WO2014117268A1 - Intelligence pour fixation des prix de produits sans correspondance exacte - Google Patents

Intelligence pour fixation des prix de produits sans correspondance exacte Download PDF

Info

Publication number
WO2014117268A1
WO2014117268A1 PCT/CA2014/050038 CA2014050038W WO2014117268A1 WO 2014117268 A1 WO2014117268 A1 WO 2014117268A1 CA 2014050038 W CA2014050038 W CA 2014050038W WO 2014117268 A1 WO2014117268 A1 WO 2014117268A1
Authority
WO
WIPO (PCT)
Prior art keywords
product
retailer
computer system
exact matching
implemented method
Prior art date
Application number
PCT/CA2014/050038
Other languages
English (en)
Inventor
Dominic Pierre Plouffe
James Harold Reed
Matthew Steven Kitching
Original Assignee
360Pi Corporation
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 360Pi Corporation filed Critical 360Pi Corporation
Priority to CA2899529A priority Critical patent/CA2899529A1/fr
Publication of WO2014117268A1 publication Critical patent/WO2014117268A1/fr

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
    • G06Q30/0283—Price estimation or determination

Definitions

  • the invention relates generally to pricing intelligence, and more particularly to an automated method and system for comparing pricing information for non-exact matching products.
  • a retailer if they are to attract customers, should provide service, convenience, selection and/or pricing that is superior to that being offered by their competitors.
  • pricing often winds up being the determining factor when a potential customer is deciding where to make his or her purchase, and this is especially true in the case of on-line retail sales since service, convenience, and selection tend to be more-or-less the same from one on-line retailer to another.
  • Pricing intelligence has therefore become an important tool in the retail environment, and is being used increasingly by retailers to improve their competitiveness while at the same time maximizing profits.
  • pricing intelligence entails a detailed analysis of pricing information using modem data mining techniques, and is differentiated from other pricing methods by the extent and accuracy of the analysis.
  • a subject retailer In order to successfully implement a pricing intelligence method, a subject retailer must have access to reliable and current pricing information, which is for instance obtained from a competitor's e-commerce site and/or from another electronically accessible database. Typically, raw pricing information is obtained in an automated fashion using web-crawlers or web-scrapers. Unfortunately, the acquisition of meaningful pricing information is complicated by a number of factors, such as for instance the need to account for shipping costs in the overall price of a product, the need to purchase multiple items in order to qualify for an advertised discount, or the need to account for the value of "free gifts" that are included with the purchase of a product, etc.
  • e-commerce sites require products to be placed into a shopping cart before pricing information is provided, or a site may impose limits on the number of requests that can be made.
  • web-crawlers or web-scrapers that are used to obtain pricing information from different e-commerce sites should be able to simulate certain human activities and/or be able to extract data that is presented in different formats. Additionally, the pricing information should be updated or refreshed from time to time in order to account for price changes, or to account for the introduction of new products or the discontinuation of existing products, etc.
  • Different exclusive models commonly have one, or sometimes a few, features either more or less than a "standard” model that is produced by the same manufacturer.
  • an exclusive model of a 50-inch plasma HDTV that is available from a particular retailer has one additional HDMI input compared to the "standard" HDTV model of the same brand that is available from another retailer.
  • the manufacturer is likely to use different model numbers for the standard HDTV model and for the exclusive HDTV model, and similarly each retailer is likely to use a different SKU.
  • matching engines are known for performing apples-to-apples type matches of products that are found on a competitor's e-commerce site. For instance, artificial intelligence, semantic analysis, data mining, and image recognition technologies can be combined to determine exact product matches even when a unique product identifier, such as a UPC, is not provided on the product page.
  • Such matching engines are also capable of recognizing size and color variants of a product. That said, the matched products are identical models with identical features sets and identical brands. As noted above, different retailers often do not offer for sale identical products.
  • a price conscious consumer may be willing to consider an alternative product that is not available from the subject retailer. For instance, a consumer who wishes to purchase a 50-inch HDTV may be willing to consider similarly featured products under the Sony ® , Samsung ® and LG ® brands. Further, the consumer may consider products that are based on LED-LCD or CCFL-LCD technology.
  • implementation of a prior art pricing intelligence method based on an apple-to-apple type product match may result in the subject retailer adjusting the price of a Sony 50- inch LED-LCD HDTV in order to be price competitive with the identical Sony ® 50- inch LED-LCD HDTV that is offered by a competitor.
  • the competitor also offers a Samsung ® 50-inch LED-LCD HDTV, or even a LG ® 50-inch CCFL- LCD HDTV, at a lower price than the Sony 50-inch LED-LCD HDTV, then the consumer is likely to choose to purchase either the Samsung ® HDTV or the LG ® HDTV from the competitor and the subject retailer will lose the sale.
  • apple- to-apple type product matches do not provide the subject retailer with a full and robust pricing intelligence solution.
  • a computer system-implemented method comprising: using a process in execution on a processor of a computer system, accessing from a database product description information and pricing information for a product sold by a first retailer; accessing using a robot module, via an electronic communication network, product description information and pricing information for a non-exact matching product that is offered for sale by a second retailer; implementing in a feature analyzer module in execution on the processor of the computer system, product matching rules with parameters that are adjustable by product; accessing implemented product matching rules for the first retailer that cover the product sold by the first retailer, and responsive to the product description information for the product and for the nonexact matching product and said implemented rules, automatically determining whether or not a match exists between the product and the non-exact matching product to within a predetermined threshold; and upon determining a match to within the predetermined threshold, automatically generating a data output linking the product and the non-exact matching product, the data output for being accessed by
