Analysis Of Chrysanthemum Market Information And Price Prediction Based On Web Text

In recent years,with the vigorous implementation of the country's rural revitalization policy,many remote towns and villages have,on the basis of their own conditions and relying on policy support,rationally developed resources and embarked on the road to prosperity.As one of the four major cut flowers,chrysanthemums have a short growth cycle,high market prices,and significant economic benefits.They are the first choice for poverty alleviation in many southwestern townships.But currently,the chrysanthemum market has problems such as lagging information,incomplete production and sales data,and lack of market dynamics researchers,which has caused chrysanthemum farmers to have insufficient understanding of the market,and chrysanthemum cultivation has certain risks.Based on text mining technology,this paper collects network information text and researches the chrysanthemum market information.Use the LDA model to extract themes of chrysanthemum-related news from government websites,and explore the evolution of the topics of related news.Second,use the word frequency statistical method to explore the attention of chrysanthemums in cities in the network information text,and visualize the heat map;Furthermore,using text analysis method,based on python software to extract the influence factors of price changes on network information text,and draw a relationship map of chrysanthemum price influence factors,including market factors,environmental factors,social factors,policy factors,and technical factors.Finally,the correlation analysis was used to study the quantifiable price influencing factors,which provided support for selecting a suitable forecasting model to predict the price of chrysanthemums.Research shows that since 2016,the theme of chrysanthemum-related news on government websites has been "Getting Rid of Poverty and Getting Rich”,indicating that there is still a lot of room for development of the chrysanthemum industry in the future,and research on it is of great significance.Analysis of urban heat shows that The three places are the most popular cities in China,and the results are in line with the current situation.Among the quantifiable influencing factors,the correlation between temperature and precipitation is weak,and the seasonal factor is significantly correlated with the price.Finally,the season in the market factor is determined.Sexual fluctuations are the most important component of chrysanthemum price changes.Based on the monthly price data of Huangju from Guangzhou Flower Research Center from January 2009 to December 2018,a time series seasonal model SARIMA(1,1,1)X(1,1,0)12 was established using time series analysis methods.Predict prices from June 2018 to June 2019 and compare them with actual price data.The results show that in the 13-month forecast data,the accuracy of 8-month forecast results is more than 90%,and the overall forecast accuracy of the seasonal model is 86.2%.It shows that the time series seasonal model can be applied to the price prediction of various chrysanthemums with seasonal fluctuations,which can provide effective information of market risk for chrysanthemum employees and provide information support for production decisions.