Abstract Semantic Atlas Method A traceable protocol for mapping a research program from abstracts (Demonstration on 2010–2026 abstracts; N = 131)
Description
A traceable, reproducible semantic atlas of 131 abstracts (2010–2026), published as a versioned Zenodo research object (Version DOI: 10.5281/zenodo.18668427; Concept DOI: 10.5281/zenodo.18667757) with registers, theme maps, similarity diagnostics, vocabulary-shift tables, and regeneration code.
Abstract
Researchers who publish many short working papers, notes, and methodological sketches often accumulate a body of work whose internal structure is obvious to the author but opaque to readers. Existing summaries tend to rely on citations, venues, or informal narrative and rarely provide a transparent map of what the texts are about and how a program evolves. This paper introduces the Abstract Semantic Atlas Method (ASAM): a lightweight, traceable protocol that converts a dated list of abstracts into a semantic atlas consisting of (i) theme families with counts and exemplars, (ii) a time-evolution trajectory (phase shift), (iii) a similarity-based core spine vs satellites structure, and (iv) a vocabulary-shift contrast between eras. Traceable means each theme label is grounded in documented top terms and anchored by exemplar abstracts so readers can relabel themes, recode assignments, add new items, and compare versions reproducibly. A worked demonstration on a 2010–2026 abstract set (N = 131) shows a coherent program structure: early emphasis on quantitative finance diagnostics, a mid-period shift to mining valuation objects, and a recent emphasis on governance and institutional design alongside research-object infrastructure methods. The contribution is methodological: ASAM provides a reusable, audit-friendly way to summarize a research program using only abstract texts.