Numerical Claim Detection in Finance: A Weak-Supervision Approach

Abstract

<p>In the past few years, Transformer based models have shown excellent performance across a variety of tasks and domains. However, the black-box nature of these models, along with their high computing and manual annotation costs have limited adoption of these models. In this paper, we employ a weak-supervision-based approach to alleviate these concerns. We build and compare models for financial claim detection task using sentences with numerical information in analyst reports for more than 1500 public companies in the United States from 2017 to 2020. The proposed aggregation function for the weak-supervision model outperforms existing aggregation functions. In addition to standard performance metrics, we provide cost-value analysis of human-annotation and weak-supervision labeling along with estimates of the carbon footprint of our models. We also analyze the performance of our claim detection models across various industry sectors given the considerable variation in numerical financial claims across industries. Our work highlights the potential of weak supervision models for research at the intersection of Finance and Data Engineering. </p>

FinEntity: Entity-level Sentiment Classification for Financial Texts

“…Moreover, most of the headlines contain only one entity, which makes it close to a sequence-level sentiment dataset. For financial entity tagging dataset, FiNER (Shah et al, 2023b) and FNXL (Sharma et al, 2023) are created for financial entity recognition and numeral span tagging respectively, but both lacks sentiment annotation. Therefore, we aims to bridge this gap by constructing a high-quality, entity-level sentiment classification dataset, which not only label the financial entity spans in sentences, but also annotate their associated sentiments.…”