中国总膳食研究食物聚类自动化探索  

Exploration of the food cluster automation in the China Total Diet Study

作　　者：李万庆[1] 沈雯捷[1] 潘俊霞[1] 闵捷[1] 李筱薇[2] 

机构地区：[1]东南大学公共卫生学院流行病与卫生统计学系,南京210009 [2]国家食品安全风险评估中心

出　　处：《卫生研究》2012年第6期895-899,共5页Journal of Hygiene Research

摘　　要：目的实现中国总膳食研究食物聚类自动化,提高食物聚类计算的质量和效率。方法通过对中国食物成分表中食物编码特点研究,以及总膳食研究食物聚类的原则进行分析,构建能够让计算机语言识别这些特点和原则的算法,实现食物聚类自动化。结果以第五次中国总膳食研究某省膳食调查数据为例,将292种食物聚为53种聚类食物,聚类结果符合总膳食研究聚类要求。结论该方法能够有效地实现中国总膳食研究食物聚类自动化计算,提高了中国总膳食研究中食物聚类的质量和效率。Objective To achieve the food cluster automation of China Total Diet Study and improve the quality and efficiency of the food cluster calculation. Method The food coding features were studied in the Chinese Food Composition Table. After analyzing the principles of food clustering on Total Diet Study, constructing an algorithm allows the computer language to identify these characteristics and principles to achieve automatic food cluster. Results The instance was selected from the data of a province in the fifth Chinese Total Diet Study. 292 food items were clustered into 53 kinds of cluster food by computer program, and the results are corresponded to the cluster requirements of total diet study. Conclusion The automatic calculation of the food cluster can be achieved by computer program more reliable and effectively in the Total Diet Study.

关 键 词：总膳食研究 食物聚类 自动化计算 

分 类 号：R151.42[医药卫生—营养与食品卫生学]