改进贝叶斯的语义推送算法设计

教育信息语义本体构建是通过语义本体构建方式去设计教育信息本体库。本体间逻辑关系表示方法,是构建出有逻辑结构的教育信息集合的过程。实现教育信息的半结构化数据归类,对不同时间采集的归类数据在规定好的模型中进行计算—词汇频度分析模型。词汇频度分析模型运用逆概率的贝叶斯思想,经过对传统贝叶斯算法与语义本体性质相结合,使MapReduce善于处理半结构化数据;经过对语义本体构建的教育信息数据结合词汇频度分析模型进行计算,获得教育信息本体的推荐能力值E i;通过对不同本体E i值进行排序,获得了推荐信息的顺序;根据推荐权重进行信息的推送工作,同时根据JS指数,经过比较基于词汇频度分析模型与目录结构推送算法的分析结果得出:词汇频度分析模型优于基于目录结构推送算法。

Design of Semantic Push Algorithm Based on Bayesian

The construction of educational information semantic ontology is to design educational information ontology database through semantic ontology construction.The representation method of logical relation between ontologies is the process of constructing the set of educational information with logical structure.The semi-structured data classification of educational information is realized,and the classification data collected at different time are calculated in a well-defined model—lexical frequency analysis model.The Bayesian idea of inverse probability is introduced in the lexical frequency analysis model.The combining of the traditional Bayesian algorithm with the semantic ontology property makes MapReduce deal with semi-structured data well.After calculating the educational information data based on semantic ontology and vocabulary frequency analysis model,the recommendation ability value(E i)of educational information ontology is obtained.By sorting different ontology E i values,the order of recommendation information is obtained.The information is pushed by the recommendation weight.According to JS index,by comparing the analysis results based on lexical frequency analysis model and directory structure push algorithm,it is concluded that the lexical frequency analysis model is superior to the push algorithm based on directory structure.