Entity description reasoning knowledge base construction and reasoning evidence quantitative information acquisition method and device

Abstract

The invention relates to an entity description reasoning knowledge base construction and reasoning evidence quantification information acquisition method and device. The method comprises the following steps that an entity description reasoning knowledge base is established by utilizing a large-scale unstructured open text, wherein the entity description reasoning knowledge base comprises an entity description knowledge base and an entity association conduction base; aiming at the input event or event description list and the event pair or event description pair list, retrieval is carried out in the entity description reasoning knowledge base through entity linkage, and reasoning evidences and conduction intensity between the events are returned. According to the method, the range of an existing logical reasoning knowledge base is widened, and the logical reasoning ability of the existing knowledge base can be improved; the method can be flexibly applied to different reasoning scenes such as single-event reasoning and double-event reasoning, and the obtaining mode can be quickly changed according to actual requirements.

Description

technical field [0001] The invention relates to a method and device for entity description reasoning knowledge base construction and reasoning evidence quantification information acquisition, which belongs to the field of natural language processing and belongs to a specific category of knowledge base construction and application reasoning methods. Background technique [0002] Knowledge reasoning is an advanced stage of artificial intelligence. Based on existing knowledge, knowledge reasoning machine technology is used to complete decision-making behaviors in limited areas, which can generate economic benefits while fully reducing human labor. For example, knowledge reasoning based on known knowledge and knowledge discovery using event-driven conduction paths can assist business reasoning and decision-making, conduct unknown risk warnings in smart investment research, and control and control public opinion on companies in public opinion analysis. monitor. [0003] In order...

AI Technical Summary

Problems solved by technology

The existing public logical reasoning knowledge base mainly focuses on the knowledge base construction of event evolution logic (such as causal evolution, conditional evolution, etc.), industrial chain conduction logic and related extensions, and the construction of other types of logical reasoning knowledge bases needs to be further developed to dig

[0006] 2. There is a lack of data in the existing event reasoning, conduction and quantification

[0007] 3. The lack of display of evidence for the conduction of the existing event reasoning

In the process of interpretable display of current event reasoning, each inferable node between events and the edges between nodes are directly used as elements to draw graphs, which is not effective enough to display the source of empirical evidence for conductive edges.

This lack makes it difficult to guarantee the credibility and reliability of reasoning

[0008] 4. Insufficient exploration and application of entity logic

Images