Hogeveen H, van Vliet J H, Noordhuizen-Stassen E N, De Koning C, Tepp D M, Brand A
Department of Herd Health and Reproduction, Utrecht University, The Netherlands.
J Dairy Sci. 1995 Jul;78(7):1441-55. doi: 10.3168/jds.S0022-0302(95)76766-2.
A knowledge-based system for the diagnosis of mastitis problems at the herd level must search for possible causes, including malfunctioning milking machines or incorrect milking technique. A knowledge-based system on general mechanisms of mastitis infection, using hierarchical conditional causal models, was extended. Model building entailed extensive cooperation between the knowledge engineer and a domain expert. The extended knowledge-based system contains 12 submodels underlying the overview models. Nine submodels were concerned with mastitis problems arising from machine milking. These models are briefly described. The knowledge-based system has been validated by other experts after which the models were adjusted slightly. The final knowledge-based system was validated to data collected at 17 commercial dairy farms with high SCC in the bulk milk. Reports containing the farm data were accompanied by recommendations made by a dairy farm advisor. This validation showed good agreement between the knowledge-based system and the dairy farm advisors. The described knowledge-based system is a good tool for dairy farm advisors to solve herd mastitis problems caused by a malfunctioning milking machine or incorrect milking technique.
用于在畜群层面诊断乳腺炎问题的基于知识的系统必须寻找可能的病因，包括挤奶机故障或不正确的挤奶技术。一个使用分层条件因果模型的关于乳腺炎感染一般机制的基于知识的系统得到了扩展。模型构建需要知识工程师和领域专家之间的广泛合作。扩展后的基于知识的系统包含12个基础概述模型的子模型。其中9个子模型涉及机器挤奶引发的乳腺炎问题。这些模型将被简要描述。该基于知识的系统已经过其他专家的验证，之后对模型进行了轻微调整。最终的基于知识的系统针对从17个商业奶牛场收集的原奶体细胞数（SCC）高的数据进行了验证。包含农场数据的报告还附有奶牛场顾问给出的建议。此次验证表明基于知识的系统与奶牛场顾问的意见高度一致。所描述的基于知识的系统是奶牛场顾问解决由挤奶机故障或不正确挤奶技术导致的畜群乳腺炎问题的一个良好工具。
