Causal event map construction method, system and device and storage medium

Abstract

The invention discloses a causal event map construction method, system and device and a storage medium, and relates to the field of natural language processing big data analysis, comprising the following steps: constructing a causal event mode knowledge base; Automatically performing causal event extraction on the input text according to the constructed causal event mode knowledge base; Performingstructured representation on the reason event statement and the result event statement which are obtained after extraction; fusing The causal event relation pairs which complete the structured representation; And storing the fused causal event data into a graph database. According to the method, manual intervention is only needed during quality control, the knowledge base is constructed in a linguistic expert arrangement mode, the event representation method can more completely represent the event, and the sentiment analysis technology is fused into event representation, so that information on the other dimension of the event can be increased. The graph database is used for storing the causal event graph, the flexibility is high, and deep causal event query can be supported.

Description

technical field [0001] The present invention relates to the field of natural language processing big data analysis, in particular to a method, system, device and storage medium for constructing a causal event map. Background technique [0002] Today's era is an era of information explosion. A large number of unstructured texts are generated every day, and a large number of objective facts or logical laws are hidden in these unstructured texts, that is, the logical world in real life is depicted internally. For example, the phrase "auspicious snow heralds a good harvest" describes a relationship between heavy snowfall and the next year's harvest, that is, "there is heavy snowfall" and "the next year's harvest will be good". A rule summed up, there are many similar events, such as: earthquakes will cause houses to collapse, and economic crises will lead to a country's economic downturn. [0003] Some of the above examples are basic common sense and some are professional knowl...

AI Technical Summary

Problems solved by technology

[0007] 1) The causal event graph is still a relatively new concept, and the construction of the event graph is still in its preliminary stage;

[0008] 2) Causal event extraction based on templates and rules is still the mainstream method. The formulation of models and rules requires a lot of labor. The model of "XX leads to XX" is too limited and needs to be expanded as much as possible;

[0009] 3) The problem of expressing causal events has not been well solved yet. There are still some problems in the true representation of event semantics by using short sentences in the original text, or by taking adjectives, verbs, and nouns as event representations;

Images