Academic influence cooperative sequencing method of nodes in scientific and technical literature heterogeneous network

Abstract

The invention discloses an academic influence cooperative sequencing method of nodes (papers, authors and periodicals / conference) in a scientific and technical literature heterogeneous network. By aiming at the heterogeneity of the scientific and technical literature network, according to the method provided by the invention, two stages of random walk are adopted for calculating and sequencing the academic influence of the nodes in the scientific and technical literature, wherein the two stages of the random walk are respectively the random walk driven by the node network topology on the scientific and technical literature heterogeneous network and the random walk driven by the academic type in a minimum network mode of the scientific and technical literature network; the weight of each edge is calculated in the scientific and technical literature heterogeneous network for configuring the random walk driven by the node network topology; and according to the minimum network mode of the scientific and technical literature heterogeneous network, an average edge betweenness of various relation edges in different fields is counted, a propagation factor is calculated, and the rand walk driven by the node academic type is configured. The method provided by the invention solves the problems caused by the heterogeneity of the heterogeneous network and avoids phenomena of node influence and in-degree positive correlation, so that the sequencing result is relatively accurate.

Description

technical field [0001] The invention provides a collaborative ranking method for academic influence of nodes (papers, authors, periodicals / conferences) in a heterogeneous network of scientific and technological documents, belonging to the field of information retrieval. Background technique [0002] Sci-tech literature networks usually contain three different types of academic nodes: authors, papers, and journals / conferences. There are also connection edges with different relationships between different types of academic nodes, such as: author-paper author and author relationship, paper-paper The relationship between citation and citation, conference-paper publication and published relationship, author-author cooperation and citation relationship. [0003] In general, an information network can be characterized by a directed graph G=(V,E), node type function τ:V→A and connection edge type function φ:E→R. Among them, the node type τ(v)∈A of each node v∈V; the edge type φ(e)∈...

AI Technical Summary

Problems solved by technology

[0006] 2. Treat the scientific literature network as a heterogeneous network, and sort the three types of nodes in the scientific literature network at the same time. Deng, Z.H, Lai B Y et al. proposed the PAV model to sort the three types of nodes at the same time. In this method The literature network includes the indirect author-author cooperative relationship, which can be composed of the author-paper writing relationship and the authoring relationship. The disadvantage is that its method of dealing with network heterogeneity is only to define different types of edges. weight, but when this weight is defined, it subconsciously compares the strength of the relationship between the same type of edges, and does not handle the relationship between different types of edges very well, and there will be a phenomenon that the influence of nodes is related to the in-degree, that is Authors and journals with more published papers will be ranked higher

Images