Review audit and coding data for computational metaphor identification
Description
These materials support the review Beyond F1: An Evidence-Centered Critical Review of Computational Metaphor Identification. They contain an eligibility audit, study-level evidence-anchor and evaluation-support coding, and the grouped counts and code underlying Figure 4. The all-period audit retained 92 eligible studies; focused screening within the 2017–2026 window retained 59 studies, including 56 method-bearing studies and 3 evaluation-focused studies. The records describe the reviewed literature; they are not a new metaphor-annotation benchmark. Search-audit reconstruction and the absence of original database exports are documented in S1. The archive does not include copyrighted article full texts or third-party benchmark datasets. Files: Supplementary_File_S1.xlsx: eligibility audit, inclusion records, review documentation, and search-audit notes. Supplementary_File_S2.xlsx: study-level primary evidential-anchor and evaluation-support coding and definitions. Figure_4_counts.csv: category sizes and grouped Y/P/N/NA support-status counts for R and V, aggregated from S2. reproduce_Figure_4.py: reproducible plotting code reading S2 and writing PDF, SVG, PNG, and grouped counts. README.txt: file inventory, dependencies, and reproduction instructions.
Files
Steps to reproduce
Install Python 3, matplotlib, numpy, and openpyxl. Times New Roman fonts are needed to match the submitted figure. Place Supplementary_File_S2.xlsx and reproduce_Figure_4.py in the same folder. Run: python reproduce_Figure_4.py Supplementary_File_S2.xlsx figure4_output . The script reads the Study_Coding_59 worksheet, checks the 59-study total and the 18/15/23/3 category counts, calculates R/V status counts, and writes Figure_4.pdf, Figure_4.svg, Figure_4.png (600 dpi), and Figure_4_counts.csv. Compare the generated counts with the supplied CSV. Coding definitions and search-audit notes are retained in S1 and S2. The procedure reproduces Figure 4 from the coded records; it does not rerun literature searches or independently recode studies.
Institutions
- Sichuan International Studies UniversityChongqing, Chongqing