  • a computer system-implemented method comprising: using a process in execution on a processor of a computer system, accessing from a database product description information and pricing information for a product sold by a first retailer; accessing using a robot module, via an electronic communication network, product description information and pricing information for a non-exact matching product that is offered for sale by a second retailer; establishing a set of rules for determining an approximate match to the product sold by the first retailer; implementing in a feature analyzer module, in execution on the processor of the computer system, the established set of rules; accessing the implemented set of rules, and responsive to the product description information for the product and for the non-exact matching product and said implemented rules, automatically determining whether or not an approximate match exists between the product and the non-exact matching product; and upon determining an approximate match, automatically generating a data output linking the product and the non-exact matching product, the data output for being accessed by a price comparison process.
  • a data processing system comprising: a module executing on a processor and accessing from a database available product description information and pricing information for the product sold by the first retailer; a robot module executing on a processor and accessing, via an electronic communication network, product description information and pricing information for the non-exact matching product that is offered for sale by the second retailer; a module executing on a processor implementing product matching rules with parameters that are adjustable by product; a module executing on a processor accessing the implemented product matching rules for the first retailer that cover the product sold by the first retailer, and responsive to the product description information for the product and for the non-exact matching product and said implemented rules, automatically determining whether or not a match exists between the product and the non-exact matching product to within a predetermined threshold; and a module executing on a processor and responsive to a determination that a match exists between the product and the non-exact matching product to within a predetermined threshold, automatically generating a
  • Figure 1 is a simplified diagram showing a system according to an embodiment of the instant invention.
  • Figure 2 is a simplified diagram showing data fields for a product on an e- commerce site.
  • Figure 3 is a simplified diagram showing a table comparing the attributes for a product with the extracted attributes for a non-exact matching product.
  • so-called "exclusive models" of a product that are available only at specific retailers are non-exact matches to the standard version of that product if the feature sets are non-identical.
  • fully featured models and entry level models of a product are considered to be non-exact matching products.
  • products that are produced by different manufacturers under different brands, and that are sold by plural retailers under those brands are also considered to be "non- exact matching" products only if the product that is produced by one manufacturer does not have a feature that is present in the product that is produced by the other manufacturer.
  • the brand of each of the two products is not considered to be a feature for the purpose of determining non-exact matches; rather, in this context two products with identical feature sets that differ only by brand are considered to be exact-matching products.
  • a consumer having no interest in any of the non-overlapping features would consider two "non-exact matching" products to be perfectly interchangeable, without regard to brand, and therefore may consider both products when making a purchase decision. As such, it is in a subject retailer's best interest to include non-exact matching products when performing pricing intelligence analysis.
  • a data processing system based on an artificial intelligence (AI) process is used to perform non-exact product matching.
  • Product description information and pricing information for a product that is sold by a subject retailer is accessed from a memory storage element having stored thereon a database of such information.
  • the product description information for a particular product includes a manufacturer model number and/or another product identifier code for that product, as well as a list of features or attributes associated with that product.
  • a robot module such as for instance a so-called “web crawler” or “spider,” is used to access product description information and pricing information for a non- exact matching product that is offered for sale by a competitor.
  • the robot module accesses a plurality of product pages on the competitor's e-commerce site and/or retrieves data from another electronically accessible database.
  • Some of the accessed product pages and/or some of retrieved data may relate to products that are not similar, within a predetermined threshold, to any product that is offered by the subject retailer. Further, some of the accessed product pages and/or some of retrieved data may relate to products that are only an exact match to products that are offered by the subject retailer.
  • some of the accessed product pages and/or some of retrieved data may relate to products that are only a non-exact match to products that are offered by the subject retailer, or to products that are an exact match to some products that are offered by the subject retailer as well as a non-exact match to other products that are offered by the subject retailer.
  • a competitor offers a television that is an exact match to a television that is offered by the subject retailer in terms of brand and features.
  • the television that is offered by the competitor is also similar to another model of the same brand or to at least one model of a different brand offered by the subject retailer.
  • Determining if two products are non-matching, exact-matching or non-exact matching to within a predetermined threshold is based on product matching rules, with parameters that are adjustable by product, implemented in a feature analyzer module that is in execution on a processor of the data processing system.
  • This module uses an AI process that is capable of learning how to determine matches between two products that are not identical one to the other. Further, the AI process is capable of learning how a particular subject retailer defines "non-exact matching.” For instance, over time the AI process is capable of learning that a first retailer considers two televisions to be non-exact matching products if they have identical display technology, identical screen size, identical resolution, and identical audio output power.
