Pith. sign in

REVIEW 4 major objections 5 minor 30 references

Integrating Natural Language Processing Techniques of Text Mining Into Financial System: Applications and Limitations

T0 review · 4 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A review of 60 papers finds asset pricing and stock prediction dominate NLP use in finance from 2018-2023, led by classification techniques and LSTM/BERT-type models, with hybrids at 47%.

desk verdict A plausible but non-reproducible mini-survey of NLP in finance: the challenge table is nice, the counts don't add up. read the letter →

arxiv 2412.20438 v1 pith:3DTP35KT submitted 2024-12-29 cs.CL cs.AIecon.GNq-fin.EC

Denisa Millo , Blerina Vika , Nevila Baci This is my paper · ORCID
classification cs.CLcs.AIecon.GNq-fin.EC
keywords naturallanguageprocessingtextminingfinancialsystemassetpricingstockpredictioninformationclassificationLSTMBERT
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper is a literature review asking which NLP models and text-mining techniques are used across components of the financial system—asset pricing, corporate finance, derivatives, risk management, portfolio theory, public and international finance—between 2018 and 2023. It claims that asset pricing, especially stock prediction, is the most studied component; information classification is the most used information-processing technique; and LSTM- and BERT-type models are the most common algorithms, with about 47% of the 60 surveyed materials combining multiple models or proposing new ones. The authors also say most work mixes probabilistic with vector-space models and textual with numerical data. On the limitations side, they update an earlier challenge list and find that data quality, financial-context adaptation, and model interpretability remain unresolved. A sympathetic reader would care because the review offers a map of where NLP-finance research concentrates and where the bottlenecks are.

What carries the argument

The carrying object is a three-axis classification scheme applied to the surveyed literature: financial-system component, information-processing NLP technique (retrieval, classification, extraction, or combination), and specific algorithm or model family (probabilistic, vector-space, LSTM, BERT-type, and hybrid variants). This scheme produces the histograms in Figures 1-3 that support the frequency claims, and it organizes the proposed workflow in Figures 4-5, where each language-analysis stage (lexical, syntactic, semantic, pragmatic, discourse) is paired with the persistent limitations the authors identify.

What would settle it

Re-run the stated literature search with a published list of included papers and the raw count per financial-system component, technique, and algorithm; if asset pricing no longer dominates or hybrid combinations do not come near 47%, the review's central descriptive claim is falsified.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central discovery is a frequency distribution over a manually collected corpus of about 60 scientific materials from 2018-2023: asset pricing (particularly stock prediction) is the financial-system component receiving the most NLP attention, corporate finance is second, and information classification is the dominant information-processing technique, with BERT-type and LSTM models the most frequently used algorithms. Roughly 47% of the surveyed materials use hybrid combinations or newly proposed methods such as NumHTML. The paper further finds that most models combine probabilistic and vector-space representations and that text signals are typically fused with numerical data, and it presents an engineering-oriented workflow in which a researcher selects a financial-system component, an NLP technique and model family, and an analysis stage, while checking the limitations attached to that stage.

Load-bearing premise

The load-bearing premise is that the roughly 60 papers the authors selected from their keyword search and manual filtering fairly represent NLP research in finance during 2018-2023, even though the inclusion criteria, exclusion rules, and raw counts behind the histograms are not reported.

