Research On Ontology Modeling Integrating Trust And Uncertain Context Quality Correction Model

This paper focuses on how to better model the context.There are some problems in the modeling and utilization of context.This paper focuses on improving the quality of context,fuzzy ontology modeling and missing context completion in three aspects: low quality context,fuzzy context and missing context.Then they are applied in the fields of photovoltaic power generation prediction,literature summary and literature retrieval.The details are as follows:First,for low-quality contexts,this paper proposes a new approach to construct a high-quality ontology model that supports the representation of context quality and can improve context quality to some extent.The high quality ontology classifies the context according to its source,which can be divided into four types: user context,network context,sensor context and neighbor context.Then,the quality representation of these classified contexts is carried out.Finally,the high-quality context is selected to replace the low-quality context to improve the quality of the context.The above research results can be applied in the field of photovoltaic power generation prediction to solve the problem of low quality environmental context data.In the simulation experiment,high quality ontology of photovoltaic power generation is established in this paper to obtain the context information needed for photovoltaic power prediction.Combined with the prediction model of gated cyclic neural network,this method can improve the prediction accuracy of photovoltaic power,so that the grid can operate more safely when photovoltaic is connected to the grid.Secondly,for fuzzy context,this paper studies a fuzzy context ontology to model and collect the context of user articles,and applies it to the field of document summaries to improve the accuracy of document summaries.In the field of literature abstract,different words will have different meanings depending on the field they are in.The method proposed in this paper can find the words that can accurately describe the content of an article by fuzzy domain membership,user interest membership and topic membership in the context ontology.Finally,through these key abstract words to get an accurate summary.Through the experiment of literature abstract,it is proved that the average accuracy of the fuzzy ontology can be improved by about 15% compared with the traditional method,so that readers can understand the content of the article accurately through the abstract.In addition to low-quality context and fuzzy context,there is another kind of completely missing context.If the citation frequency of newly published literature is empty,it belongs to the completely missing context.In this paper,the citation frequency of newly published articles is inferred and predicted,and then the citation frequency is applied in the literature retrieval,so as to improve the accuracy of literature retrieval.Specifically,by combining the literature context ontology with the fuzzy neural network reasoning model,the missing context is obtained by using other existing context reasoning and these literature retrieval contexts are applied to the literature retrieval field.The literature retrieval experiment shows that the average accuracy of this method is about 20% higher than that of the BM25 method alone,and the literature retrieved by this method can be more in line with the user's interest.