Publication clustering method based on literature citation relationship

Abstract

The invention relates to a publication clustering method based on a literature citation relationship. The correlation between publications is calculated by collecting the paper information published in the publications and the citations between the papers, the citation relationship of T 'paper-paper' is converted line by line to get the relationship between T line 'publication-publication', basedon this, the relationship W between n publications is obtained, the relationship network with the publications as nodes is established, and spectral clustering is used for performing publication dividing on the relationship network W with the publications as the nodes and iterative clustering on the publications to obtain clustering results. The method adopts data based on citation relationship between a large number of papers to calculate the relationship between the publications, and then uses a relationship-based clustering algorithm to obtain groupings between the publications; compared with manual classification, in the method, the proposed classification based on citation relationship can more objectively and timely reflect the relationship between the research directions of variouspublications, so that the obtained grouping results have objectivity and timeliness.

Description

technical field [0001] The invention belongs to the technical field of digital computing equipment or data processing equipment or data processing method especially suitable for specific functions, and particularly relates to a publication clustering method based on document citation relationship for data mining to classify publications. Background technique [0002] In the process of academic research by the majority of scientific researchers, academic journals are an important resource for presenting research trends in professional fields and guiding scientific research work. Researchers generally choose to pay attention to corresponding journals according to their field classification, such as reading papers in the journal and submitting papers to the journal. [0003] In the prior art, the classification of publications based on different fields is mainly based on manual classification. Generally, the research fields involved in the publications are divided according to ...

AI Technical Summary

Problems solved by technology

[0005] The technical problem solved by the present invention is that in the prior art, the classification of publications based on different fields is mainly based on manual classification, which leads to the continuous emergence of new popular research directions and interdisciplinary research as the research work progresses. The emergence of journals requires timely adjustments and updates in the field of publications. The traditional manual classification method has insufficient consideration and time lag, which is not conducive to the development of scientific research work that requires timeliness and innovation. An optimized publication clustering method based on literature citation relationship

Images