Method and device for establishing event repository

Abstract

The invention provides a method and a device for establishing an event repository. The method comprises the following steps of recognizing a text to make out an event excavation candidate statement; segmenting the recognized event excavation candidate statement to obtain a syntax fragment; generating event knowledge instances according to the syntax fragment and the pre-defined event knowledge structure, wherein the quantity of the event knowledge instances is identical to that of verb-object fragments and subject-predicate fragments contained in the syntax fragment; obtaining the corresponding event excavation target statement according to the verb-object fragments and the subject-predicate fragments contained in the syntax fragment, classifying the components of the obtained event excavation target statement, and writing the classified component into the event knowledge instance so as to complete the establishment of the repository. The event repository can be established through the pre-defined event knowledge structure, an accurate search result can be provided for the users on the basis of the event repository, and the search experience of the user can be improved.

Description

technical field [0001] The invention relates to the field of computer technology, in particular to a method and device for constructing an event knowledge base. Background technique [0002] Search engine technology is undergoing a revolution around the knowledge graph. The core of the revolution is to shift from the traditional unstructured web text-oriented to the new generation of knowledge search oriented to "structured knowledge", so as to provide users with Provide more accurate search results and improve its search experience. At present, the core technology of knowledge map technology (knowledge map in the narrow sense) includes the entity knowledge base containing entity knowledge (hereinafter referred to as the entity base) and its related application technologies. [0003] In the entity library, an entity knowledge is mainly composed of its unique semantic number, that is, semantic identification (ID), attributes and attribute values ​​with clear semantic meaning...

AI Technical Summary

Problems solved by technology

[0005] Taking the deep question-and-answer application as an example, if a user asks questions such as "Which subsidiaries did Baidu acquire in 2013?" Traditional entity library knowledge cannot give the answer

[0006] It can be seen that when performing in-depth knowledge search based on traditional entity databases, it is often impossible to provide users with search results, thereby reducing the user's search experience.

Images