Multi-dimensional Visual Exploration Of The IUCN Red List Species Information And Related Literature

Research on data visualization has been used in many aspects of life today,but for multi-dimensional data such as species information and its literature information,the existing research methods use a single visualization chart or software for static chart presentation,and researchers cannot obtain correlation information between data,limiting the deep analysis of problems in the professional field.How to enable researchers to better use species data for biological protection,and effectively mining relevant information from species data and literature data has become an urgent problem to be solved.Therefore,the study will be carried out from the following aspects.First of all,in view of the problem that only considering the volume of literature published in a country or a certain time period cannot obtain the overall research trend,combined with bibliometric method and statistical analysis,the data processing and design visualization interactive exploration of the information with dimensions such as time and space in the literature data set are carried out,and the multi-dimensional dynamic correlation analysis of the literature data is realizedSecondly,in order to solve the problem that viewing a single keyword cannot meet the researchers’ comprehensive understanding of the current research status,the BERT pretraining model + hierarchical clustering method is used to complete the clustering of keywords and through visual interaction design,so as to achieve the goal of exploring the internal relationship between keywords and analyzing the laws contained therein.Third,the method of fused BERT and CNN models is adopted to label the abstract part of the literature with "pseudo-label",and then,the new data set is trained to ensure that the abstract text is classified according to the name of the species,so as to solve the problem of insufficient utilization of literature information through simple regular matching,and reduce the difficulty of researchers finding corresponding literature.Finally,in view of the above experimental results,a visual analysis system of biological information based on multi-level information browsing is constructed,and the hierarchical visual view of the rising sun map,radar map,Sankey map and relationship diagram is used to explore the relevant information of the IUCN Red List and its research literature by dynamically associating the interaction method,improve the ability to obtain data,and solve the problem that Cite Space software cannot interactively explore and analyze the results.