WO2020000201A1

WO2020000201A1 - Commodity search system based on e-commerce platform

Info

Publication number: WO2020000201A1
Application number: PCT/CN2018/092924
Authority: WO
Inventors: 陈钦鹏
Original assignee: 深圳齐心集团股份有限公司
Priority date: 2018-06-26
Filing date: 2018-06-26
Publication date: 2020-01-02

Abstract

A commodity search system based on an E-commerce platform, comprising a commodity attribute labelling module (1), a commodity similarity identification module (2), a summarisation and intelligent determination module (3) and a pushing module (4). The commodity attribute labelling module (1) is used for conducting commodity attribute labelling on each commodity in the E-commerce platform according to a definition rule of commodity attribute labelling; the commodity similarity identification module (2) is used for creating, according to each commodity attribute labelling of the commodity, a chain of commodities similar to the commodity; and the summarisation and intelligent determination module (3) is used for obtaining a commodity attribute identification retrieval expression by means of analysis according to a search condition and the definition rule of commodity attribute labelling, filtering out commodities that satisfy a retrieval condition and summarising and pushing same. In the system, the commodity attribute labelling module (1) is used for attribute labelling of a commodity, which makes it possible to obtain the accurate positioning of the commodity, and the retrieval accuracy thereof is has very high.

Description

Product search system based on e-commerce platform

Technical field

The present invention relates to an e-commerce system, and more particularly, to a product search system based on an e-commerce platform.

Background technique

In the prior art, the product search of an e-commerce platform only uses a simple text to perform a product search, or further enhances the technology, and uses pictures to perform a product search. The technical methods of these product search often require customers to You can select the right product after browsing; it is extremely time-consuming. With the advancement of technology, this search method will be eliminated; the current product search technology in e-commerce transactions has the technical drawback of searching for suitable products and taking a long time. Technical problems urgently to be solved by those skilled in the art.

Summary of the invention

The technical purpose of the present invention is to overcome the technical problem that the product search in the prior art can only be performed by simple text or pictures, which takes a lot of time to identify the product; and provides a method for attribute search on the product, which has a fast search speed and accurate positioning. Product search system based on e-commerce platform.

To achieve the above technical objectives, the technical solution of the present invention is: a product search system based on an e-commerce platform, including a product attribute labeling module, a product similar identification module, a summary and intelligent judgment module, and a push module;

The product attribute labeling module is configured to label each product in the e-commerce platform according to the definition rules of the product attribute label;

The commodity similar identification module is used to mark each commodity attribute of a commodity, build a commodity chain similar to the commodity, and add a corresponding commodity similar identifier to the commodity and the similar commodity according to each commodity attribute;

The summary and intelligent judgment module is configured to analyze and obtain the product attribute identifier search formula according to the search conditions and the definition rules of the product attribute label, and search in the product attribute label and similar identifiers, and filter out the products that meet the search conditions and perform Aggregate push

The push module pushes the closest similar products that have been selected to meet the search conditions to the user interface, and performs intelligent sorting.

Further, the product attribute labeling includes at least one or more of the category to which the product belongs, the name of the product, the product store, the price of the product, and the sales volume of the product, wherein the category of the product is classified according to the function and use of the product, and the classified category is from the superior To the lower level, it is divided into multiple general categories, one or more sub-categories arranged below the general category, and one or more sub-categories arranged on the facets of the small categories in a tree-like distribution.

Further, each general category, sub-category, and sub-category respectively corresponds to an identifier, and the identifier of the sub-category is the general category identifier + the sub-category serial number, wherein the sub-category serial number is defined according to the level and order of the sub-category And the number of digits of the subclass serial number is 2 ⁿ , n is the class to which the subclass belongs; the identifier of the subclass is the general class ID + the subclass serial number + the subclass serial number, where the subclass serial number is based on the subclass's class level and its It is defined by ordering, specifically, the number of digits of the subclass serial number is 2 ^m , and m is the level to which the subclass belongs.

Further, the product attribute is marked as [identification of category to which product belongs, identification of product name, identification of product store, identification of price level, identification of sales level].

