Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

An Algorithmic Framework for Systematic Literature Reviews: A Case Study for Financial Narratives

Gabin Taibi, Joerg Osterrieder

TL;DR

This paper addresses the fragmentation in the study of financial narratives by presenting an algorithmic systematic literature review framework that integrates NLP, transformer-based representations, dimensionality reduction, and clustering to produce a reproducible literature selection. Applied to a Scopus-derived corpus on financial narratives, the framework distills an initial 288 publications down to 16 core papers categorized into narrative understanding and narrative modeling, illustrating how narratives provide predictive content beyond traditional sentiment or topic analyses. The study demonstrates methodological progress toward holistic narrative modeling, including semantic shift, emotional content, and narrative networks, and shows that narrative-based signals can improve forecasting of asset prices, volatility, and macro variables. Limitations include data representativeness, interpretability of complex NLP models, and regime validation, with future work aimed at standard benchmarks and automation to scale literature synthesis in narrative economics. Overall, the framework offers a practical, reproducible approach for conducting literature reviews in finance that can inform both theory and market applications.

Abstract

This paper introduces an algorithmic framework for conducting systematic literature reviews (SLRs), designed to improve efficiency, reproducibility, and selection quality assessment in the literature review process. The proposed method integrates Natural Language Processing (NLP) techniques, clustering algorithms, and interpretability tools to automate and structure the selection and analysis of academic publications. The framework is applied to a case study focused on financial narratives, an emerging area in financial economics that examines how structured accounts of economic events, formed by the convergence of individual interpretations, influence market dynamics and asset prices. Drawing from the Scopus database of peer-reviewed literature, the review highlights research efforts to model financial narratives using various NLP techniques. Results reveal that while advances have been made, the conceptualization of financial narratives remains fragmented, often reduced to sentiment analysis, topic modeling, or their combination, without a unified theoretical framework. The findings underscore the value of more rigorous and dynamic narrative modeling approaches and demonstrate the effectiveness of the proposed algorithmic SLR methodology.

An Algorithmic Framework for Systematic Literature Reviews: A Case Study for Financial Narratives

TL;DR

This paper addresses the fragmentation in the study of financial narratives by presenting an algorithmic systematic literature review framework that integrates NLP, transformer-based representations, dimensionality reduction, and clustering to produce a reproducible literature selection. Applied to a Scopus-derived corpus on financial narratives, the framework distills an initial 288 publications down to 16 core papers categorized into narrative understanding and narrative modeling, illustrating how narratives provide predictive content beyond traditional sentiment or topic analyses. The study demonstrates methodological progress toward holistic narrative modeling, including semantic shift, emotional content, and narrative networks, and shows that narrative-based signals can improve forecasting of asset prices, volatility, and macro variables. Limitations include data representativeness, interpretability of complex NLP models, and regime validation, with future work aimed at standard benchmarks and automation to scale literature synthesis in narrative economics. Overall, the framework offers a practical, reproducible approach for conducting literature reviews in finance that can inform both theory and market applications.

Abstract

This paper introduces an algorithmic framework for conducting systematic literature reviews (SLRs), designed to improve efficiency, reproducibility, and selection quality assessment in the literature review process. The proposed method integrates Natural Language Processing (NLP) techniques, clustering algorithms, and interpretability tools to automate and structure the selection and analysis of academic publications. The framework is applied to a case study focused on financial narratives, an emerging area in financial economics that examines how structured accounts of economic events, formed by the convergence of individual interpretations, influence market dynamics and asset prices. Drawing from the Scopus database of peer-reviewed literature, the review highlights research efforts to model financial narratives using various NLP techniques. Results reveal that while advances have been made, the conceptualization of financial narratives remains fragmented, often reduced to sentiment analysis, topic modeling, or their combination, without a unified theoretical framework. The findings underscore the value of more rigorous and dynamic narrative modeling approaches and demonstrate the effectiveness of the proposed algorithmic SLR methodology.
Paper Structure (18 sections, 2 figures, 1 table)

This paper contains 18 sections, 2 figures, 1 table.

Table of Contents

  1. Introduction
  2. Methodology
  3. Initial research sourcing
  4. Algorithmic selection framework
  5. Textual analysis: research properties statements
  6. Data preparation
  7. Research clustering
  8. Data cleaning and extraction
  9. Results
  10. Selection Phase Results
  11. Narrative Understanding Papers
  12. Narrative Modeling Papers
  13. Discussion
  14. Toward a Shared Understanding of Financial Narratives
  15. Methodological Advances in Narrative Quantification
  16. ...and 3 more sections

Figures (2)

  • Figure 1: Temporal distributions of selected research papers.
  • Figure 2: Schematic representation of the paper selection process. The process includes database querying, filtering by inclusion/exclusion criteria, algorithmic selection via NLP and clustering, and manual exclusion of inaccessible studies or workshop proceeds.