Inplausible method based on few-sample relation prediction model

Abstract

The invention belongs to the field of knowledge maps, and particularly relates to an interpretable method based on a few-sample relation prediction model. The method comprises the following steps: evaluating the interpretability of a few-sample relation prediction model to obtain an interpretable evaluation result; improving the model according to the interpretable evaluation result; acquiring a problem from a user, and inputting the problem into the improved few-sample relation prediction model to obtain a credible prediction result of the problem; according to the method, multiple comparison models are selected for analysis, evaluation indexes are calculated by changing data volumes and data contents of the few-sample relation prediction model and the comparison models, and influences of different data volumes and different data contents on the models are analyzed; calculating evaluation indexes by changing hyper-parameters, such as an activation function, a pooling strategy and regularization, of a convolutional neural network in the few-sample relation prediction model, and analyzing the influence of the hyper-parameters on the model; the credibility of the relation prediction result of the model is improved, and the practicability is high.

Description

technical field [0001] The invention belongs to the field of knowledge graphs, and in particular relates to an interpretable method based on a few-sample relation prediction model. Background technique [0002] Knowledge graph is a combination of theories of applied mathematics, graphics, information visualization technology, information science and other disciplines with quantitative citation analysis, co-occurrence analysis and other methods, and uses a visual graph to visualize the core structure and development history of the discipline. , cutting-edge fields and the overall knowledge framework to achieve the modern theory of multi-disciplinary integration. [0003] Interpretability means having enough understandable information to solve a problem. Specifically in the field of artificial intelligence, the interpretable deep model can give the decision basis for each prediction result. For example, the search engine gives the corresponding answer according to a question,...

AI Technical Summary

Problems solved by technology

However, most neural networks are black-box models, and their internal decision-making process is difficult to understand

Without sufficient interpretability, the application of these models in some areas with high security requirements will be limited

Images