Further, the price level identifier is divided according to the commodity price and the price, and the sales level identifier is divided into the commodity sales and the sales volume.

Further, the similar identifier of the product is [identical identifier of product category, similar identifier of product name, similar identifier of product store, similar identifier of price level, similar identifier of sales level].

Further, the product categories of the same or similar products belonging to the same category are similarly identified by the same identifier; the products of the same name or similar products are similarly identified by the same identifier; the product stores of the same product are similarly identified by the same identifier.

Further, the products of the same or similar category and the same price level are marked with the same price level similar mark, and the products of the same or similar category and the same sales level are marked with the same sales level similar mark.

Further, the aggregation and intelligent judgment module first determines which one or more of the category to which the product belongs, the product name, the product store, the price level, and the sales level.

Further, the summary and intelligent judgment module obtains the search criteria of the product attribute identifier according to the definition rules of the product attribute labeling based on the search conditions, and then retrieves the products that satisfy the search conditions according to the search criteria of the product attribute identifier search. Products that are similar to this product, and products that meet the search conditions are filtered out and pushed together.

The product search system based on the e-commerce platform of the present invention has the following beneficial effects:

1. The present invention uses a product attribute labeling module to mark attributes of a product. According to the product attribute labeling module, accurate positioning of the product can be obtained, and the retrieval accuracy is very high.

2. The present invention uses a product attribute labeling module to similarly identify a product and build a product chain similar to the product, which is helpful for recommending products with similar search conditions to customers. Among them, there are multiple different product chains similar to the product. The commodity chain of attributes can meet the multi-faceted needs of customers.

3. The present invention does not perform a product search through simple text or pictures, but a method of attribute identification search, which consumes less time than the traditional simple text or picture search for goods.

Additional aspects and advantages of the present invention will be given in part in the following description, part of which will become apparent from the following description, or be learned through the practice of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 is a block diagram of the modular structure of the invention;

Among them, 1. Product attribute labeling module; 2. Product similar identification module; 3. Summary and intelligent judgment module; 4. Push module;

detailed description

Hereinafter, embodiments of the present invention will be described in detail. Examples of the embodiments are shown in the drawings, wherein the same or similar reference numerals indicate the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the drawings are exemplary and are intended to explain the present invention, but should not be construed as limiting the present invention.

In one embodiment of the present invention, as shown in FIG. 1, a product search system based on an e-commerce platform includes a product attribute labeling module 1, a product similar identification module 2, a summary and intelligent judgment module 3, and a push module 4.

The product attribute labeling module 1 is configured to label each product in the e-commerce platform according to the definition rules of the product attribute label.

The product attribute labeling includes at least one or more of the category to which the product belongs, the name of the product, the product store, the price of the product, and the sales volume of the product. The category of the product is classified according to the function and use of the product, and the categories are classified from higher to lower levels. Divided into multiple general categories, one or more sub-categories placed below the general category, and one or more sub-categories placed on the facets of the small categories in a tree-like distribution, and each general category and sub-category The subclasses and subclasses each correspond to an identifier, and the subclass identifier is the general class identifier + the subclass serial number. The identification of the subclass is the general class ID + the subclass serial number + the subclass serial number.

The product name includes the general name of the product and the approximate names of multiple products. The identifier of the product name is the common name of the product / approximate name of each product.

Each commodity store also corresponds to an identifier, and the identifier of the commodity store may be the website address, IP address, name of the commodity store, other code corresponding to the commodity store, and the like.

The price of the product also corresponds to a logo. The logo of the price of the product is divided according to the price and is set as the price level logo.

The sales volume of a product also corresponds to a logo. The sales volume of a product is classified according to the sales volume and is designated as the sales level logo.

The product attribute of the product is marked as [category identification of the product, product name identification, product store identification, price level identification, sales level identification].

