Research Of Data Mining In Nutrition Catering System

Currently, chronic disease has become a public health problem in China, and theexpert said that the nature of the chronic diseases is the imbalance of trace elements inthe human body. Trace element in the human bodyis almost derived from water andfood, and reasonable diet can ensure body adequate intake of trace elements. So builda personal nutrition and health catering system from the perspective of trace elementscan effectively help people eat healthy and adequate intake of trace elements to keepthem away from chronic diseases and help chronic patients recover.Many fields, including medical field, have been generating and retaining a largeamount of data by the development of computer technology, hospitals retained a largenumber of patient datum, including the data of trace elements, by EHR, informationsystems and digital devices. By analyzing these data, we are often able to reveal manypotential laws. Data mining is the technology which can find potential and valuableinformation by analyzing large amounts of data. Association rule mining is one of themost active data mining research methods which is used to find hidden, unknown,valuable information by exploring the association between the massive amounts of data.However, most hospitals do not have capacity of analysis, summary and applications ofdata.This paper hopes to study content of trace elements in chronic diseases tosummarize the relationship between chronic disease and trace elements, and apply therelationship to the personal nutrition, catering system to help people eat healthy. Firstly,this paper gives the definition of nutrition and health catering, shows problems of it andtell us it is so important to use mining technology in nutrition science and Diseasesscience to find laws and apply it in the field of nutrition and health catering. Secondly,this paper proposes an improved Apriori algorithm based on learning and in-depth studyof the classical association rule mining algorithm. The thinking of Apriori algorithm issimple, it find association rules by support value and confidence value in frequentitemset which generated by an iterative method which is called drill-search. Above that,it is easy to find that there is a big efficiency bottleneck caused by lots times of scanningdatabase and a large number of candidate itemset generated. The nature of improvedApriori algorithm in this paper is reducing the number of items which need scan in thedata mining, in order to improve the efficiency of the algorithm while ensuring the correct rate of data mining.This paper wants to find the relationship between trace elements and chronicdiseases by the improved algorithm. This paper chooses diabetes as example of chronicdiseases to research, it describes data pre-processing and data mining process in detail,analyze and interpret mining results.Finally, this paper apply mining association rules to the personal health nutritioncatering system, and gives the application results.