Complex question and answer query method and device based on time sequence knowledge graph

Abstract

The invention discloses a complex question and answer query method and device based on a time sequence knowledge graph, and the method comprises the steps: obtaining a time sequence question, and inputting the time sequence question into a trained question and answer system model; wherein the question and answer system model comprises a time sequence knowledge embedding model, an embedding extraction model, a time change model and an answer scoring model; obtaining entity embedding and time embedding in the time sequence problem by utilizing the time sequence knowledge embedding model, and obtaining potential embedding of the time sequence problem by utilizing the embedding extraction model; and transforming the time embedding into the target time of the time sequence problem by using the time variation model; and scoring the entity embedded by the entity and the target time by using an answer scoring model, and outputting the entity or time with the highest score as an answer to the time sequence question. According to the method, potential embedding is converted into the target time of the problem, the processing capacity of the complex time sequence problem can be improved, and the processing process of the complex problem is simplified.

Description

technical field [0001] The invention relates to the technical field of information technology and data services, and in particular, to a complex question-answering query method and device based on a time-series knowledge graph. Background technique [0002] A time-series knowledge graph is a collection of structured events with time attributes, which can be represented as a set of quadruplets G={(s,r,o,t)}, where s represents the head entity, o represents the tail entity, and r represents the relationship, t represents time, such as (Messi, Award, World Footballer, 2009). The time-series knowledge graph can be regarded as a knowledge base, and the question-answering system based on the time-series knowledge graph can make it easier for people to obtain the knowledge inquired, that is, to ask questions to the question-and-answer system in natural language without writing professional query sentences. [0003] At present, the mainstream time series knowledge graph question an...

AI Technical Summary

Problems solved by technology

A complex question-answering query method based on time-series knowledge graphs is proposed. First, the potential time and entity embeddings in complex questions are extracted, and then the neural network related to entities, relationships, and question types is used to infer changes in time information in questions, so that Transform latent embeddings to the target time of the problem, while solving complex problems in the same way as simple problems, that is, using temporal graph embeddings to answer simplified queries without cumbersome multi-step query or sub-problem decomposition

Images