For example, the definition rule of product attribute labeling is that the general category in the category to which the product belongs includes daily necessities, clothing, home appliances, home improvement, food, and so on. Among them, the logo of daily necessities is NE, the logo of clothing is CL, the logo of home appliances is EA, the logo of home improvement is FC, and the logo of food is FD. The identification of the small class is the general class ID + the serial number of the small class, where the serial number of the small class is defined according to the rank of the small class and its order. Specifically, the number of digits in the serial number of the small class is 2 ⁿ , and n is the rank of the small class. . The identification of the subclass is the general class identification + the subclass serial number + the subclass serial number, where the subclass serial number is defined according to the level and order of the subclass, specifically, the number of digits of the subclass serial number is 2 ^m , where m is The level to which the subclass belongs.

It should be noted that the definition rules of the above-mentioned product attribute labels may not be limited to such definition rules, and may be defined by themselves as needed.

The following is an example to further illustrate the product attribute labeling of the product. Assume that the product attribute labeling of a product is [NE / NE02 / NE0201, broom / broom, IP1, A, E], where NE / NE02 / NE0201 is of the category to which the product belongs. Logo, expressed as the first sub-class under the second sub-category under the daily necessities category, where NE is the logo of the daily necessities in the general category, NE02 is the logo of the second sub-category under the general categories, and NE0201 is under the general The identification of the first level sub-category below the second sub-category; broom / broom is expressed as a product name identification, where broom is the general name of the product and the broom is an approximate name of the product; IP1 is the product store identification; A is the price level identification ( That is, the product price level is A level); E represents the sales level identification (that is, the product sales level is E level).

The product similarity identification module 2 is used for labeling each product attribute of a product, constructing a similar product chain to the product, and adding a corresponding product similar identifier to the product and its similar product according to each product attribute.

The similar product ID is [the similar product type, the product name is similar, the product store is similar, the price level is similar, and the sales level is similar.] Among them, the products belonging to the same category or similar products have similar product categories with the same identifier and are connected through a similar product chain of product categories. Products with the same or similar product names have similar product name similar identifiers, and are connected by a product name similar product chain. Products with the same product store have the same product store with similar identifiers, and are connected through the product store with similar product chains. Products with the same or similar categories and products with the same price class are marked with the same price class and similar identifiers, and pass similar product price levels. Similar product chain connections. Products with the same or similar product category and products with the same sales level are marked with the same sales level and similar identifiers, and are connected by similar product chains with similar product sales levels.

From this, multiple commodity chains with different attributes can be established, that is, one or more of the commodity chains that have similar identifiers for product categories, similar identifiers for product names, similar identifiers for product stores, similar identifiers for price levels, and similar identifiers for sales levels. For example, The search condition is simultaneous search of the product name and price. During the search, you can immediately obtain through the product chain at the same time products that meet the corresponding product category similar identifiers and corresponding price level similar identifiers. The positioning is very accurate, and the retrieval speed is fast and efficient. high. Aggregation and intelligent judgment module 3, which is used to analyze and obtain the product attribute identification search formula according to the search conditions and the definition rules of the product attribute labeling, and search in the product attribute labeling and similar identifiers, and select and summarize the products that meet the search conditions. Push.

Specifically, the type searched by the search condition is determined according to the type of the search condition, that is, which one or more of the category to which the product belongs, the product name, the product store, the price level, and the sales level is determined.

Then, the search conditions are based on the definition rules of the product attribute label to obtain the product attribute identification search formula, and then the product that meets the search conditions is retrieved according to the product attribute identification search formula. At the same time, the product similar to the product is obtained according to the similar identifier of the product, and Products that meet the search conditions are filtered out and pushed in aggregate.

For example, the search condition is "broom", which is obtained according to the above-mentioned definition rules of product attribute labeling. The broom belongs to the first-level sub-category of the second sub-category below the general category of daily necessities, and the identity of the category to which the product belongs is NE / NE02 / NE0201 is searched by the aggregation and intelligent judgment module according to its product attribute identifier search formula NE / NE02 / NE0201 to obtain all the products that meet the category ID of the product as NE / NE02 / NE0201, and then push them collectively.

The push module 4 pushes the closest similar products that have been selected to meet the search conditions to the user interface and intelligently sorts them.

