Zhou Cuiping, Li Shaobo, Xie Cankun, Yuan Panliang, Liao Zihao
Guizhou University, State Key Laboratory of Public Big Data, Guiyang, Guizhou, China.
Guizhou Institute of Technology, Guiyang, Guizhou, China.
PeerJ Comput Sci. 2025 Apr 30;11:e2813. doi: 10.7717/peerj-cs.2813. eCollection 2025.
Data asset value assessment is of strategic significance to the development of data factorization, in order to solve the problems of strong assessment subjectivity and low assessment efficiency and accuracy in traditional assessment methods. This article introduces the SLPDBO-BP data asset assessment model for data asset value assessment. Firstly, the sinusoidal chaos mapping strategy, the Levy flight strategy and the fusion of adaptive weight variation operators are integrated to increase the population diversity of the algorithm, broaden the search range, and augment the global optimization capability of the algorithm. Secondly, in an attempt to comprehensively evaluate the optimization performance of SLPDBO, a series of numerical optimization experiments are carried out with 20 test functions and with popular optimization algorithms and dung beetle optimizer (DBO) algorithms with different improvement strategies. Finally, in order to verify the effectiveness of the proposed algorithm in data asset value assessment, the SLPDBO algorithm is combined with backpropagation (BP) to establish the SLPDBO-BP model for data asset value assessment, and the acquired data sets are used in the proposed model for data asset value assessment. The experimental results show that the SLPDBO-BP model performs well in assessment accuracy, and its assessment indexes mean absolute error (MAE), root mean square error (RMSE) and mean absolute percentage error (MAPE) are reduced by 35.1%, 37.6% and 38.7%, respectively, compared with the dung beetle optimizer backpropagation (DBO-BP) model, and its evaluation efficiency is improved, and the proposed model demonstrates better evaluation simulation effects by remarkably outperforming other models in terms of evaluation accuracy and error level.
数据资产价值评估对数据分解的发展具有战略意义，旨在解决传统评估方法中评估主观性强、评估效率和准确性低的问题。本文介绍了用于数据资产价值评估的SLPDBO-BP数据资产评估模型。首先，集成正弦混沌映射策略、莱维飞行策略和自适应权重变异算子融合，以增加算法的种群多样性，拓宽搜索范围，增强算法的全局优化能力。其次，为全面评估SLPDBO的优化性能，使用20个测试函数以及具有不同改进策略的流行优化算法和蜣螂优化器（DBO）算法进行了一系列数值优化实验。最后，为验证所提算法在数据资产价值评估中的有效性，将SLPDBO算法与反向传播（BP）相结合，建立用于数据资产价值评估的SLPDBO-BP模型，并将获取的数据集用于所提模型进行数据资产价值评估。实验结果表明，SLPDBO-BP模型在评估准确性方面表现良好，其评估指标平均绝对误差（MAE）、均方根误差（RMSE）和平均绝对百分比误差（MAPE）与蜣螂优化器反向传播（DBO-BP）模型相比分别降低了35.1%、37.6%和38.7%，其评估效率得到提高，并且所提模型在评估准确性和误差水平方面显著优于其他模型，展现出更好的评估模拟效果。
