基于文本挖掘技术的高血压用药规律研究

Medication rules research on hypertension based on text mining

大数据时代的来临日益凸显数据挖掘技术的价值。文本挖掘作为数据挖掘的研究分支,对非结构化数据的知识发现有重要意义。高血压患病人群广,发病率高,治疗药物种类繁杂,寻找其中的用药规律,是临床医学的一个重要方向。基于文本挖掘技术,从在线医疗网站获取医患互动论坛数据,进行文本预处理,基于TF-IDF算法发现高血压常用中西药、非药物治疗、并发症用药特点等,结合关联规则算法挖掘"症-药"关系,有益于高血压的临床判断及用药研究。另外,验证了在线医疗网站医患互动数据用于疾病研究的可用性和效果。

The era of big data is coming which increasingly emphasizes the value of data mining technology. As a research branch of data min- ing, text mining is so important to the discovery of data-unstructured knowledge. Hypertension with the character of high incidence is one of the main diseases which damage human health. Many different kinds of drugs exist and the discovery of medication rules is one of the most impor- tant research directions. Based on text mining technology, this paper obtained data from doctor-patient interactive forum on online medical web- sites and preprocessed the text. Then, found common used Chinese and western medicine, non drug therapy, characteristics of drug use for complication and so on by using TF-IDF method, mined the relationship between symptoms and drugs with association rules algorithm. All the studies will be beneficial to the clinical judgment of hypertension and drug research. In addition, the paper verifies the availability and effec- tiveness of data from doctor-patient interactive forum used in the medical research.