  • the AI process is capable of learning that a second retailer considers two televisions to be non-exact matches if they have identical screen size, identical resolution, and the same number of HDMI inputs.
  • different retailers may define differently how the non-exact matching is to be determined, such as for instance based on the specific features that each retailer considers to be most important.
  • the subject retailer may provide initial parameters of the product matching rules for each different product or each different group of products. Alternatively, default initial values are used.
  • a user interface presents the subject retailer with available parameters for each product or group of products and the subject retailer selects and/or enters values for different parameters via the interface.
  • the parameters are presented in a checklist format via a display device of a computer system.
  • the AI process performs product matching for the subject retailer based on the initial parameters.
  • the AI process determines a set of features for a given product that is offered by the subject retailer.
  • the set of features is defined, for instance, by the subject retailer and includes the features that the subject retailer considers to be important.
  • the set of features includes a plurality of sub-sets of features in a hierarchical ranking. For example, three feature sub-sets are defined and the closeness of the non-exact match depends on how many of the sub-sets of features are common to two products. If both products have all three feature sub-sets in common then the non-exact match is considered to be a close match.
  • both products have only the top two highest ranked feature sub-sets in common then the non-exact match is considered to be less close.
  • the minimum requirement is that both products have the top ranked feature sub-set in common; otherwise the two products are considered to be non-matching products.
  • the AI process determines whether or not a match exists between the two products, to within a predetermined threshold, based on a comparison of the features that are present in each of the two products.
  • the subject retailer provides initial values for the different parameters of the matching rule, and the AI process learns how to determine matches between two products based on an analysis of all of the features that are available for the relevant class of products. Alternatively, default or null values are used initially.
  • the AI process generates a list of possible matches for a product that is offered by the subject retailer, including products that are believed to be close matches, products that are considered to be less close matches, and products that are considered to be non-matching.
  • the subject retailer reviews the products that are contained in each section of the list, and confirms the assignment of products on a product-by-product basis.
  • the subject retailer may indicate, for instance, that some of the products considered to be non-matching are actually close matches or less close matches, or that some of the products that are considered to be close matches are actually less close matches or non-matches, etc.
  • the AI system adjusts the parameters of the matching rule. When this process is iterated over time, as new products are introduced or old products are discontinued, the AI process continues to learn the attributes or features that are important for correctly determining non-exact matches between products.
  • the AI process learns based on feedback that is provided by the subject retailer, the same AI process is capable of learning each different retailer's definition of what constitutes a non-exact match.
  • the length of time that is required for the AI process to learn how to determine non-exact matches between products depends, partially, on the type of products that are being compared. Products with large feature or attribute sets, such as for instance automobiles or HDTVs, require longer learning periods than products with small feature sets, such as for instance coffee mugs or staplers.
  • the quality and consistency of the feedback that is provided by the subject retailer affects the length of time that it takes for the AI process to learn how to determine non-exact matches between products.
  • a data output linking the product and the non-exact matching product is generated.
  • the data output is for being accessed by a price comparison process at a later time, or optionally a price comparison is performed based on the data output and a result of the price comparison is presented in a human-intelligible form.
  • the predetermined threshold for determining a match between a product and a non-exact matching product may be defined in a variety of different ways. In generally, products that are determined to be exact matching products are excluded. According to one implementation, a matching-score value is determined based on a comparison of the features of the product and the features of the non-exact matching product. In this implementation the predetermined threshold defines a range of matching-score values that is indicative of a match between the product and the nonexact matching product. As such, matching scores that are outside of this range are considered to be indicative of either exact matching products or non-matching products.
  • the predetermined threshold defines which of the feature groups must be present in the non-exact matching product.
  • the predetermined threshold may require the non-exact matching product to have all of the features of at least the feature group that is at the top of the hierarchical arrangement of feature groups in order to be considered a match with the product.
  • a data output is generated that is indicative of a difference between a first feature set relating to the non-exact matching product and a second feature set relating to the product that is sold by the subject retailer. Based on this data output, for instance, a value of a feature that is non-overlapping between the first feature set and the second feature set may be determined. Further, statistical analysis may be used to compare the pricing information and features of a plurality of different non-exact matching products, which have different but similar combinations of features, to determine values for a plurality of different features. Feature value information is useful for pricing non- identical products.
  • the feature value information may be used to identify which features should be added to a product feature set or removed from a product feature set in order to provide an optimized feature set that increases profit on a per unit basis.
