A Blockchain-Based Method for Trusted Storage and Intelligent Analysis of Ideological and Political Education Training Data
Year: 2026
Type: conference-paper
Abstract: Addressing the core needs of trusted storage and intelligent analysis of ideological and political education training data, this paper proposes an integrated blockchain-based solution. A three-layer consortium blockchain trusted storage model is constructed to adapt to the multi-source heterogeneity, temporal sequence, and large-scale characteristics of training data. A lightweight layered consensus mechanism (LHC) is designed to optimize the high communication overhead and latency of traditional PBFT consensus. Through node layering and a three-stage simplified process, the consensus communication complexity is reduced from O(n²) to O(n). Simultaneously, an attention-driven temporal mining algorithm for ideological and political elements (AIPEM) is designed, constructing an end-to-end analysis framework of “trusted data input - multi-source feature fusion - attention weighting - temporal classification - ideological and political element mapping.” Relying on multi-source feature fusion, self-attention weighting, and lightweight Transformer temporal classification, accurate quantitative evaluation of ideological and political elements is achieved. Experimental results show that the LHC mechanism, while maintaining a 100% data tampering detection rate, achieves a 45.1% increase in peak throughput and a 44.8% reduction in average evidence storage latency compared to the traditional PBFT. The AIPEM algorithm achieves a 92.3% accuracy rate in classifying ideological and political elements, representing improvements of 8.1 and 5.6 percentage points compared to RNN and CNN models, respectively, while reducing overall computational complexity by 42%. This research provides technical support for the intersection of educational digitization and industrial data management, effectively solving the core challenges of reliable management of practical training data and accurate analysis of ideological and political elements, and possesses significant engineering application and promotion value.
Author: Xin Liu
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
SDG: Quality education
Open Access status: hybrid
OpenAlex ID: W7204181523