Editorial extensions

If this is right

  • If the frequency claims hold, asset pricing and stock prediction are where NLP-finance methods are maturing fastest, leaving public finance, derivatives, and other understudied components as open ground for new applications.
  • The dominance of information classification suggests that labeling and categorizing financial texts is the proven core, so progress on extraction, retrieval, and combined techniques could shift the field's center of gravity.
  • The prevalence of LSTM and BERT-type models plus 47% hybrids implies that domain-specific architectures and fine-tuned pretrained models, not generic pipelines, are the current performance frontier.
  • Persistent limits on data quality, financial lexicons, time-varying distributions, and interpretability mean that gains from larger models will be capped until annotation standards and context-adaptive methods improve.
  • The proposed engineering workflow gives a practical starting point: pick the component, technique, and analysis stage explicitly, and anticipate the known failure modes at each stage before building the system.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The pattern the paper observes is likely driven by data availability rather than economic importance: open stock and news data make asset-pricing experiments cheap, so the same distribution may not reflect where NLP could add the most value.
  • Because the paper does not publish its screening criteria or the list of 60 materials, the 47% hybrid figure and component shares are not yet externally auditable; a reproducible protocol with raw counts would settle their stability.
  • The challenge table implies a concrete research agenda: build sector-specific financial lexicons and annotation sets, adapt models to non-IID and time-varying text streams, and design interpretability tools for finance-domain users.
  • One testable extension is to run the same classification scheme on the 2024-2025 literature to see whether ChatGPT-era LLMs displace LSTM and BERT as the default financial-text models.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper is a literature review of natural language processing (NLP) and text-mining applications in the financial system over 2018–2023. It claims that asset pricing, especially stock prediction, is the most studied component; information classification is the most used NLP technique; LSTM and BERT-type models are the most common algorithms; and hybrid/'other' techniques account for 47% of the 60 materials analyzed. It also lists open challenges such as data quality, context adaptation, and model interpretability, and proposes a workflow for analyzing financial text.

Significance. If the quantitative claims were properly supported, the paper would offer a convenient descriptive snapshot of a fast-moving interdisciplinary area. The qualitative challenge list is plausible and consistent with prior reviews, and the proposed workflow could be a useful starting point for practitioners. However, the paper's central contribution is empirical: the frequency distributions in Figures 1–3 are the main results. As it stands, these numbers cannot be checked because the sample is not documented, the search query is incomplete, and at least one figure contains a clear arithmetic inconsistency. The paper therefore currently falls short of the reproducibility standard expected for quantitative survey claims.

major comments (4)
  1. [Section 2.1] The sample construction is not reproducible. The query is given with a placeholder (“*FS element mentioned in chapter 1*”) rather than an actual search string, and no inclusion/exclusion criteria, list of the 109 screened materials, list of the ~60 classified materials, or raw per-category counts are provided. This makes the central descriptive claims (asset pricing most studied, information classification most used, LSTM/BERT most common, 47% hybrid share) unverifiable. The authors should supply the full query, screening protocol, and a supplementary table with the coding of each included material.
  2. [Figure 3] The percentages printed in Figure 3 sum to 102% (2+10+3+2+2+2+2+5+2+10+12+3+47), indicating either overlapping categories or an arithmetic/coding error. Since the 47% 'other/hybrid' share is a headline result, the figure must be corrected and the underlying counts reported so the sum can be checked.
  3. [Table 1 and Section 3] Table 1 marks 'restriction to confidential data' as 'solved' in the current research, but the text in Section 3 states that this limitation 'is not fully appointed yet' and later says it 'resonate[s] a need to address them as quickly as possible.' This is an internal contradiction about one of the paper's substantive claims concerning open challenges and should be resolved.
  4. [Section 2.2] The claim that 'most of the research materials combined probabilistic with vector-space models' is presented without any counts, definitions of 'probabilistic' and 'vector-space,' or coding rules. Without a transparent classification protocol, a reader cannot assess whether this statement is supported by the reviewed materials; it needs to be either operationalized or reformulated as a qualitative observation.
minor comments (5)
  1. [Section 2.1] The sentence 'There are included citations and patents with the query' is confusing; the method later describes screening 'around 250 scientific materials' of which '35 of them books and the rest papers and articles.' Please clarify whether patents were actually included and how.
  2. [References] References [21] and [28] are the same work (Malandri et al., 2018) but are cited as if they were distinct sources. Please merge them and renumber.
  3. [Section 2.2] The term 'NEUS' appears in the list of techniques but is never defined. GNUS (Generalized News-based Sentiment Analysis) is also mentioned; please make the abbreviations consistent and define each one at first use.
  4. [Abstract and Section 2.2] The abstract mentions 'bidirectional encoder models' while the body refers to 'BERT types'; please use consistent terminology.
  5. [Figure captions and text] Figure captions contain stray commas and incomplete phrasing (e.g., 'years 2018-2023,'). Also, 'club inter-domain results' in the Discussion is an unclear phrase; please reword.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the paper is a descriptive literature survey whose claims summarize external cited work; no prediction or derivation reduces to its own inputs.