  • an existing product with the optimized feature set may be available from a manufacturer, and the subject retailer may make arrangements to carry that product. In other instances, the subject retailer may approach a
  • an adjustment to be applied to the price of the product that is sold by the subject retailer may be generated.
  • a subject retailer 102, a competitor 104 and a pricing intelligence provider 106 are in communication one with another via a communications network such as for instance wide area network (WAN) 108.
  • the subject retailer 102 has an e-commerce site 110 in communication with a memory storage device 112 having stored thereon a database including at least inventory data 114 relating to products that are offered by the subject retailer.
  • the inventory data 114 includes pricing information as well as descriptions for each of the products that are offered by the subject retailer.
  • the competitor 104 has an e-commerce site 116 in communication with a memory storage device 118 having stored thereon a database including at least inventory data 120 relating to products that are offered by the competitor.
  • the inventory data 120 includes pricing information as well as descriptions for each of the products that are offered.
  • the pricing intelligence provider 106 includes a data processing system 122 for comparing the price of a product sold by the subject retailer 102 to the price a non-exact matching product that is offered for sale by another retailer, such as for instance the competitor 104.
  • the data processing system 122 is in communication with a memory storage device 124 having stored thereon parameter data 126 and other data that is required during use of the data processing system 122.
  • FIG. 2 shown is a simplified diagram of a competitor's product page 200 for a product.
  • the product page 200 contains information relating to the product, such as for instance a photograph or another visual representation 202 of the product, the price 204 of the product, a description 206 of the product including an indication of features or attributes of the product, and optionally other information 208 relating to delivery terms, offer conditions, other limitations etc.
  • a robot module such as for instance a WebCrawler or spider, of the data processing system 122 extracts the data 202-208 from the product page 200.
  • FIG. 3 shown is a simplified comparison of the attributes or features of a product that is offered by the subject retailer 102 compared to the features that are believed to be present in the product that is represented by the product page 200, based on the data 202-208 extracted from the product page 200.
  • the product that is offered by the subject retailer 202 has attributes i-v.
  • the competitor's product is believed to have attributes i, ii and v; attribute iii is not present in this product and attribute iv may be present in this product.
  • the data processing system 122 of the pricing intelligence provider 106 implements product matching rules with parameters that are adjustable by product.
  • a data output linking the product and the non-exact matching product is generated automatically.
  • the data output is for being accessed by a price comparison process, which is in execution on a processor of the pricing intelligence provider 106.
  • the pricing intelligence provider generates the data output for being accessed by a price comparison process that is in execution on another processor, such as for instance a processor of the subject retailer 102.
  • the ability to automatically match a product and a non-exact matching product supports a number of enhanced functions, such as for instance i) determining the value of individual features or groups of features, ii) determining price adjustments to be competitive with non-identical products, iii) identifying desirable attributes that together are not available, iv) assessing the assignment of products to a specific taxonomy, etc.
  • matches are determined between a product and a non-exact matching product, it is possible to determine differences between the two products in terms of the type and/or quantity of features that are offered. For instance, it may be determined that for $50 more than the subject retailer is charging for a particular HDTV, the competitor is offering a similar HDTV with one additional HDMI input. Based on a larger number of similar comparisons, it may be possible to determine an approximate value of a feature. That is to say, using statistical methods etc. may reveal that an approximately $45-$55 value is associated with the HDMI input.
  • the value of other features may also be determined in this way.
  • Determining matches between non-exact matching products also facilitates determining price adjustments. It may be the case that the subject retailer is competitively priced relative to the price that other retailers are charging for the same product. Unfortunately, if similar but non-exact matching products are offered by other retailers at a better price point then the subject retailer may suffer poor sales. By ensuring price competitiveness with non-exact matching products in addition to exact matching products, the subject retailer is likely to increase the number of sales. Of course, if it is necessary to reduce the price of the product below a point of being profitable, then the subject retailer may instead discontinue selling the product and instead offer a product that may be priced more competitively with the non-exact matching products.
  • Statistical analysis may also be employed to identify features that are present in some products but not in others, and to further analyze the available groupings of features to identify if particularly desirable groupings of features are available. That is to say, once a particularly desirable grouping of features is identified, the subject retailer may attempt to determine if there are any products on the market that offer that particularly desirable grouping of features. If such a product is available, then the subject retailer may attempt to acquire the right to offer the product. Alternatively, the subject retailer may approach a manufacturer to request an exclusive model having the particularly desirable grouping of features if such a grouping is not currently available on the market. [0034] Additionally, determining matches between non-exact matching products may facilitate an assessing the taxonomy to which a product is assigned.
  • a HDTV offered by the subject retailer and classified as a consumer electronic product may be found to be a non-exact match to a competitors HDTV with at least and additional feature and classified as a business display product. If the business display product taxonomy allows the subject retailer to charge a premium price on the price for the same product, then the subject retailer may assign the HDTV to the business display taxonomy.

