Research on Data Resource Pricing Method Based on SSA-XGBoost Model  

作　　者：Jian YANG Yajuan CHEN Liwei CHANG Yali LÜ 

机构地区：[1]School of Information,Shanxi University of Finance and Economics,Taiyuan 030006,China

出　　处：《Journal of Systems Science and Information》2025年第1期116-136,共21页系统科学与信息学报(英文)

基　　金：Supported by the National Social Science Fund of China(23BJY205);the MOE(Ministry of Education in China)Project of Humanities and Social Sciences(21YJCZH197,22YJAZH080);Shanxi Provincial Research Foundation for Basic Research(202303021221184)。

摘　　要：Data pricing is a key link to promote the efficient circulation of data in the market.However,the existing methods are still insufficient in terms of pertinence,dynamism and comprehensiveness.Therefore,we proposed a data pricing prediction model based on sparrow search optimization XGBoost,aiming to provide a reference for pricing decisions in data market.First,we crawled the data transaction information of Youedata.com and performed preprocessing operations such as outlier processing,one hot encoding and logarithmic transformation on the dataset;Secondly,we conducted exploratory data analysis to understand the distribution of data and their correlation.Then,we used the LASSO algorithm to select features for the dataset and constructed a data pricing prediction model based on SSA-XGBoost.Finally,we compared and analyzed it with six machine learning models including LightGBM,GBDT,MLP,KNN,LR and XGBoost.The experimental results show that in terms of the R-squared,the prediction results of the proposed SSA-XGBoost model exceed the above six models by 4.9%,7.4%,7.1%,23.8%,12.8%,and 2.3%respectively,and are superior to the state-of-the-art work.Furthermore,the evaluation results of the five indicators of MSE,RMSE,MAE,MAPE,and RMSPE are better than other models,showing higher stability.

关 键 词：data pricing lasso SSA-XGBoost machine learning 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]