full rationale

The paper makes no derivation and fits no parameters. Its central claims—asset pricing is the most studied financial-system component, information classification is the most used NLP technique, and LSTM/BERT-type models are the most used algorithms—are empirical tallies of roughly sixty external scientific materials described in Sections 2.1 and 2.2. The categories (asset pricing, corporate finance, derivatives, risk management, public finance, etc.) are defined in the introduction from cited prior work [1] and standard NLP terminology, and the reported frequencies are presented as observations about the surveyed literature, not as outputs computed from the paper's own conclusions. There is no equation in which a predicted quantity is defined in terms of a fitted input, no parameter fitted to a subset of data and then called a prediction, and no 'uniqueness theorem' invoked to force a choice. The paper also contains no load-bearing self-citations: none of the references are authored by Millo, Vika, or Baci, and the few same-journal citations ([13], [14]) are external works used as examples, not as justification for the paper's frequency claims. The proposed engineering path in Figures 4 and 5 is an expository recommendation built on Jurafsky and Martin [2] and limitations gathered from the reviewed papers, not a derivation that presupposes its own result. The review's weaknesses—undocumented inclusion criteria, no listing of the sixty materials, and the Figure 3 percentages summing to 102%—are reproducibility and accuracy concerns, not circularity. Because no claim in the paper is equivalent by construction to an input, the circularity score is 0.

Assumptions & free parameters 0 free parameters · 2 assumptions · 0 invented entities

The review relies on two unverified assumptions about sample representativeness and classification reliability. There are no free parameters because the paper makes no fitted model or derivation, and no invented entities.

assumptions (2)
  • domain assumption The Google Scholar query and manual filtering produce a representative sample of 2018-2023 NLP-in-finance research.
    The frequency claims in Section 2.2 depend on the sample being representative, but inclusion/exclusion criteria are not specified.
  • domain assumption Manual classification of papers into FS components and NLP techniques is reliable and consistent.
    No coding protocol, definitions, or inter-rater checks are reported; Figure 1-3 categories may overlap ('Combination', 'Other').

how reviews work

0 comments
Cite this review

Pith. "Pith review of Integrating Natural Language Processing Techniques of Text Mining Into Financial System: Applications and Limitations." pith.science (2026). https://pith.science/paper/3DTP35KT

@misc{pith2026241220438,
  author       = {Pith},
  title        = {Pith review of: Integrating Natural Language Processing Techniques of Text Mining Into Financial System: Applications and Limitations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3DTP35KT}},
  note         = {Machine review of arXiv:2412.20438}
}
read the original abstract

The financial sector, a pivotal force in economic development, increasingly uses the intelligent technologies such as natural language processing to enhance data processing and insight extraction. This research paper through a review process of the time span of 2018-2023 explores the use of text mining as natural language processing techniques in various components of the financial system including asset pricing, corporate finance, derivatives, risk management, and public finance and highlights the need to address the specific problems in the discussion section. We notice that most of the research materials combined probabilistic with vector-space models, and text-data with numerical ones. The most used technique regarding information processing is the information classification technique and the most used algorithms include the long-short term memory and bidirectional encoder models. The research noticed that new specific algorithms are developed and the focus of the financial system is mainly on asset pricing component. The research also proposes a path from engineering perspective for researchers who need to analyze financial text. The challenges regarding text mining perspective such as data quality, context-adaption and model interpretability need to be solved so to integrate advanced natural language processing models and techniques in enhancing financial analysis and prediction. Keywords: Financial System (FS), Natural Language Processing (NLP), Software and Text Engineering, Probabilistic, Vector-Space, Models, Techniques, TextData, Financial Analysis.

