基于WEB的水产养殖病害诊断专家系统  

An expert system of disease treatment in aquaculture based on WEB

作　　者：阎笑彤 徐翔[1] 郭显久[1,2] 孙凯乐[1] 

机构地区：[1]大连海洋大学信息工程学院,辽宁大连116023 [2]辽宁省海洋信息技术重点实验室,辽宁大连116023

出　　处：《大连海洋大学学报》2016年第2期225-230,共6页Journal of Dalian Ocean University

基　　金：国家海洋公益性行业科研专项(201205023-4);辽宁省科技计划项目(2012216012)

摘　　要：为解决水产养殖中的病害问题,采用TFIDF加权算法和动态规划的设计理念,结合余弦相似度模型,设计并实现了基于WEB的水产养殖病害诊断专家系统。该系统包括专家诊断模块、查询模块、浏览模块和专家在线更新模块,能自动对用户输入症状进行关键词提取并通过TFIDF算法加权,然后采用余弦相似度模型公式计算,得出输入文本与知识库中每篇文档的相似度,最后把相似度高的前3个疾病及其治疗方案作为诊断结果反馈给用户。经实验验证,该系统可以快速地对用户输入的病害症状进行正确诊断,可用于水产养殖生产实践中。An expert system of disease treatment in aquaculture is designed and implemented based on WEB by TFIDF weighted algorithm, design idea of dynamic programming as well as cosine similarity model. The system in-cluding expert diagnosis module, query module, browse and experts online update module can automatically extract the keywords of user input symptoms by weighting TFIDF algorithm, and then conclude the input text with each of the document similarity in the knowledge base using cosine similarity model formula, and finally give the high simi-larity of the first three diseases and their treatment as a diagnostic feedback to the user. The experimental results showed that the system quickly made correct diagnosis for the disease symptom of user input.

关 键 词：水产养殖 病害诊断 专家系统 知识库 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]