Knowledge dynamic evolution method based on multi-time granularity and application

Abstract

The invention discloses a knowledge dynamic evolution method based on multi-time granularity and application. The dynamic knowledge evolution method comprises the following steps: firstly, constructing knowledge tetrads in a knowledge base, then vectorizing each knowledge tetrad and splitting each knowledge tetrad into a triple corresponding vector and a time vector, then carrying out initialization representation on the triple corresponding vectors by utilizing a convolutional network, and carrying out granularity unification on the time vectors; embedding the time vector with uniform granularity into a representation vector corresponding to the triad, and inputting the representation vector into a triad representation method for representation; and finally, knowledge tetrad mining is carried out, a regression model is constructed in combination with all entity representation information and periodic historical entity representation information, entities of multiple time steps in the future are predicted, and dynamic evolution of knowledge tetrad is realized. According to the knowledge dynamic evolution method based on the multiple time granularities, the knowledge of the multiple time granularities is fused for characterization, the characterization capacity of the model can be improved, and the entity at the future moment can be better predicted.

Description

technical field [0001] The invention relates to the technical field of artificial intelligence and knowledge graph, in particular to a knowledge dynamic evolution method and application based on multi-time granularity. Background technique [0002] With the advent of the 5G era, the data in cities has exploded. The collection, cleaning, and modeling analysis of high-dimensional, heterogeneous, and multi-modal data in cities can provide important guidance for solving pain points and difficult problems in cities. significance. For example, using deep learning algorithms to predict the flow of people and vehicles in a city can better plan the traffic in the city; predict the air quality and water quality in the city, which can better manage the urban environment. Cities have the characteristics of wide areas, many people, and large regions. At every moment, a large amount of multi-source heterogeneous data is generated. Therefore, a large amount of urban time series data will ...

AI Technical Summary

Problems solved by technology

However, these methods are based on knowledge quadruples of the same time granularity. They cannot represent knowledge quadruples with multiple time granularities in a knowledge base, resulting in a lot of important information that cannot be effectively integrated into the model. Among them, the accuracy of the model to predict entities or relationships in the future is low, and it cannot accurately capture the dynamic evolution of knowledge

However, in the real world, when constructing a city-related knowledge base, knowledge of multiple time granularities generally exists, and the existing technology is still unable to characterize the knowledge of multiple time granularities.

Images