Landscapes

  • Business, Economics & Management (AREA)
  • Development Economics (AREA)
  • Strategic Management (AREA)
  • Engineering & Computer Science (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Economics (AREA)
  • Game Theory and Decision Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Marketing (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

L'invention concerne un procédé mis en œuvre au moyen d'un système informatique, qui consiste à utiliser un processus en cours d'exécution sur un processeur d'un système informatique pour accéder, à partir d'une base de données, aux informations relatives à la description du produit et aux informations relatives au prix d'un produit vendu par un premier détaillant. À l'aide d'un module de robot, les informations relatives à la description d'un produit 5 sont accessibles par l'intermédiaire d'un réseau de communication électronique pour un produit sans correspondance exacte qui est mis en vente par un second détaillant. Les règles de mise en correspondance de produit avec des paramètres qui sont réglables par produit sont mises en œuvre dans un module d'analyseur de caractéristique en cours d'exécution sur le processeur du système informatique. Une détermination est réalisée de manière automatisée pour savoir si une correspondance existe ou non entre le produit et le produit sans correspondance exacte 10, conformément à un seuil prédéfini. Lorsqu'une correspondance est déterminée conformément au seuil prédéfini, une sortie de données reliant le produit et le produit sans correspondance exacte est générée automatiquement, la sortie de données étant accessible par un processus de comparaison de prix.
PCT/CA2014/050038 2013-01-29 2014-01-21 Intelligence pour fixation des prix de produits sans correspondance exacte WO2014117268A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CA2899529A CA2899529A1 (fr) 2013-01-29 2014-01-21 Intelligence pour fixation des prix de produits sans correspondance exacte

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US201361757819P 2013-01-29 2013-01-29
US61/757,819 2013-01-29

Publications (1)

