FinMR
FinQA is a dataset specifically designed for financial reasoning and question answering. It comprises questions, financial background information, and corresponding answers. The dataset combines text and visual data, with visual data presented as images stored in JSON files. Its structure includes a unique identifier, shared background, shared image path, question text, multiple‑choice options, correct answer, and detailed explanation. Annotations are performed by financial experts to ensure high accuracy and consistency. The dataset may contain inherent biases from source financial documents; users should exercise caution when generalizing model outputs and consider domain‑specific adaptation.
Dataset description and usage context
Financial Multimodal Mathematical Reasoning QA Dataset💰
Dataset Description
FinQA is a dataset for financial reasoning and question answering. It contains questions, financial background information, and corresponding answers. The dataset includes text and visual data, with the visual data represented by images in a JSON file.
Dataset Structure
The dataset contains the following fields:
- ID: Unique identifier for each question.
- Share Context: Background information related to the question.
- Share Image: Path to the associated visual data.
- Question Text: The question to be answered.
- Options: Multiple‑choice options.
- Answer: Correct answer.
- Explanation: Detailed explanation of the answer.
Annotation Process
Annotations are performed by financial experts to ensure high accuracy and consistency.
Bias, Risks, and Limitations
The dataset may contain inherent biases originating from its source financial documents. Users should be cautious when generalizing results.
Recommendations
Avoid over‑generalization of model outputs and consider domain‑specific adaptation.
Citation
BibTeX:
@article{article_id,
title = {Enhancing Multimodal Financial Math Reasoning with Reflection Module and Error Log},
author = {Shuangyan Deng, Haizhou Peng, ChunHou Liu, Jiachen Xu},
year = {2024},
journal = {arXiv},
primaryClass={cs.CV}
}
Dataset Card Authors
[The University of Auckland]
Dataset Card Contact
Pair the dataset with AI analysis and content workflows.
Once the source passes your review, move straight into summarization, transformation, report drafting, or presentation generation with the JuheAI toolchain.