Published September 19, 2025 | Version 1.0.0
Dataset Open

STARQA: A Question Answering Dataset for Complex Analytical Reasoning over Structured Databases

Authors/Creators

  • 1. ROR icon Bloomberg (United States)
  • 2. IIT Delhi

Description

The STARQA dataset consists of complex analytical reasoning questions and answers for three structured databases: IMDB, Olist, and Eurosoccer.

Each entry contains the following fields:

  1. db_id: Name of the database used to generate the question
  2. question: The analytical question posed
  3. question_path: Path to the corresponding code snippet in the folder
  4. gold: the answer to the question

Files

README.md

Files (678.6 kB)

Name Size Download all
md5:ed44442db561a5cb63e4243c74b98a87
7.7 kB Download
md5:286a82b569eb3e9706cf7193bb39b798
10.6 kB Download
md5:21af09a213bfd260c5e1b01f7a77ae8c
12.8 kB Download
md5:15f915b0772b9e1df36e4d6ff066d52a
96.4 kB Download
md5:fd8687b508603d607bda1329c1f4d70f
187.3 kB Download
md5:ad5db8a53ef7650b56d92072b04d23f3
255 Bytes Preview Download
md5:5be8e8662674519fddac8fa646a74892
238.4 kB Download
md5:2018e97a33a60945f002866d0a6a9075
5.6 kB Preview Download
md5:a5837d79a574520853e06a61276b7fce
119.5 kB Download

Additional details

Related works

Cites
Dataset: https://developer.imdb.com/non-commercial-datasets/ (URL)
Dataset: https://www.kaggle.com/datasets/hugomathien/soccer (URL)
Dataset: https://www.kaggle.com/datasets/olistbr/brazilian-ecommerce (URL)

Dates

Available
2025-09-19