| With the continuous expansion of the trading size of e-commerce in China,the online consumer groups have been growing year by year,and the e-commerce industry has become more and more mature.The rapid development of e-commerce will surely promote people's production and lifestyle changes,and accelerate the transformation of people's consumption concepts.At the same time,the e-commerce development model also conforms to the concept of innovation and development in our country.The rapid development of the e-commerce domain will surely play an important role in China's economic restructuring.Therefore,based on the online transaction data,this paper enriches the traditional statistical sources through the online transaction data,so as to promote the statistical research based on the online transaction price,and strive to enrich the research work based on the online transaction data with the background of online transaction.This paper mainly completed these work:First,it analyzes the current situation of e-commerce development in our country,sorts out the trends and prospects of the application of online data under the trend of Big Data,and summarizes the technical characteristics of the existing online data acquisition methods.On this basis,this article considers starting from the online data extraction,combining the platform advantage of big data distributed system and the technical characteristics of Nutch crawler to build the data capture framework of Nutch crawler in distributed cluster.In practice,it is better to capture the transaction price data of Taobao E-commerce,which lays the foundation for the follow-up work of this article.Second,analyze online transaction data.In this paper,based on the market share of mobile phone in the current year and the types of mobile phone brands captured each day,10 mobile phone brands were selected as the research objects,and the corresponding mobile phone price processing methods were worked out to get the transaction data of mobile phones.Then the paper analyzes the volatility of online transaction price,discusses the difference and connection between online transaction and traditional entity transaction,at the same time analyzes the anomalous fluctuation in price by K-means clustering algorithm,and summarizes the existence of anomalous price fluctuation,and lays the foundation for the follow-up anomaly price fluctuation.Third,the statistical analysis of price data.In the third chapter,we focus on the characteristics of abnormal fluctuation price and summarize the classification of abnormal fluctuation.And according to the characteristics of data fluctuation in single week,it isfound that the abnormal fluctuation price appears in one to two times within one week.Due to the short-term changes in cell phone prices,there is a great correlation.Therefore,in this paper,the idea of local fitting is considered and the local polynomial fitting method is used to correct the abnormal fluctuations according to the data points in the local area.Fourthly,after completing the above problems,this article refers to the preparation of China's consumer price index(CPI),formulating an index calculation model based on online transaction price.The main calculation is the chain price index of daily,and the monthly average fixed base index and chain index of price.At the same time,we compare the index before and after the abnormal fluctuation correction,and analyze the result of the correction.According to the calculation result,it is found that the online transaction price index based on 10 kinds of mobile phone brand data is close to the price index published by the government,and the research results accord with the original intention and reach the expected result. |
PDF Full Text Request