification device based on random forest algorithm
CN107358472A (en) * 2017-07-19 2017-11-17 北京京东尚科信息技术有限公司 Data processing method and device

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111695695A (en) * 2020-06-09 2020-09-22 北京百度网讯科技有限公司 Quantitative analysis method and device for user decision behaviors
CN111695695B (en) * 2020-06-09 2023-08-08 北京百度网讯科技有限公司 Quantitative analysis method and device for user decision behaviors
CN112613953A (en) * 2020-12-29 2021-04-06 北京环球国广媒体科技有限公司 Commodity selection method, system and computer readable storage medium

Similar Documents

Publication Publication Date Title
WO2022237667A1 (en) Method and apparatus for determining order fulfillment warehouse
CN109840730B (en) Method and device for data prediction
CN109636430A (en) Object identifying method and its system
CN109961299A (en) The method and apparatus of data analysis
CN109767150A (en) Information-pushing method and device
CN107590678A (en) Method of Commodity Recommendation and system
CN108776692A (en) Method and apparatus for handling information
CN113095893A (en) Method and device for determining sales of articles
CN108197298A (en) A kind of smart shopper exchange method and system based on natural language processing
CN110097302A (en) The method and apparatus for distributing order
CN110348921A (en) The method and apparatus that shops's article is chosen
CN110020876A (en) A kind of information generating method and device
CN109961307A (en) The appraisal procedure and device of object-oriented
CN109961199A (en) A kind of method and apparatus for analyzing data fluctuations
CN111612385A (en) Method and device for clustering to-be-delivered articles
CN109840724A (en) Method and apparatus for output information
CN110276566A (en) Information output method and device
US9213979B2 (en) Array attribute configurator
CN110490682B (en) Method and device for analyzing commodity attributes
CN110189203A (en) Information processing method, device, medium and calculating equipment
CN110880119A (en) Data processing method and device
CN110363558A (en) A kind of method and apparatus generating commodity association message
CN109767249A (en) The method and apparatus for predicting cost performance
CN110135772A (en) Method and apparatus for generating information
CN103455566B (en) Information displaying method and device

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