计算机工程与应用 ›› 2017, Vol. 53 ›› Issue (4): 238-243.DOI: 10.3778/j.issn.1002-8331.1506-0258

• 工程与应用 • 上一篇    下一篇

智能导医系统中TF-IDF权重改进算法研究

徐奕枫1,刘利军1,黄青松1,2,傅铁威1   

  1. 1.昆明理工大学 信息工程与自动化学院,昆明 650500
    2.昆明理工大学 云南省计算机技术应用重点实验室,昆明 650500
  • 出版日期:2017-02-15 发布日期:2017-05-11

Research on TF-IDF weight improvement algorithm in intelligent guidance system

XU Yifeng1, LIU Lijun1, HUANG Qingsong1,2, FU Tiewei1   

  1. 1.Faculty of Information Engineering and Automation, Kunming University of Science and Technology, Kunming 650500, China
    2.Yunnan Key Laboratory of Computer Technology Applications, Kunming University of Science and Technology, Kunming 650500, China
  • Online:2017-02-15 Published:2017-05-11

摘要: 面向患者的智能导医系统通过人工智能技术,依据患者症状计算可能疾病,引导患者准确挂号。目前智能导医系统多采用患者输入描述自身症状或者提问的方式,该方式易出现患者输入与医学专业症状词不匹配的问题,导致计算出的疾病可信度较低。针对这一问题,提出重心后移和医学专业语料库相结合的方法,对同义词匹配,映射出与患者症状对应的症状词;根据症状不论重要与否在每一疾病中仅出现一次的特点,提出基于患者关注度的症状词频计算方法;针对传统TF-IDF算法在待分类疾病类中数量分布不均时提取疾病效果差的问题,提出基于疾病类间分布的症状权重改进算法。实验结果表明,改进算法在疾病推荐正确率和可信度两方面具有更好的效果。

关键词: 智能导医系统, 人工智能, 重心后移, 同义词匹配, TF-IDF算法

Abstract: Intelligent guidance system oriented to patients using artificial intelligence technology, according the patient’s symptoms to calculate possible disease, guide the patients to registrars correctly. At present, most of the intelligent guide system through input the patient’s symptoms or put forward questions in the website. This approach is prone to appear problems which the patient’s input and medical professional symptoms of the words do not match, resulting the disease of calculated low reliability. In order to solve this problem, this paper puts forward the barycenter backward algorithm combination with medical professions thesaurus to the literal similarity algorithm to identify synonyms, mapping the symptoms to the symptoms of the patient; According to the symptoms of the disease, whether important or not only in each of the characteristics of a disease, propose the calculation method of user attention based on the frequency of symptom; the traditional TF-IDF algorithm takes problem which symptoms uneven in classify diseases, put forward symptoms weight improving algorithm based on disease distribution. The experiment results show that the proposed and improved algorithm has better effect on performance.

Key words: artificial intelligence system, artificial intelligence, barycenter backward algorithm, synonym similarity matching, TF-IDF algorithm