Application Research Of Agricultural Information Set In Jilin Province Based On Text Classification Algorithm

In recent years,the rapid development of Internet technology has been applied in our daily life in various fields.Internet technology is also widely used in the field of agriculture,agricultural informatization began to appear in people's vision.The state attaches great importance to the development of China's agricultural informatization construction,so it has invested a lot of manpower and material resources,and many excellent researchers have devoted themselves to the field of agricultural informatization.Online agricultural information has entered the public eye in the form of exponential growth,but people do not know how to get the information they need from the vast and disorderly data set in the fastest time and the most convenient way.According to the current situation,it is very important to select the optimal text classification method to obtain relevant agricultural information.The text classification algorithm model can be used to automatically classify the text data sets of agricultural information in jilin province,which can improve the accuracy of text classification results,reduce the time of text classification experiments,and greatly improve the utilization rate of agricultural knowledge information.The main work is as follows:(1)Preparation stage of jilin province agricultural information text data set and text preprocessing stage of jilin province agricultural information set:First,a web crawler is used to crawl the documents under the relevant columns of jilin province as the agricultural information set,and then the text set of jilin province's agricultural information is processed by using Jieba word segmentation method.In addition,by using the pause word table,the modal particles,punctuation marks,Numbers and other words that cannot represent the characteristics of the text are removed from the text of agricultural information.Then,the dimension reduction operation is carried out on the text set that has completed word segmentation and word segmentation.Finally,the text is represented by text vector-quantization.(2)Realization process of classification operation of text set of agricultural information in jilin province:Text classification experiments are carried out on the vector files generated by preprocessing technology using text classification algorithm.The decision tree classification method and bayesian classification algorithm inmachine learning method are studied.Among the deep learning methods,convolutional neural network classification method,cyclic neural network classification method and long and short term memory network classification method are adopted.C-LSTM model and RCNN model in the hybrid model.The text classification experiment of jilin province agricultural information set was carried out with the established model.By comparing the experimental results of each classification model and comparing the advantages and disadvantages of each model,it is found that the mixed model has the best classification accuracy.