Unlock instant, AI-driven research and patent intelligence for your innovation.

A method and device for improving knowledge graph

A knowledge map and algorithm technology, applied in the field of knowledge map improvement, can solve the problems of inability to achieve accurate product search, simple and rough tags, etc., and achieve the effect of accurate recommendation and rich tags

Active Publication Date: 2021-10-26
GUANGZHOU PINWEI SOFTWARE
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] In the existing technology, the labels of products included in the product knowledge graph or user knowledge graph usually only identify the basic characteristics of the product such as brand, color, material, size, and basic style, and the labels are based on the information provided by the seller and user when uploading the product. , which are directly marked on the product after manual review. However, users are paying more and more attention to the product. If the user wants to search for products with characteristics other than the basic characteristics of the product, such as searching for a product with a specified pattern, It will lead to the inability to realize precise search for products because the knowledge map does not contain the corresponding tags; in addition, it is also impossible to realize accurate recommendations for users to purchase products based on the user's attention to the products.
[0004] Therefore, in the existing technology, there is a problem that the product labels of the knowledge graph are relatively simple and rough. It is urgent to improve the knowledge graph and enrich the labels in the knowledge graph, so as to realize the accurate search of products by users and the accurate recommendation of products purchased by users.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A method and device for improving knowledge graph
  • A method and device for improving knowledge graph
  • A method and device for improving knowledge graph

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0071] The embodiment of the present invention provides a method for improving knowledge graph, refer to figure 1 As shown, the method includes:

[0072] 101. Extract the label of the product from the product information at least including the product image.

[0073] Specifically, the product information only includes the product image, and the product label is extracted from the product information including the product image, and the process may include:

[0074] Extract image features from product images;

[0075] Get labels corresponding to image features.

[0076] The product information includes a product image and text description information of the product, and the product label is extracted from the product information including at least the product image. The process may include:

[0077] Extract image features from product images; and

[0078] Extract keywords in text description information;

[0079] Determine the label of the product according to the image fe...

Embodiment 2

[0085] The embodiment of the present invention provides a method for improving knowledge graph, refer to figure 2 As shown, the method includes:

[0086] 201. Extract a product label from product information at least including a product image.

[0087] Among them, the products in the product image can be clothing, food, cosmetics, furniture, daily necessities or other products.

[0088] The label of the product is used to describe the product. The label can identify the product name, brand, size, color, style, version, pattern, or be used to indicate the definition of the general or summary feature range in the product. The style can include "original". "Wind", "military style", "retro style", "sweet and cute", etc.; used to indicate the definition of the range of general or summary features in the product, such as "thin", "bright color", etc.

[0089] In addition, the label may also be other information that can be used to identify the commodity, and the embodiment of the ...

Embodiment 3

[0150] The embodiment of the present invention provides a method for improving knowledge graph, refer to image 3 As shown, the method includes:

[0151] 301. Extract a product label from product information at least including a product image.

[0152] Specifically, this step is the same as step 201, and details are not repeated here.

[0153] In the embodiment of the present invention, since the label of the product is extracted from the product information including the product picture and text description information, more information and characteristics of the product can be reflected, so the label of the product is further enriched, diversified, and refined. .

[0154] 302. Add the extracted tag to the tag library.

[0155] Specifically, when it is recognized that the extracted tags are not included in the tag library of the knowledge graph, the extracted tags are added to the tag library.

[0156] The present invention does not limit the specific determination proces...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a knowledge map perfecting method and device, belonging to the technical field of electronic commerce. The method includes: extracting a product label from product information including at least a product image; adding the extracted label to a knowledge graph, where the knowledge graph includes a product knowledge graph and / or a user knowledge graph; setting a label corresponding to the label in the knowledge graph relation. The embodiment of the present invention can enrich the tags in the knowledge graph, so as to realize the user's accurate search for commodities and realize the accurate recommendation of the user's purchased commodities.

Description

technical field [0001] The invention relates to the technical field of electronic commerce, and in particular, to a method and device for perfecting a knowledge map. Background technique [0002] Knowledge graph is a graph-based data structure. Its nodes represent entities or concepts, and edges represent various semantic relationships between entities / concepts. By using knowledge graphs in search engines, it can better It can query complex related information, understand user intent from a semantic level, and improve search quality. For example, in the field of e-commerce, using the product knowledge graph can help users search for corresponding products, and using the user knowledge graph can help recommend products to users. [0003] In the prior art, the labels of the commodities included in the commodity knowledge graph or the user knowledge graph usually only identify the basic characteristics such as the brand, color, material, size, and basic style of the commoditie...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06F16/36G06Q30/06
CPCG06Q30/0631G06F16/367
Inventor 徐然崔燕红张智祺黄惠燕郭安琪
Owner GUANGZHOU PINWEI SOFTWARE