An Evaluation of Reasoning Capabilities of Large Language Models in Financial Sentiment Analysis

被引:4
|
作者
Du, Kelvin [1 ]
Xing, Frank [2 ]
Mao, Rui [1 ]
Cambria, Erik [1 ]
机构
[1] Nanyang Technol Univ, Sch Comp Sci & Engn, Singapore, Singapore
[2] Natl Univ Singapore, Dept Informat Syst & Analyt, Singapore, Singapore
关键词
financial sentiment analysis; large language models; prompt engineering; RISK; PERFORMANCE;
D O I
10.1109/CAI59869.2024.00042
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Large Language Models (LLMs) have garnered significant attention within the academic community due to their advanced capabilities in natural language understanding and generation. While empirical studies have shed light on LLMs' proficiency in complex task reasoning, a lingering question persists in the field of Financial Sentiment Analysis (FSA): the extent to which LLMs can effectively reason about various financial attributes for FSA. This study employs a prompting framework to investigate this topic, assessing multiple financial attribute reasoning capabilities of LLMs in the context of FSA. By studying relevant literature, we first identified six key financial attributes related to semantic, numerical, temporal, comparative, causal, and risk factors. Our experimental results uncover a deficiency in the financial attribute reasoning capabilities of LLMs for FSA. For example, the examined LLMs such as PaLM-2 and GPT-3.5 display weaknesses in reasoning numerical and comparative attributes within financial texts. On the other hand, explicit prompts related to other financial attributes showcase varied utilities, contributing to LLMs' proficiency in discerning financial sentiment.
引用
收藏
页码:189 / 194
页数:6
相关论文
共 50 条
  • [21] LLMs to the Moon? Reddit Market Sentiment Analysis with Large Language Models
    Deng, Xiang
    Bashlovkina, Vasilisa
    Han, Feng
    Baumgartner, Simon
    Bendersky, Michael
    COMPANION OF THE WORLD WIDE WEB CONFERENCE, WWW 2023, 2023, : 1014 - 1019
  • [22] Sentiment analysis of online responses in the performing arts with large language models
    Seong, Baekryun
    Song, Kyungwoo
    HELIYON, 2023, 9 (12)
  • [23] Revisiting Sentiment Analysis for Software Engineering in the Era of Large Language Models
    Zhang, Ting
    Irsan, Ivana clairine
    Thung, Ferdian
    Lo, David
    ACM TRANSACTIONS ON SOFTWARE ENGINEERING AND METHODOLOGY, 2025, 34 (03)
  • [24] Evaluating Large Language Models in Process Mining: Capabilities, Benchmarks, and Evaluation Strategies
    Berti, Alessandro
    Kourani, Humam
    Haefke, Hannes
    Li, Chiao-Yun
    Schuster, Daniel
    ENTERPRISE, BUSINESS-PROCESS AND INFORMATION SYSTEMS MODELING, BPMDS 2024, EMMSAD 2024, 2024, 511 : 13 - 21
  • [25] Benchmarking Large Language Models on CFLUE - A Chinese Financial Language Understanding Evaluation Dataset
    Zhu, Jie
    Li, Junhui
    Wen, Yalong
    Guo, Lifan
    FINDINGS OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS: ACL 2024, 2024, : 5673 - 5693
  • [26] Are Large Language Models Capable of Causal Reasoning for Sensing Data Analysis?
    Hu, Zhizhang
    Zhang, Yue
    Rossi, Ryan
    Yu, Tong
    Kim, Sungchul
    Pan, Shijia
    PROCEEDINGS OF THE 2024 WORKSHOP ON EDGE AND MOBILE FOUNDATION MODELS, EDGEFM 2024, 2024, : 24 - 29
  • [27] Leveraging hierarchical language models for aspect-based sentiment analysis on financial data
    Lengkeek, Matteo
    Knaap, Finn van der
    Frasincar, Flavius
    INFORMATION PROCESSING & MANAGEMENT, 2023, 60 (05)
  • [28] LogicBench: Towards Systematic Evaluation of Logical Reasoning Ability of Large Language Models
    Parmar, Mihir
    Patel, Nisarg
    Varshney, Neeraj
    Nakamura, Mutsumi
    Luo, Man
    Mashetty, Santosh
    Mitra, Arindam
    Baral, Chitta
    PROCEEDINGS OF THE 62ND ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS, VOL 1: LONG PAPERS, 2024, : 13679 - 13707
  • [29] Towards Reasoning in Large Language Models: A Survey
    Huang, Jie
    Chang, Kevin Chen-Chuan
    FINDINGS OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS, ACL 2023, 2023, : 1049 - 1065
  • [30] Conversations on reasoning: Large language models in diagnosis
    Restrepo, Daniel
    Rodman, Adam
    Abdulnour, Raja-Elie
    JOURNAL OF HOSPITAL MEDICINE, 2024, 19 (08) : 731 - 735