Financial sentiment analysis with FUNNEL : filtered UNion for NER-based ensemble labeling
(2025) In Digital Finance 7(4). p.725-744- Abstract
This paper introduces FUNNEL (Filtered UNion for NER-based Ensemble Labeling), a novel ensemble-based framework for labeling financial news that enhances the reliability of stock-specific sentiment signals. The framework integrates weak keyword-based heuristics with a transformer-based named entity recognition model (spaCy) through a weighted voting scheme, balancing precision and recall. Manual evaluation of 1,400 article–label pairs demonstrates that FUNNEL outperforms the original FNSPID labels in both accuracy and coverage. Applied across seven major companies, the framework reveals systematic differences in sentiment signals produced by FinBERT, RoBERTa, and VADER. These results indicate that integrating different labeling... (More)
This paper introduces FUNNEL (Filtered UNion for NER-based Ensemble Labeling), a novel ensemble-based framework for labeling financial news that enhances the reliability of stock-specific sentiment signals. The framework integrates weak keyword-based heuristics with a transformer-based named entity recognition model (spaCy) through a weighted voting scheme, balancing precision and recall. Manual evaluation of 1,400 article–label pairs demonstrates that FUNNEL outperforms the original FNSPID labels in both accuracy and coverage. Applied across seven major companies, the framework reveals systematic differences in sentiment signals produced by FinBERT, RoBERTa, and VADER. These results indicate that integrating different labeling strategies enhances dataset reliability, coverage, and stability, providing a scalable framework for financial sentiment analysis.
(Less)
- author
- Nordansjö, William LU ; Fourong, Fredrik and Qasim, Muhammad LU
- organization
- publishing date
- 2025-12
- type
- Contribution to journal
- publication status
- published
- subject
- keywords
- Deep learning, FUNNEL, Labeling, Natural language processing, Sentiment analysis, Transformer models
- in
- Digital Finance
- volume
- 7
- issue
- 4
- pages
- 20 pages
- publisher
- Springer International Publishing
- external identifiers
-
- scopus:105020287797
- ISSN
- 2524-6984
- DOI
- 10.1007/s42521-025-00162-3
- language
- English
- LU publication?
- yes
- id
- 71d557c9-dde6-4e26-b4d0-995e6d1e895c
- date added to LUP
- 2025-12-10 09:50:03
- date last changed
- 2025-12-10 09:51:23
@article{71d557c9-dde6-4e26-b4d0-995e6d1e895c,
abstract = {{<p>This paper introduces FUNNEL (Filtered UNion for NER-based Ensemble Labeling), a novel ensemble-based framework for labeling financial news that enhances the reliability of stock-specific sentiment signals. The framework integrates weak keyword-based heuristics with a transformer-based named entity recognition model (spaCy) through a weighted voting scheme, balancing precision and recall. Manual evaluation of 1,400 article–label pairs demonstrates that FUNNEL outperforms the original FNSPID labels in both accuracy and coverage. Applied across seven major companies, the framework reveals systematic differences in sentiment signals produced by FinBERT, RoBERTa, and VADER. These results indicate that integrating different labeling strategies enhances dataset reliability, coverage, and stability, providing a scalable framework for financial sentiment analysis.</p>}},
author = {{Nordansjö, William and Fourong, Fredrik and Qasim, Muhammad}},
issn = {{2524-6984}},
keywords = {{Deep learning; FUNNEL; Labeling; Natural language processing; Sentiment analysis; Transformer models}},
language = {{eng}},
number = {{4}},
pages = {{725--744}},
publisher = {{Springer International Publishing}},
series = {{Digital Finance}},
title = {{Financial sentiment analysis with FUNNEL : filtered UNion for NER-based ensemble labeling}},
url = {{http://dx.doi.org/10.1007/s42521-025-00162-3}},
doi = {{10.1007/s42521-025-00162-3}},
volume = {{7}},
year = {{2025}},
}