Studies Of Recommendation Technology Based On Diseases And Pests Knowledge Graph Of Wheat

Wheat is one of the most important food crops in the world.China attaches great importance to the development of wheat planting industry.The sowing and tillage methods of wheat planting will have a direct impact on wheat yield,and the occurrence of wheat diseases and insect pests may lead to wheat yield reduction or even no harvest.In the process of modern agricultural development,it is necessary to strengthen the comprehensive understanding and understanding of wheat planting technology,especially the wheat farmers’ knowledge of wheat diseases and insect pests,so as to ensure the quality and yield of wheat.Due to the differences in cultural and educational levels of agricultural producers and the problems of new ways of dissemination of scientific and technological achievements,they rely more on agricultural production experience than on the relevant agricultural knowledge in the network in actual production.With the development of intelligent agriculture,the recommendation system is introduced to solve the efficiency problem of agricultural knowledge dissemination.The recommendation system can recommend according to the characteristics of different users.It can promote the effective dissemination of wheat disease and insect pest knowledge and improve wheat farmers’ understanding of related knowledge.Taking the knowledge of wheat diseases and insect pests as the research object,this paper studies the technology of naming entity recognition of wheat diseases and insect pests,and puts forward a recommended model for wheat disease and pest knowledge,which provides technical support for the effective dissemination of wheat pest control and other related knowledge.The main work of this article includes the following two parts:(1)Study on named entity recognition technology of wheat diseases and insect pests.The concept of knowledge graph is introduced into the field of wheat diseases and insect pests.Firstly,the characteristics of the knowledge of wheat diseases and insect pests are analyzed,and the construction of the ontology database in the field of wheat diseases and insect pests is completed.The knowledge graph is formed after knowledge extraction based on the ontology database of wheat diseases and insect pests.A named entity recognition model of BERTBi LSTM-CRF wheat diseases and insect pests based on BIO tagging strategy is constructed,and an experiment is designed to compare with the baseline model.The results show that the model is superior.The accuracy of model is 0.9086,the recall of model is 0.9001 and F1 score of model is 0.9043.(2)A feature fusion recommendation model based on knowledge graph is proposed.In order to solve the problems of data sparsity and large granularity of user interest description in the current recommendation model,the knowledge graph is introduced into the recommendation model,and the user interest is modeled based on the knowledge graph.After the semantic features and knowledge features of mining are fused,the gated loop unit of fusion attention mechanism is used to capture the high-level information contained in the knowledge graph.Enrich the user representation to improve the recommendation effect.Based on the common academic dataset Movie Lens and the dataset related to the knowledge of wheat diseases and insect pests,the experiments have verified the importance of each module of the recommended model built in this paper,and the model can improve the effect of recommendation and ease the data sparsity,with the AUC of 0.8857 and an accuracy of 0.8131.In view of the problems faced in the knowledge dissemination of wheat diseases and insect pests,this paper studies the technology of named entity recognition of wheat diseases and insect pests,and puts forward a recommendation model based on the feature fusion of knowledge graph.With the help of this model to promote the effective dissemination of wheat diseases and pests knowledge and help the practice of agricultural production.