The investigation of Ecological Agriculture (EA) patterns can reveal the differences, aggregation, and diversity of agricultural development, providing specific paths in agricultural development and environmental protection to achieve the Sustainable Development Goals. Although field surveys, literature analysis, and the method using administrative statistics can be employed to investigate EA records and determine EA distributions comprehensively, they still rely on manual operations that are generally unable to support the rapid and large-scale identification of EA patterns required by current agricultural sustainable researches. To address this issue, this paper proposes a novel and rapid approach for Ecological Agriculture Pattern Investigation Based on Web-text (WEAPI), with the ability to automatically acquire EA pattern records, including pattern type, occurrence time, precise location, and other relevant information. The proposed method is employed in a national-scale case study to investigate trends in Chinese Ecological Agriculture (CEA). Results of the study reveal WEAPI's ability to detect new trends in CEA via the latest news and the corresponding distributions. The WEAPI method can also exhibit the unknown patterns of the current Chinese agricultural development. Further validation experiments demonstrate that the proposed method achieves over 95 % precision in the pattern parse processes and an 87 % coverage rate at the town level of the official CEA pattern list. Moreover, WEAPI can provide dynamic changing analyses on the annual evolution of the EA patterns in each type. Despite limitations under sparse records in partial classes, the results reveal WEAPI as a promising and powerful tool for agricultural research and agricultural development planning.
Keywords: CEA distribution; Chinese Ecological Agriculture (CEA) patterns; Data mining; Ecological Agriculture (EA); WEAPI; Web text.
Copyright © 2022. Published by Elsevier B.V.