In the data transaction process within a data asset trading platform, quantifying the trustworthiness of data source nodes is challenging due to their numerous attributes and complex structures. To address this issue, a distributed data source trust assessment management framework, a trust quantification model, and a dynamic adjustment mechanism are proposed. The model integrates the Analytic Hierarchy Process (AHP) and Dempster-Shafer (D-S) evidence theory to determine attribute weights and calculate direct trust values, while the PageRank algorithm is employed to derive indirect trust values. The direct and indirect trust values are then combined to compute the comprehensive trust value of the data source. Furthermore, a dynamic adjustment mechanism is introduced to continuously update the comprehensive trust value based on historical assessment data. By leveraging the collaborative efforts of multiple nodes in the distributed network, the proposed framework enables a comprehensive, dynamic, and objective evaluation of data source trustworthiness. Extensive experimental analyses demonstrate that the trust quantification model effectively handles large-scale data source trust assessments, exhibiting both strong trust differentiation capability and high robustness.