Loading...
Loading...
Contributors:
Dataset URI:
10.6084/m9.figshare.34764636.v1The following description is suitable for the Figshare “Description” field. It explains the scope, structure, scientific value and reproducibility of the dataset.This dataset accompanies the systematic review entitled “Artificial intelligence for crop pest and disease management: Diagnostic reliability, field transferability and decision value.” It provides structured bibliographic records, screening decisions, evidence classifications, methodological assessments and reproducible Python code supporting a critical evaluation of artificial intelligence (AI) applications in crop protection.The review examines three interconnected questions: how reliably AI systems identify crop pests and diseases, whether their performance is maintained under independent field conditions, and whether AI-generated information contributes to better pest and disease management decisions and agricultural outcomes.The literature search covered Scopus, PubMed and IEEE Xplore, with 52 final search exports representing 36,159 bibliographic occurrences. Deduplication yielded 10,977 unique publication records. Broad screening classified 5,814 records as retained, 2,138 as excluded and 3,025 as uncertain. A focused evidence set comprised 132 publications, of which 101 were included in the final synthesis. The remaining records comprised 29 publications for which full texts were unavailable, one excluded as outside the review scope and one excluded following retraction.The 101 included publications were organised into four evidence domains: diagnostic reliability (20 studies), independent field validation (20), adaptation and transferability (28), and AI-supported pest and disease management (33).The repository contains six principal supplementary components:S1: Search strategies, database records and search-file manifest.S2: Bibliographic occurrences, deduplication, screening decisions and selection history.S3: Evidence extraction framework and methodological codebook.S4: Focused evidence matrix documenting crop species, pest and pathogen targets, diagnostic approaches, validation characteristics, transferability and evidence maturity.S5: Comparison and claim-audit records addressing reference standards, independent validation, methodological limitations and links between AI predictions and management outcomes.S6: Customisable Python code for bibliographic processing, conservative duplicate identification, data validation and reconciliation of screening and evidence records.Additional documentation includes a data dictionary, software requirements, execution instructions, configuration template, validation outputs and file-integrity checksums.The supplied Python code enables researchers to reproduce the documented screening and evidence-reconciliation checks and adapt the analytical procedures to other systematic reviews. The original database export files, licensed full-text articles and complete source-level extraction documents are not included. Consequently, independent reproduction of the original search intake and verification of every full-text claim requires access to those materials. Screening classifications and evidence assessments should be interpreted alongside the documented uncertainties and methodological qualifications.The dataset is intended to support transparency, independent scrutiny and methodological reuse. It offers a structured basis for examining the strengths and limitations of AI research in crop protection, identifying gaps between experimental performance and field validation, and evaluating the evidence needed to translate AI-based diagnostic technologies into reliable and practical tools for farmers.
Keyword(s):
Artificial intelligence not elsewhere classified; Artificial intelligence; Crop pest and disease management; Diagnostic reliability; Field transferability; Decision support systems; Sustainable agricultural development.
MeSH Term(s):
Grants and Funding:
No Grants or Funding Data Available
Associated Publication(s):
No Associated Publication Data Available
Supplementary Data:
No Supplementary Data Available
Related Dataset(s):
No Related Dataset Data Available
The National Library of Medicine (NLM) provides information about biomedical datasets. Inclusion on an NLM Web page does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: NLM Web Policies (nih.gov)