In the present invention, a product attribute labeling module is used for attribute labeling of a product. According to the product attribute labeling module, accurate positioning of the product can be obtained, and the retrieval accuracy is very high.

The present invention uses a product attribute labeling module to similarly identify a product and build a product chain similar to the product, which is helpful for recommending products with similar search conditions to customers. The product chain similar to the product has multiple different attributes. Commodity chain can meet customers' multiple needs.

Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limitations on the present invention. Those skilled in the art will not depart from the principles and purposes of the present invention. In the case of the present invention, changes, modifications, replacements, and variations can be made to the above embodiments within the scope of the present invention. The scope of the invention is defined extremely equally by the appended claims.

Claims

A product search system based on an e-commerce platform, which is characterized by including a product attribute labeling module, a product similar identification module, a summary and intelligent judgment module, and a push module;

The product attribute labeling module is configured to label each product in the e-commerce platform according to the definition rules of the product attribute label;

The product similar identification module is configured to mark a product chain similar to the product according to each product attribute label of the product, and add a corresponding product similar label to the product and the similar product according to each product attribute label;

The summary and intelligent judgment module is configured to analyze and obtain the product attribute identifier search formula according to the search conditions and the definition rules of the product attribute label, and search in the product attribute label and similar identifiers, and filter out the products that meet the search conditions and perform Aggregate push

The push module pushes the closest similar products that have been selected to meet the search conditions to the user interface, and performs intelligent sorting.
The product search system based on the e-commerce platform according to claim 1, characterized in that the product attribute labeling includes at least one or more of a category to which the product belongs, a product name, a product store, a product price, and a product sales amount, wherein the product The category is classified according to the function and purpose of the product. The categories are divided into multiple general categories, one or more sub-categories below the general category, and level 1 on the small facets. Or multi-level subclasses, distributed in a tree-like manner.
The product search system based on the e-commerce platform according to claim 2, characterized in that each general category, sub-category, and sub-category corresponds to an identifier, and the sub-category identifier is the general category identifier + the sub-category serial number, Among them, the subclass serial number is defined according to the class to which the subclass belongs and its order, and the number of digits of the subclass serial number is 2 n , where n is the class to which the subclass belongs; the identifier of the subclass is the general class identifier + the subclass serial number + The subclass serial number, where the subclass serial number is defined according to the subclass's rank and its order. Specifically, the number of digits of the subclass serial number is 2 m , and m is the rank of the subclass.
The product search system based on the e-commerce platform according to claim 1, characterized in that the product attributes are marked as [category identification of product, product name identification, product store identification, price level identification, sales level identification].
The product search system based on the e-commerce platform according to claim 4, characterized in that the price level indicator is divided according to the price of the product and the sales level indicator is divided into the sales according to the sales volume of the product.
The product search system based on the e-commerce platform according to claim 1, characterized in that the similar identifiers of the products are [product category similar identifiers, product name similar identifiers, product store similar identifiers, price level similar identifiers, and sales levels Similar ID].
The product search system based on the e-commerce platform according to claim 6, characterized in that the product category similar identifiers of the products of the same or similar category to which the products belong are the same; the product name similar identifiers of the products having the same or similar product names are the same The same store with the same product has the same logo.
The product search system based on the e-commerce platform according to claim 6, characterized in that the products belonging to the same or similar category and the products of the same price level are marked with the same price level similar identifier, and the products belonging to the same or similar category belong to the same category. Products with the same sales level are marked with similar logos for the same sales level.
The product search system based on the e-commerce platform according to claim 1, wherein the aggregation and intelligent judgment module first determines which of the category the product belongs to, the product name, the product store, the price level, and the sales level One or more.
The product search system based on the e-commerce platform according to claim 1, characterized in that: the aggregation and intelligent judgment module obtains a product attribute identifier search formula based on the search criteria according to the definition rule of the product attribute label, and then searches according to the product attribute identifier The products obtained by the search can be obtained by the formula search, and the products similar to the products are obtained according to the similar identification of the products, and the products that meet the search conditions are screened out and pushed together.