SciQ (Scientific Question Answering) by Kaggle | Academic Research

About this Dataset

SciQ (Scientific Question Answering)

SciQ: A Dataset for Science Question Answering

The Next Generation Science Standards

Source

Huggingface Hub: link

About this dataset

The SciQ dataset contains 13,679 crowdsourced science exam questions about Physics, Chemistry and Biology, among others. The questions are in multiple-choice format with 4 answer options each. For the majority of the questions, an additional paragraph with supporting evidence for the correct answer is provided.

How to use the dataset

This dataset consists of science questions and their corresponding distractors, correct answers, and supports. The questions are designed to evaluate a person's knowledge of science. The distractors are designed to confuse the test taker and lead them away from the correct answer. The correct answers are provided so that the test taker can check their work. The supports are designed to help the test taker understand the question and find the correct answer

Research Ideas

Train a model to answer scientific questions

Acknowledgements

License

> License: CC0 1.0 Universal (CC0 1.0) - Public Domain Dedication
> No Copyright - You can copy, modify, distribute and perform the work, even for commercial purposes, all without asking permission. See Other Information.

Columns

File: validation.csv

Column name	Description
question	The question text. (String)
distractor3	One of the distractors for the question. (String)
distractor1	One of the distractors for the question. (String)
distractor2	One of the distractors for the question. (String)
correct_answer	The correct answer for the question. (String)
support	The supporting text for the question. (String)

File: train.csv

Column name	Description
question	The question text. (String)
distractor3	One of the distractors for the question. (String)
distractor1	One of the distractors for the question. (String)
distractor2	One of the distractors for the question. (String)
correct_answer	The correct answer for the question. (String)
support	The supporting text for the question. (String)

File: test.csv

Column name	Description
question	The question text. (String)
distractor3	One of the distractors for the question. (String)
distractor1	One of the distractors for the question. (String)
distractor2	One of the distractors for the question. (String)
correct_answer	The correct answer for the question. (String)
support	The supporting text for the question. (String)

Tables

Test

@kaggle.thedevastator_sciq_a_dataset_for_science_question_answering.test

339.43 KB
1000 rows
6 columns


CREATE TABLE test (
  "question" VARCHAR,
  "distractor3" VARCHAR,
  "distractor1" VARCHAR,
  "distractor2" VARCHAR,
  "correct_answer" VARCHAR,
  "support" VARCHAR
);

Train

@kaggle.thedevastator_sciq_a_dataset_for_science_question_answering.train

3.63 MB
11679 rows
6 columns


CREATE TABLE train (
  "question" VARCHAR,
  "distractor3" VARCHAR,
  "distractor1" VARCHAR,
  "distractor2" VARCHAR,
  "correct_answer" VARCHAR,
  "support" VARCHAR
);

Validation

@kaggle.thedevastator_sciq_a_dataset_for_science_question_answering.validation

334.11 KB
1000 rows
6 columns


CREATE TABLE validation (
  "question" VARCHAR,
  "distractor3" VARCHAR,
  "distractor1" VARCHAR,
  "distractor2" VARCHAR,
  "correct_answer" VARCHAR,
  "support" VARCHAR
);