Figures

Figures reproduced from arXiv: 2412.20438 by the authors.

Figure 1
Figure 1. Histogram of scientific materials extracted and classified [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Histogram of scientific materials extracted and [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Histogram of scientific materials extracted and classified [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: The first step is to determine the FS component to work on with and the NLP technique regarding information processing and model algorithms. In the left side FS components, in the right side the most used NLP techniques for each component. In the bottom the most probab…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 28 canonical work pages

  1. [1]

    Financial Economics

    T. Hens, M.O. Rieger, “Financial Economics ”, Springer Texts in Business and Economics, Springer Berlin Heidelberg, pp. 5-6, Heidelberg, Berlin, Germany, 2016

    work page 2016

  2. [2]

    Speech and Language Processing an Introduction to Natural Language Processing

    J. Martin, D. Jurafsky, “Speech and Language Processing an Introduction to Natural Language Processing”, Computational Linguistics, and Speech Recognition, 3rd ed., pp. 3-263, 2020

    work page 2020

  3. [3]

    Deep Learning in Economics: A Systematic and Critical Review

    Y. Zheng, Z. Xu, A. Xiao, “Deep Learning in Economics: A Systematic and Critical Review ”, Arif. Intell. Rev., Vol. 56, pp. 9497-9539, September 2023

    work page 2023

  4. [4]

    Forecasting with Economic News

    L. Barbaglia, S. Consoli, S. Manzan, “Forecasting with Economic News” , Journal of Business and Economic Statistics, Vol. 41, pp. 708-719, July 2023

    work page 2023

  5. [5]

    NumHTML: NumericOriented Hierarchical Transformer Model for Multi-Task Financial Forecasting

    L. Yang, J. Li, R. Dong, Y. Zhang, B. Smyth, “NumHTML: NumericOriented Hierarchical Transformer Model for Multi-Task Financial Forecasting”, AAAI, Vol. 36, pp. 11604-11612, June 2022

    work page 2022

  6. [6]

    Textual Analysis for China’s Financial Markets : a Review and Discussion

    A. Huang, W. Wu, T. Yu, “Textual Analysis for China’s Financial Markets : a Review and Discussion”, China Finance Review International, Vol. 10, No. 1, pp. 1-15, 2020

    work page 2020

  7. [7]

    BERT’s Sentiment Score for Portfolio Optimization: A Fine-Tuned View in Black and Litterman Model

    F. Colasanto, L. Grilli, D. Santoro, G. Villani, “BERT’s Sentiment Score for Portfolio Optimization: A Fine-Tuned View in Black and Litterman Model”, Neural Comput. and Applic., Vol. 34, pp. 17507-17521, October 2022

    work page 2022

  8. [8]

    Analysis of the Fin ancial Information Contained in the Texts of Current Reports: A Deep Learning Approach

    M. Wujec, “Analysis of the Fin ancial Information Contained in the Texts of Current Reports: A Deep Learning Approach”, JRFM, Vol. 14, p. 582, December 2021

    work page 2021

Show all 30 references
  1. [9]

    Machine Learning Meets the Journal of Public Budgeting and Finance: Topics and Trends Over 40 Years

    C. Chen, S. Xiao, B. Zhao, “Machine Learning Meets the Journal of Public Budgeting and Finance: Topics and Trends Over 40 Years”, Public Budgeting and Finance, Vol. 43, pp. 3-23, December 2023

    2023

  2. [10]

    An AI -Enabled Stock Prediction Platform Combining News and Social Sensing with Financial Statements

    T.I. Theodorou, A. Zamichos, M. Skoumperdis, A. Kougioumtzidou, K. Tsolaki, D. Papadopoulos, T. Patsios, G. Papanikolaou, A. Konstantinidis, A. Drosou, D. Tzovaras, “An AI -Enabled Stock Prediction Platform Combining News and Social Sensing with Financial Statements”, Future I...

