| In today’s social of information explosion, as a focus area of research in21st century, biological medicine literature is growing at an alarming rate. But in the face a huge number of documents, People no longer satisfied in the database to retrieve the data access and other simple operations, but also want the computer to automatically analyze the intelligence of the large amounts of data in the database in order to reveal the hidden knownledge in these data, the more important the message that the overall features of these data description and its development trend forecast. This study attempted to use the association rules method mining asthma-related literature to exploit their research hotpots in recent years. The purpose is to seek a more satisfactory effect of computer-processing literature to analyze specific thematic areas of research hotspot to resolve the excessive workload and subjectivity on manual reading.This study was based on asthma-related literature which published between2004and2008from PubMed. Download XML format bibliography, extracting the various literatures of the MeSH major topics and the corresponding MeSH subheadings, and statistics MeSH major topic/MeSH subheadings appear frequency, frequency intercept more than100times as a high-frequency words. High-frequency words and the each articles to form the word-articles matrix, using SPSS Clementine software to do association rule mining, generating more than800of the Rules. Construction of high-frequency words co-occurrence matrix, using SPSS statistical software to do hierarchical clustering of the matrix. Analyze asthma research hotspots in six major areas. Then, based on dendrogram bottom combined MeSH terms as a antecedent respectily in association rules to find a shared consequent with high confidence, the composition of the form (MeSH term A/MeSH term B)â†'MeSH term C, named as "co-association rules". Then analysis all areas of research hotspot based on the "co-association rules".Asthma research hotspot concentrated in six major areas:(1) the immunology pathogenesis of asthma;(2) asthma triggers and epidemiology;(3) airway remodeling;(4) asthma diagnosis and assessment;(5) asthma treatment;(6) occupational asthma. The analysis of the various areas of research hotspot through each "co-association rules" is detailed in the text. Search asthma-related literature from respiratory core journals which SCI impact factor greater than3.8and published during2004-2008. Extract and classify subjects through reading the abstracts of the literature artificial. It’s showed that more than90%of the literature’s subjects occur in hot spots summarized in this study.In this study, using association rules analysis of asthma-related literature research hotspot excavation analysis has good accuracy.It has made a new attempt for the computer processing document analysis specific thematic areas of research hotspot. Compared to the method, this study based on co-word cluster analysis can give more relations between MeSH terms that can be more specific analysis of research hotspot. |
PDF Full Text Request