An Algorithmic Framework for Systematic Literature Reviews: A Case Study for Financial Narratives
Gabin Taibi, Joerg Osterrieder · 2026
Read An Algorithmic Framework for Systematic Literature Reviews: A Case Study for Financial Narratives by Gabin Taibi, Joerg Osterrieder (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-03794.txt.
Introduction
Gabin Taibi and Joerg Osterrieder propose an algorithmic framework for systematic literature reviews, the careful surveys scholars use to summarize a field. Done by hand, reviews are slow and their study selection is hard to reproduce or audit. Their method combines natural language processing, grouping algorithms, and clarity tools to automate and structure how papers are picked and studied.
They demonstrate the framework on financial narratives, an emerging topic about how shared stories of economic events move markets and asset prices, drawing on the Scopus database. The review found real progress in modeling such narratives with language techniques, but also fragmentation, with the concept often cut down to mere sentiment. A repeatable review pipeline could help many fields keep up with publication growth without lowering standards.
Catalog record
| CREATOR | Gabin Taibi, Joerg Osterrieder |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 55 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2601-03794
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
An Algorithmic Framework for Systematic Literature Reviews: A Case Study for Financial Narratives Gabin Taibi, Joerg Osterrieder 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 1 1 1 https://www.scopus.com/ 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. Keywords— Systematic Literature Review, Text Processing, NLP, Financial Narrative, Financial Market Dynamics 1 Introduction The influence of narratives on financial markets has recently become a prominent area of study in both economics and finance. The analysis of narrative includes understanding how stories evolve, spread, and impact financial markets over time. By examining the mechanisms through which they form and propagate, researchers aim to uncover their role in shaping expectations, driving investor behavior, and consequently impact market cycles. In Social Science, narratives are structured accounts or interpretive frameworks through which people understand and act in the social world. The work of 15 advances that they are not simply representations but ontological structures: social life itself is storied. Narratives constitute social identities, guide actions, and embed individuals in relational settings over time and space. These narratives are constituted through emplotment (causal linking of events), selective appropriation, and their temporal and spatial connectivity. In economics, narratives are studied under the term Narrative Economics, a field focusing on how economic agents use stories to navigate and make sense of complex environments. 14 argues that economic fluctuations cannot be fully understood through quantitative models alone, as narratives play a critical role in shaping market movements. Thus, 19 highlights that conviction narratives—coherent and emotionally compelling stories—enable investors to act with confidence despite uncertainty, reinforcing collective behaviors in markets and 13 completes this statement describing economic narratives as “stories that offer interpretations of economic events”. We extend this concept by focusing on financial narratives, a subset of economic narratives, which we define as structured interpretations or explanatory frameworks concerning financial markets or economic events, based on available information. Still, the notion of perfect market efficiency has been challenged by 5 , who argue that if all information were instantly reflected in prices, investors would have no incentive to acquire costly private information. This critique aligns more closely with the semi-strong form of the Efficient Market Hypothesis, which allows for delayed information diffusion. Furthermore, individual market participants do not perceive and process information uniformly, as their interpretations are shaped by personal expectations and objectives, contextual knowledge, and prior experience, leading them to adopt or reject specific perspectives selectively. Yet, we observe that, over time, these individual interpretations eventually converge into dominant narratives that influence collective beliefs which, in turn, shape financial market behavior. Despite growing interest in the field, the concept of financial narratives remains underdefined, and the methodologies used to analyze them are diverse and often inconsistent. The purpose of this systematic literature review is to address this gap by offering a clearer picture of how narratives are conceptualized, modeled, and employed in the study of financial markets. In particular, this review aims to clarify how financial narratives are defined and processed across the literature, and assess their reported impact on financial or macroeconomic variables. It also provides a comprehensive overview of the textual analysis techniques employed, with particular attention to the evolution of methods following the introduction of transformer-based architectures. In doing so, the review assesses whether significant methodological advances have been made since early contributions, aiming to identify both conceptual gaps and best practices in this emerging research area. The study addresses the following research questions: (i) How can NLP and textual analysis techniques be used to quantify and model financial narratives? (ii) Can financial narratives modeling enhance financial market dynamics understanding? More importantly, the relevance of this research lies not only in its theoretical insights but also in its methodological contributions. In fact, the review advances automated knowledge discovery by advanced NLP techniques and machine learning to systematically identify, evaluate, and extract relevant academic literature. This approach contributes to the development of scalable, data-driven techniques for literature synthesis and lays the groundwork for more reproducible research. The proposed framework incorporates transformer-based sentence embedding, dimensionality reduction, and clustering to evaluate relevance against pre-defined research criteria. By integrating these components, the framework not only enables a more systematic and reproducible literature review process, but also improves the assessment of selection quality, ensuring that the selected studies align closely with the research questions. Furthermore, textual analysis is employed to provide researchers with a structured understanding of the remaining publications, thus facilitating the organization and synthesis of findings, as well as the assessment of the thematic relevance and quality of the selected studies. The review is structured as follows: Section 2 presents the algorithmic framework for systematic literature reviews, including the query design, filtering steps, clustering approach, and quality assessment procedures. Section 3 applies the methodology to the domain of financial narratives, analyzing research trends, modeling techniques, and conceptual challenges identified in the selected literature. Section 4 discusses the implications of the findings, limitations of existing approaches, and potential directions for future research in financial narrative modeling. 2 Methodology Our approach in this literature review is inspired by the work of 2 , who structured the review process into eight distinct steps: defining the research question, developing and implementing the review methodology, conducting literature exploration and analysis, applying selection criteria (inclusion/exclusion), assessing the quality of selected studies, extracting relevant data, synthesizing findings, and reporting insights. This structured framework, actually based on 21 , starts by defining the research problem and formulating clear research questions. A structured review methodology is then developed to guide literature search and analysis. Thus, specific inclusion criteria are applied to filter studies and assess their relevance and quality. Once the final set of papers is selected, key information is extracted and synthesized to generate meaningful insights, organized and presented in a structured report. Table 1: Summary of article selection criteria. Criteria Decision Inclusion of pre-defined keywords in title, abstract, or keyword list Inclusion Article not published in a scientific journal Exclusion Article in a language other than English Exclusion Article published before 2010 Exclusion Duplicates Exclusion Low relevance to research questions Exclusion Unavailability of the article online Exclusion Workshop Proceeds Exclusion However, because the core phase of the current methodology, the study selection, depends heavily on the researcher’s judgment, the overall process remains subjective, difficult to replicate, and prone to bias. To address this, we extend the framework