TITLE:
Digital-Intelligence Empowerment for Rural Revitalization: Estimating First-Launch Economy Productivity Efficiency via ETL-MLP Time-Series Reconstruction
AUTHORS:
Mo Shi, Hetao Yuan, Xuao Chen
KEYWORDS:
Rural Revitalization, First-Launch Economy, Digital-Intelligence Empowerment, ETL Pipeline, MLP Regressor
JOURNAL NAME:
Open Journal of Social Sciences,
Vol.14 No.3,
March
19,
2026
ABSTRACT: Evaluating the economic productivity efficiency of the First-Launch Economy is crucial for rural revitalization. Yet, empirical efforts are frequently hindered by the scarcity, fragmentation, and temporal lag of regional production statistics. To address this critical data gap, this study develops a reproducible digital-intelligence framework leveraging Extract-Transform-Load (ETL) pipelines and Machine Learning reconstruction. As the First-Launch Economy of the plush toy industry in Hanbin District, Ankang City, this study fuses national macro-export anchor data with web-scraped local enterprise records. The methodology in this study employs an ETL pipeline for rule-based time-series cleaning, utilizing tailored heuristics like last-observation-carried-forward (LOCF) to standardize and impute sparse data. Subsequently, through exhaustive parameter traversal, a Multi-Layer Perceptron (MLP) Regressor is trained on 96 monthly observations (2018-2026). Subsequently, the MLP regressor effectively downscales sparse monthly inputs to generate 2921 synthetic daily production records for the time span from 2018 to 2026, while the 2921 daily records strictly preserve macroeconomic temporal structures. The reconstructed high-resolution dataset overcomes traditional data granularity limitations, enabling complex downstream econometric tasks such as short-term forecasting, shock-response simulation, and counterfactual policy assessment. Due to this, the framework in this study provides a validated, transferable blueprint for the digital empowerment of rural economies, while outlining future pathways to integrate remote-sensing and firm-level survey data to overcome source heterogeneity further.