Automatic Modeling Of Knowledge Ontology And Its Fuzzy Retrieval Technology

With the rapid development of intelligent manufacturing,digitalization and Internet of Things,automatic knowledge push is becoming more and more important in the engineering application of the above technologies.At present,the main achievements of our country are still focused on various forms of knowledge active indexing and task-linked knowledge push.There is no systematic method and application system for intelligent knowledge push in the situation based on people,behavior and application objects.The main reasons are complex situation and manufacturing knowledge,difficult-extracting mapping relationship between situation and knowledge and the low accuracy rate of fuzzy retrieval.In view of the above problems,this paper carries out the following research work.In order to realize orderly organization of manufacturing knowledge,the ontology modeling method of manufacturing knowledge based on word vector clustering is studied.This paper uses Word2 Vec training the manufacturing text corpus and quickly generate keyword vectors.In order to improve the accuracy of word vectors,cross-comparison experiments are used to determine the optimal parameters of Word2 Vec model.K-Means clustering based on keyword vector is used to extract the basic concept of class center and the binary relationship between concepts.By using the concepts and relationships acquired,the ontology model of manufacturing knowledge is established and the orderly organization of manufacturing knowledge is realized.Experiment shows that the proposed method can realize the automatic modeling of manufacturing knowledge ontology.In order to support situtation-based associative retrieval,the modeling method of associative query semantic network is studied.A semantic network representation and storage model supporting ordered modeling is proposed.In order to reduce the scale of the Semantic Web and improve the accuracy of the Semantic Web,the cosine distance is used to obtain the vector similarity of concept words,and the concept with minimum correlation degree is separated and deleted based on similarity.A conceptual semantic relativity measurement model based on word frequency is established,and an associative query semantic network is built based on the measurement results,and the rationality of automated modeling is verified by experiments.This paper studies a fuzzy retrieval method based on the mapping of associative query semantic network and knowledge ontology.The Association query semantic network model is used to infer and associate the situation query conditions,and the association query phrases are obtained by fuz zy expansion in the query semantic network.The mapping rules from association query phrases to knowledge ontology are established,and the set of knowledge ontology nodes is mapped by association query phrases to realize the fuzzy retrieval based on the situation query conditions.The research content of this paper has practical significance for enriching the knowledge push method system and improving the intelligent level of knowledge push.At the same time,the ontology automation modeling method studied in this paper has certain reference significance for complex modeling requirements in other fields.