
Structured evidence extraction record underlying the systematic survey "Arabic Information Extraction: A Systematic Survey of Techniques, Resources, and Challenges (2018–2025)".
The file contains one row for each of the 42 papers in the final review corpus, with 19 extracted fields per paper corresponding to the standardized extraction template described in Section 2 of the manuscript. Fields cover bibliographic metadata, task and domain labelling, dataset and tool information with availability status, modelling approach, evaluation setup and metrics, three-axis contribution assignments, and survey placement.
Scope: this record covers included papers only. Records excluded during title/abstract screening (n = 26) and full-text eligibility assessment (n = 18) are not listed; the screening pipeline and its counts are reported in Section 2 and Figure 1 of the manuscript.
A README sheet within the file documents every column, the conventions used, and the provenance of the extraction.