Research And Implementation Of Intelligent Diagnosis System For Agricultural Enterprises' Technological Requirements

At present,the contribution rate of agricultural scientific and technological progress has increased from 53.5% in 2012 to 59.2% in 2019,but there is still a big gap compared to the contribution rate of 70%-80% in developed countries.Accelerating the transformation and upgrading of agriculture and promoting quality and efficiency are the necessary conditions for a large agricultural country to turn into a modern agricultural power.The key to agricultural modernization lies in the progress and innovation of science and technology,and the transfer of technology is the key to plugging technology wings into agriculture.At present,the research results of scientific research institutions such as colleges and universities cannot be fully applied to actual production,and there is a phenomenon that the technology owner and the technology demander are disconnected.Based on this,this paper analyzes and summarizes the influencing factors of agricultural technology demand and the key technologies that use data mining to mine technical demand,and designs and implements an intelligent diagnosis system for agricultural technology demand,with a view to making a modest contribution to the transfer of agricultural technology.The research work of this thesis mainly includes the following four aspects:(1)Explain the relevant subjects of agricultural technology transfer activities.Through a large amount of literature reading,the factors affecting the technological needs of agricultural technology consumers are sorted out,and the main factors are summarized into internal attributes and external environments.(2)Research the key technologies of using data mining for agricultural technology demand diagnosis,and elaborate the technical principles and algorithm implementation.(3)Use python programming to realize the key technology of agricultural technology demand intelligent diagnosis system,and complete the effect evaluation.There are two key technologies implemented here: the use of text clustering algorithms to cluster companies(groups)in the industry dimension;the use of conditional random field models(CRF)and models based on word embedding + BiLSTM + CRF,respectively Extracting "demand subject objects" and "demand intentions" in the technical demand description texts of agricultural enterprises.From the evaluation results,it can be seen that the word embedding + BiLSTM + CRF model has better performance.(4)Analyze the main functional modules of the intelligent diagnosis system for agricultural technology demand,design and implement the system.