首先，一种从模式丰富的异质信息网络中提取元路径的通用框架被设计出。 在

First, a general framework for extracting meta-paths from heterogeneous information networks with rich patterns is designed. In the complex heterogeneous information network, a network model based on local information is constructed to replace the original network model, which solves the problem that the network model in the model-rich heterogeneous information network is difficult to apply. <br>In addition, a novel node type evaluation function that integrates the degree of relevance and support is designed to automatically select the type.

First, a common framework for extracting metapaths from pattern-rich heterogeneity information networks is designed. Under the complex heterogeneity information network, a network mode based on local information is constructed to replace the original network mode, which solves the problem that the network mode in the heterogeneic information network is difficult to apply.<br>In addition, a novel node type evaluation function with comprehensive correlation and support degree is designed to automate the selection of types.

Firstly, a general framework for extracting meta paths from heterogeneous information networks with rich patterns is designed. In the complex heterogeneous information network, a network model based on local information is constructed to replace the original network mode, which solves the problem that the network mode in heterogeneous information network with rich patterns is difficult to apply.<br>In addition, a novel node type evaluation function is designed, which can automatically select the type<br>