Update dataset card
Browse files
README.md
CHANGED
@@ -1,36 +1,54 @@
|
|
1 |
-
|
2 |
-
|
3 |
-
|
4 |
-
|
5 |
-
|
6 |
-
|
7 |
-
|
8 |
-
|
9 |
-
|
10 |
-
|
11 |
-
|
12 |
-
|
13 |
-
|
14 |
-
|
15 |
-
|
16 |
-
|
17 |
-
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
-
|
22 |
-
|
23 |
-
|
24 |
-
-
|
25 |
-
|
26 |
-
|
27 |
-
|
28 |
-
|
29 |
-
|
30 |
-
-
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
|
2 |
+
# FinQA Dataset (Processed)
|
3 |
+
|
4 |
+
## Dataset Description
|
5 |
+
|
6 |
+
### Dataset Summary
|
7 |
+
The FinQA dataset is designed for numerical reasoning over financial data, containing questions that require complex reasoning over tables and text from financial reports.
|
8 |
+
|
9 |
+
### Dataset Statistics
|
10 |
+
- Total examples: 8281
|
11 |
+
- Training set size: 7134 examples (combined train + dev)
|
12 |
+
- Test set size: 1147 examples
|
13 |
+
|
14 |
+
### Dataset Structure
|
15 |
+
Each example contains:
|
16 |
+
- Required columns:
|
17 |
+
- query: The question to be answered (derived directly from qa.question)
|
18 |
+
- context: Combined context including pre-text, table, and post-text, formatted with random section headers and separators for variety
|
19 |
+
- output: The execution answer (derived from qa.exe_ans)
|
20 |
+
- Original FinQA fields:
|
21 |
+
- id: Unique example identifier
|
22 |
+
- pre_text: Text appearing before the table
|
23 |
+
- post_text: Text appearing after the table
|
24 |
+
- table: Tabular data in string format
|
25 |
+
- program: The reasoning program to derive the answer
|
26 |
+
- exe_ans: The execution result
|
27 |
+
|
28 |
+
### Context Formation
|
29 |
+
The context field is created by concatenating:
|
30 |
+
1. Pre-text with a randomly selected header (e.g., "Background:", "Context:", "Pre-text:")
|
31 |
+
2. Table data with a randomly selected header (e.g., "Data Table:", "Tabular Data:", "Table:")
|
32 |
+
3. Post-text with a randomly selected header (e.g., "Additional Information:", "Follow-up:", "Post-table:")
|
33 |
+
|
34 |
+
These sections are joined using random separators (##,
|
35 |
+
|
36 |
+
, or --) to create variety.
|
37 |
+
|
38 |
+
## Dataset Creation
|
39 |
+
|
40 |
+
### Source Data
|
41 |
+
This dataset is derived from the FinQA dataset created by Chen et al. The original dataset is available at [FinQA GitHub Repository](https://github.com/czyssrs/FinQA).
|
42 |
+
|
43 |
+
### Citation
|
44 |
+
|
45 |
+
```
|
46 |
+
@article{chen2021finqa,
|
47 |
+
title={FinQA: A Dataset of Numerical Reasoning over Financial Data},
|
48 |
+
author={Chen, Zhiyu and Chen, Wenhu and Smiley, Charese and Shah, Sameena and Borova, Iana and Langdon, Dylan and Moussa, Reema and Beane, Matt and Huang, Ting-Hao and Routledge, Bryan and Wang, William Yang},
|
49 |
+
journal={Proceedings of EMNLP 2021},
|
50 |
+
year={2021}
|
51 |
+
}
|
52 |
+
```
|
53 |
+
### Licensing Information
|
54 |
+
This dataset is released under the MIT License, following the original FinQA dataset licensing terms.
|