Construction method of commodity knowledge graph

Abstract

The present invention discloses a construction method of a commodity knowledge graph, which comprises objective commodity classification knowledge construction and subjective user viewpoint knowledge construction of the commodity knowledge graph. The subjective knowledge construction comprises a process of mining viewpoint objects and viewpoint words from user comments by using a method based on multi-classifier integration, and a process of directly obtaining URL information of viewpoint holders, viewpoint express time and / or viewpoints from websites. The objective knowledge construction comprises a process of obtaining commodity classification hierarchies combined with depth and breadth by using a method based on multi-source heterogeneous classification hierarchy fusion, and a process of directly obtaining commodity instances from the websites. The structured knowledge graph constructed for commodities covers the objective commodity classification knowledge and the subjective user viewpoint knowledge, accurate interpretation of the commodity instances can be provided for upper layer service, viewpoint description about the commodities and attributes thereof by multiple users can also be understood, and therefore more rapid, accurate and comprehensive commodity knowledge service is provided for the users.

Description

technical field [0001] The invention belongs to the technical field of databases, and in particular relates to a method for constructing a commodity knowledge map. Background technique [0002] With the rapid development and popularization of network technology, the Web has become one of the largest knowledge bases in fact. However, knowledge on the Web exists in an unstructured or semi-structured form, which greatly restricts their automatic and intelligent application. Reorganizing the knowledge contained on the Web and transforming it into structured information has received a lot of attention in recent years. For example, Google integrates multiple open source knowledge bases (such as Freebase and Wikipedia) and its search data to build a knowledge map, and Microsoft released Satori, both of which are committed to realizing the knowledge structure of strings by structuring the knowledge in web pages. Semantic matching improves search quality. In academia, Yago develop...

AI Technical Summary

Problems solved by technology

At present, there are two main problems in the application of user opinion information in comment data: (1) the information overload problem caused by massive comment data; this problem makes it difficult for users to accurately and comprehensively obtain the user opinion information they expect ; (2) User opinion is hidden in unstructured text, which seriously hinders the automatic and intelligent application of user opinion information

Images