Abstract Objectives Big data consisting of unstructured data such as documents, images, sound sources, etc. are difficult to apply to existing structured data analysis programs and analyzed using techniques such as text mining, web mining, opinion mining, and network analysis. In metabolic syndrome management, interventions such as lifestyle correction are more important than drug treatment. Therefore, this study studied nutrition that can help prevent metabolic syndrome through atypical data analysis using Medline data. Methods From 1977 to December 2019, a total of 992 abstracts were extracted among the papers available from the Pubmed using the Mesh words of metabolic syndrome, nutrition, and prevention as a search term. Text mining using the Netminer 4.0 program resulted in the extraction of 7846 nouns and the final nouns of nutrients or foods defined as the frequency of occurrence are more than 30 times. For the selected words, we constructed and analyzed a network using links that connectivity values of 0.05 or more. Results Of the 27 words related to nutrition in 992 papers, most five frequent nouns were Calcium, Magnesium, Mediterranean diet, Zinc, and Dairy product. In the network analysis, the five keywords in the centrality analysis were Dairy products, Fish, Vegetables, Fruit and Copper. The 27 words were grouped into eight groups, and four groups of one or more words were identified: A first group consisting of Calcium, Copper, Flavonoid, Iron, Magnesium and Selenium, and the second group of Zinc, DHEA, EPA, Fish and Omega 3. The third group consisting of Polyphenol, Prebiotics, Probiotics, and Yogurt, and the last group consisting of Dairy products, Fruits, Mediterranean diet, Milk, Nut, Sodium, Sugar, and Vegetables. Conclusions Numerous minerals, omega 3, Probiotics and vegetables, fruits and dairy products were identified in the nutrition papers related to the prevention of metabolic syndrome. Funding Sources This study has not supported by any funds.
Read full abstract