一种基于贝叶斯网络的匹配引擎设计

匹配引擎不是简单的搜索，而是全新的深层次信息挖掘。该文构建一种基于贝叶斯网络模型的匹配引擎。项目需求中有4种类型的节点集合，通过建模，设计一个4层贝叶斯网络，主要用来对输入样本进行打分筛选。在该模型中，层与层之间都有表示具体意义的条件概率将其联系起来，在一定程度上解决了检索结果歧义和检索结果少的问题。实验证实了该模型可以更准确地搜索信息。

Matching Engine Design Based on Bayesian Network

Matching engine is not simpls simple search,but absolutely in-depth information mining.This paper explains a construction of a matching engine based on simple Bayesian network.By modeling,it designs the four kinds of node set into a four-tier Bayesian network,mainly to mark and select the input samples.In this model,all tiers have meaningful conditional probability between them,which,to some extent,has solved wrong results and few results problems.Experiments prove the model can search more precisely.