Publication Number Publication Date
WO2014117268A1 true WO2014117268A1 (fr) 2014-08-07

Family

ID=51261352

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CA2014/050038 WO2014117268A1 (fr) 2013-01-29 2014-01-21 Intelligence pour fixation des prix de produits sans correspondance exacte

Country Status (2)

Country Link
CA (1) CA2899529A1 (fr)
WO (1) WO2014117268A1 (fr)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170017969A1 (en) * 2015-07-17 2017-01-19 Overstock.Com, Inc. Methods and systems for auditing and editing content of an e-commerce site
US20220292565A1 (en) * 2019-08-22 2022-09-15 Nec Corporation Processing device, and processing method

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020095411A1 (en) * 2001-01-16 2002-07-18 Caldwell David Edward Natural language product comparison guide synthesizer
US20020107861A1 (en) * 2000-12-07 2002-08-08 Kerry Clendinning System and method for collecting, associating, normalizing and presenting product and vendor information on a distributed network
US20110264598A1 (en) * 2010-04-21 2011-10-27 Microsoft Corporation Product synthesis from multiple sources
US20120173472A1 (en) * 2007-01-26 2012-07-05 Herbert Dennis Hunt Similarity matching of a competitor's products

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020107861A1 (en) * 2000-12-07 2002-08-08 Kerry Clendinning System and method for collecting, associating, normalizing and presenting product and vendor information on a distributed network
US20020095411A1 (en) * 2001-01-16 2002-07-18 Caldwell David Edward Natural language product comparison guide synthesizer
US20120173472A1 (en) * 2007-01-26 2012-07-05 Herbert Dennis Hunt Similarity matching of a competitor's products
US20110264598A1 (en) * 2010-04-21 2011-10-27 Microsoft Corporation Product synthesis from multiple sources

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170017969A1 (en) * 2015-07-17 2017-01-19 Overstock.Com, Inc. Methods and systems for auditing and editing content of an e-commerce site
US20220292565A1 (en) * 2019-08-22 2022-09-15 Nec Corporation Processing device, and processing method

Also Published As

Publication number Publication date
CA2899529A1 (fr) 2014-08-07

Similar Documents

Publication Publication Date Title
US20140214617A1 (en) Pricing intelligence for non-identically identified products
US11238617B2 (en) Normalization of color from a digital image into a universal digital color system for specification and matching
US9436704B2 (en) System for normalizing, codifying and categorizing color-based product and data based on a universal digital color language
US20220138831A1 (en) Method of Providing Fashion Item Recommendation Service Using User's Body Type and Purchase History
JP6315524B2 (ja) 色パレットを使用した自動的な画像ベースの推奨
CN112837118B (zh) 企业用户的商品推荐方法和装置
US11727459B2 (en) Search query-based replacement part interface
JP7544355B2 (ja) 在庫およびプロファイル情報を安全管理するための方法およびシステム
US9087357B2 (en) System for normalizing, codifying and categorizing color-based product and data based on a universal digital color language
US9607404B2 (en) System for normalizing, codifying and categorizing color-based product and data based on a universal digital color system
US20170330262A1 (en) Method for Relevancy Ranking of Products in Online Shopping
WO2018129456A1 (fr) Système et procédé d'identification de produit basés sur des images
US20150379733A1 (en) Automatic image-based recommendations using a color palette
CN107862566A (zh) 一种商品推荐方法和系统
US10817522B1 (en) Product information integration
US20210383452A1 (en) Commodity recommendation system
US20180068370A1 (en) Product part picture picker
US20210056588A1 (en) Method, apparatus and computer program for providing advertising information based on images
US20130144759A1 (en) Product purchase device and product purchase method
US20220067976A1 (en) Normalized nesting cube and mapping system for machine learning to color coordinate products, patterns and objects on a homogenized ecommerce platform
US20030023492A1 (en) Method and system for collecting and processing marketing data
WO2014117268A1 (fr) Intelligence pour fixation des prix de produits sans correspondance exacte
US11170428B2 (en) Method for generating priority data for products
KR20180092053A (ko) 지능형 쇼핑 관리시스템
KR20210003706A (ko) 코디네이션 패션 아이템을 추천하는 방법

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 14746409

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2899529

Country of ref document: CA

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 14746409

Country of ref document: EP

Kind code of ref document: A1