Design And Implementation Of Alzheimer's Disease Document Visualization Platform

Alzheimer's disease has so far no effective treatment,and its fatality rate is extremely high.In recent years,the research on AD has gradually warmed up,the number of AD related documents is as high as 120,000.How to obtain useful information for scientific research workers and medical researchers from massive literature has become crucial.In order to learn its research trends from a large amount of AD related literature,it is the main purpose of this thesis to design and implement a visualization platform that can clearly show the trends of AD literature.In order to understand the dynamics of Alzheimer's disease,the relevant literature was analyzed and tracked.The visualization platform of this thesis is devoted to the research of relative trends of AD literature,involving Scientometrics,LDA topic models,DEA model,and visualization technologies.The main research contents of this article mainly include the following three aspects.(1)This article has studied the visualization platform technology and means.Among them,scientometrics involves the analysis of scientific knowledge structure and the evolution of scientific knowledge.This thesis analyzes the evolution of scientific knowledge based on literature,selects the currently popular topic model for in-depth discussion and research,and studies the DEA model.(2)Researched and implemented the AD visualization system.According to the actual needs of the system,this article designed the overall architecture of the AD visualization platform,and completed the design and implementation of major modules such as data collection,database design,data analysis,and platform visualization in the system.(3)Based on the method of metrology,Biopython was used to obtain the AD related documents of NCBI website,and the visual trends of the relative trends of the number of documents,the amount of national documents issued,the number of journals cited,and keywords were analyzed.Based on the DEA model,a scientific analysis of the change trend was conducted.Based on the theme evolution model established by the LDA theme model,the AD related documents are divided according to time,the hidden themes are obtained using the LDA model,the parameter estimation is performed using the Gibbs algorithm,and the continuous document-theme,theme-word probability distributions are obtained.Thematic evolution is demonstrated in terms of subject strength and subject similarity.