Research On The Influencing Factors Of Online Second-hand Book Prices In China

With the concept of circular economy becoming more and more popular,the market development of second-hand books ushered in a new opportunity.With the advent of the Internet era,massive resources have flooded into the market and various types of online second-hand book trading platforms have sprung up,greatly promoting the rapid rise of the second-hand book market.However,there are numerous secondhand book trading platforms on the market,and the phenomenon of non-standard pricing of second-hand books seriously hinders the development of the second-hand book market.Therefore,it is of great significance to explore the factors affecting the price of second-hand books and establish reasonable pricing mechanism of secondhand books.This paper conducts research based on the book data of three online platforms,Douban Books,Duozhuayu and Kongfuzi,and provides a theoretical basis for the formulation of second-hand book prices through quantitative analysis.The top four textbooks,literature,children’s books and art books in the second-hand book market were selected,and the characteristics of the overall books were descriptively analyzed,and then the textbook books with the largest proportion in the second-hand book market were mainly studied to construct a second-hand book characteristic price model,screen out the significant features,and quantitatively analyze the significant features.The book text data of book title,book introduction and book catalog were segmented,and the content characteristics of books were quantified according to the number of keywords of hit subjects,and the content characteristics and second-hand book prices were analyzed.Finally,based on the two ensemble learning algorithms of XGBoost and LightGBM,the data sets of significant features without adding content features and adding content features are used to predict the price of second-hand books,and a second-hand book price prediction model based on the integrated algorithm is established,and the prediction effect is compared and analyzed.The results show that in the characteristic price model,the features of Duozhuayu,Duozhuayu inventory,open book,whether the book is translated,book pricing,Douban score,Duozhuayu transactions,age of the audience,Kongfuzi quantity and the number of people who want in Douban have a significant impact on the price of used books.The coefficient of complex correlation between the content of books and the price of second-hand books is 0.28,indicating a strong correlation.From the model prediction results,the prediction effect of the model with content features added is better than that of the model without content features added,and the performance effect of XGBoost model is better than LightGBM.The average absolute error(MAE)of XGBoost model on the data set with content features added is 1.266.Goodness of fit(R2)was 0.920.Finally,this paper suggests that consumers pay attention to the appearance of books when buying second-hand books,and can understand the current situation of the second-hand book market by paying attention to different second-hand book platforms,and choose the most cost-effective second-hand books;It is suggested that the secondhand book platform should make its pricing mechanism transparent and establish a more reasonable second-hand book pricing system.Readers and second-hand book platform can jointly promote the better development of the second-hand book market.