    2021

  3. [11]

    Application of Machine Learning and Government Finance Statistics for Macroeconomic Signal Mining to Analyze Recessionary Trends and Score Policy Effectiveness

    C. Vuppalapati, A. Ilapakurti, S. Vissapragada, V. Mamaidi, S. Kedari, R. Vuppalapati, S. Kedari, J. Vuppalapati, “Application of Machine Learning and Government Finance Statistics for Macroeconomic Signal Mining to Analyze Recessionary Trends and Score Policy Effectiveness”, ...

    2021

  4. [12]

    Is Trust a Valid Indicator of Tax Compliance Behavior? A Study on Taxpayers’ Public Perception Using Sentiment Analysis Tools

    I. Florina, C. S tefana, C. Codru ta, “Is Trust a Valid Indicator of Tax Compliance Behavior? A Study on Taxpayers’ Public Perception Using Sentiment Analysis Tools”, A.M. Dima, M. Kelemen, (Eds.), “Digitalization and Big Data for Resilience and Economic Intelligence ”, Spring...

    2022

  5. [13]

    Human Disease Detection Using Artif icial Intelligence

    V. Jain, B. Jha, S. Joshi, S. Miglani, A. Singal, S. Babbar, M. Demirci, M.C. Taplamacioglu, “Human Disease Detection Using Artif icial Intelligence” , International Journal on Technical and Physical Problems of Engineering (IJTPE), Issue 55, Vol. 15, No. 2, pp. 125- 133, June 2023

    2023

  6. [14]

    Intelligent Breast Cancer Screening Based on Deep Neural Networks

    H.S. Rahli, N. Benamrane, “Intelligent Breast Cancer Screening Based on Deep Neural Networks”, International Journal on Technical and Physical Problems of Engineering (IJTPE), Issue 57, Vol. 15, No. 4, pp. 404 - 409, December 2023

    2023

  7. [15]

    Comprehensive Review of Text-Mining Applications in Finance

    A. Gupta, V. Dengre, H.A. Kheruwala, M. Shah, “Comprehensive Review of Text-Mining Applications in Finance”, Financ Innov, Vol. 6, p. 39, December 2020

    2020

  8. [16]

    Asset Pricing Via Deep Graph Learning to Incorporate Heterogeneous Predictors

    J. Huang, R. Xing, Q. Li, “Asset Pricing Via Deep Graph Learning to Incorporate Heterogeneous Predictors”, Int J . of Intelligent Sys ., Vol. 37, pp. 8462 -8489, November 2022

    2022

  9. [17]

    A News-Based Machine Learning Model for Adaptive Asset Pricing

    L. Zhu, H. Wu, M.T. Wells, “A News-Based Machine Learning Model for Adaptive Asset Pricing” , Arxiv Preprint Arxiv:2106.07103, 2021

    2021 arXiv

  10. [18]

    Asset Pricing and Deep Learning

    C. Zhang, “Asset Pricing and Deep Learning” , arXiv:2209.12014, Vol. 24, [q-fin.ST], September 2022

    2022 arXiv

  11. [19]

    Predicting Shareholder Litigation on Insider Trading from Financial Text: An Interpretable Deep Learning Approach

    R. Liu, F. Mai, Z. Shan, Y. Wu, “Predicting Shareholder Litigation on Insider Trading from Financial Text: An Interpretable Deep Learning Approach ”, Information and Management, Vol. 57, p. 103387, December 2020

    2020

  12. [20]

