Big Data Analysis Of Agricultural Product Market Transactions Based On Hadoop

Big data technology has changed the lives of human beings and brought great convenience to dealing with problems in life.China's agricultural product market transactions have been in an unstable state,especially for the price of agricultural products.In practice,as the amount of data increases over time,the transaction data of agricultural products market is diverse.Traditional databases have defects in large data storage,resulting in partial data loss and inaccurate data statistics.Big data technology makes up for the shortcomings of traditional database storage,can handle massive amounts of data,is no longer limited to the operation of a single machine,expands the amount of data that can process data through clustering,and big data technology also solves the difference in data types.Formal problems.Big data technology can also collect and process data in real time.The processing is parallel processing.The advantage of parallel processing is that it can improve the speed of data processing.Massive data processing can greatly reduce data operation time.Real-time data processing can be timely.Provide accurate data analysis and decision making.Big data technology provides accurate predictive analytics and is the key to gaining competition in trade in agricultural market transactions.Taking advantage of big data technology,this paper uses big data technology to process data.The big data technology in this paper is mainly based on Hadoop platform for massive data processing analysis of agricultural product market transactions.The core of Hadoop platform is MapReduce and DHFS.In the data storage process of this paper,HDFS distributes the transaction price data of agricultural products market.In order to simplify the system operation process,HBase module is added to the Hadoop platform,which can perform random and real-time read and write operations on data.In the prediction result,HBase is used to complete the data result storage operation;in data processing,MapReduce processes the data,and completes the data processing through map()and reduce()functions;meanwhile,the Hive module is added to the Hadoop platform.Hive uses SQL queries to simplify the process of data extraction during data extraction.In order to be able to analyze the data more intuitively,the R language is added to the Hadoop platform,which constitutes the complete Hadoop+HBase+Hive+R language development platform for Hadoop in the agricultural market transaction data analysis.The R language makes the data modeling process.More simple,using the visual characteristics of the R language,this article visually analyzes the data througha line graph.At the same time,the ARIMA model and the exponential model are used to analyze the model on the platform.The ARIMA model and the exponential model are modeled and model predicted.The accuracy of the model is calculated according to the sampling data and forecast data.And the line graph analysis of the data verified that the ARIMA model is a better predictive model for the pepper transaction price.Through the prediction of the model,it can be used in the big data analysis of agricultural products trading.According to the prediction characteristics of different models,find suitable models to make medium-and long-term or short-term forecasts for agricultural market transactions,make reasonable analysis through medium and long-term or short-term forecasts,make reasonable decisions on agricultural products trade and government regulation of agricultural products cultivation.