    Model Explainability in Deep Learning Based Natural Language Processing

    S. Gholizadeh, N. Zhou, “Model Explainability in Deep Learning Based Natural Language Processing” , arXiv:2106.07410 [cs], June 2021

    2021 arXiv

  13. [21]

    Public Mood- Driven Asset Allocation: The Importance of Financial Sentiment in Portfolio Management

    L. Malandri, F.Z. Xing, C. Orsenigo, C. Vercellis, E. Cambria, “Public Mood- Driven Asset Allocation: The Importance of Financial Sentiment in Portfolio Management”, Cong Compute, Vol. 10, pp. 1167 -1176, December 2018

    2018

  14. [22]

    Quantifying Sentiment with News Media Across Local Housing Markets

    C.K. Soo, “Quantifying Sentiment with News Media Across Local Housing Markets”, The Review of Financial Studies, Vol. 31, pp. 3689-3719, October 2018

    2018

  15. [23]

    CatBoost Model and Artificial Intelligence Techniques for Corporate Failure Predic tion

    S.B. Jabeur, C. Gharib, S. Mefteh Wali, W.B. Arfi, “CatBoost Model and Artificial Intelligence Techniques for Corporate Failure Predic tion”, Technological Forecasting and Social Change, Vol . 166, p p. 120 -658, May 2021

    2021

  16. [24]

    BERT-Based Financial Sentiment Index and LSTM - Based Stock Return Predictability

    J.Z.G. Hiew, X. Huang, H. Mou, D. Li, Q. Wu, Y. Xu, “BERT-Based Financial Sentiment Index and LSTM - Based Stock Return Predictability”, Arxiv:1906.09024 [q- fin.ST], July 2022

    1906 arXiv

  17. [25]

    Automatic Domain Adaptation Outperforms Ma nual Domain Adaptation for Predicting Financial Outcomes

    M. Sedinkina, N. Breitkopf, H. Sch utze, “Automatic Domain Adaptation Outperforms Ma nual Domain Adaptation for Predicting Financial Outcomes” , The 57th International Journal on “Technical and Physical Problems of Engineering” (IJTPE), Iss. 61, Vol. 16, No. 4, Dec. 2024 6 Ann...

    2024

  18. [26]

    Using Financial News Sentiment for Stock Price Direction Prediction

    B. Fazlija , P. Harder, “Using Financial News Sentiment for Stock Price Direction Prediction” , Mathematics, Vol. 10, pp. 21-56, June 2022

    2022

  19. [27]

    ChatGPT: Unlocking the Future of NLP in Finance

    A. Zaremba , E. Demir, “ChatGPT: Unlocking the Future of NLP in Finance” , SSRN Journal, Modern Finance, 2023, Vol. 1, No. 1, pp. 93-98, 2023

    2023

  20. [28]

    Public Mood– Driven Asset Allocation: The Importance of Financial Sentiment in Portfolio Management

    L. Malandri, F.Z. Xing, C. Orsenigo, C. Vercellis, E. Cambria, “Public Mood– Driven Asset Allocation: The Importance of Financial Sentiment in Portfolio Management”, Cong Compute, Vol. 10, pp. 1167 -1176, December 2018

    2018

  21. [29]

    Bank Financial Risk Prediction Model Based on Big Data

    H. Peng, Y. Lin, M. Wu, “Bank Financial Risk Prediction Model Based on Big Data” , Scientific Programming, Vol. 20, No. 22, pp. 1-9, February 2022

    2022

  22. [30]

    Transformers-Based Approach for a Sustainability Term- Based Sentiment Analysis (STBS A)

    B. Sandwidi, S. Pallitharammal Mukkolakal, “Transformers-Based Approach for a Sustainability Term- Based Sentiment Analysis (STBS A)”, The Second Workshop on NLP for Positive Impact (NLP4PI), (Abu Dhabi, United Arab Emirates (Hybrid)), pp. 157 -170, Association for Computation...

    2022

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.