Datasets:

cardiffnlp
/

databench

question string	answer string	type string	columns_used sequence	column_types sequence	sample_answer string	dataset string
Is the person with the highest net worth self-made?	True	boolean	[ "finalWorth", "selfMade" ]	[ "number[uint32]", "boolean" ]	False	001_Forbes
Does the youngest billionaire identify as male?	True	boolean	[ "age", "gender" ]	[ "number[UInt8]", "category" ]	True	001_Forbes
Is the city with the most billionaires in the United States?	True	boolean	[ "city", "country" ]	[ "category", "category" ]	True	001_Forbes
Is there a non-self-made billionaire in the top 5 ranks?	True	boolean	[ "rank", "selfMade" ]	[ "number[uint16]", "boolean" ]	False	001_Forbes
Does the oldest billionaire have a philanthropy score of 5?	False	boolean	[ "age", "philanthropyScore" ]	[ "number[UInt8]", "number[UInt8]" ]	False	001_Forbes
What is the age of the youngest billionaire?	19.0	number	[ "age" ]	[ "number[UInt8]" ]	32.0	001_Forbes
How many billionaires are there from the 'Technology' category?	343	number	[ "category" ]	[ "category" ]	0	001_Forbes
What's the total worth of billionaires in the 'Automotive' category?	583600	number	[ "category", "finalWorth" ]	[ "category", "number[uint32]" ]	0	001_Forbes
How many billionaires have a philanthropy score above 3?	25	number	[ "philanthropyScore" ]	[ "number[UInt8]" ]	0	001_Forbes
What's the rank of the wealthiest non-self-made billionaire?	3	number	[ "selfMade", "rank" ]	[ "boolean", "number[uint16]" ]	288	001_Forbes
Which category does the richest billionaire belong to?	Automotive	category	[ "finalWorth", "category" ]	[ "number[uint32]", "category" ]	Food & Beverage	001_Forbes
What's the country of origin of the oldest billionaire?	United States	category	[ "age", "country" ]	[ "number[UInt8]", "category" ]	United Kingdom	001_Forbes
What's the gender of the billionaire with the highest philanthropy score?	M	category	[ "philanthropyScore", "gender" ]	[ "number[UInt8]", "category" ]	M	001_Forbes
What's the source of wealth for the youngest billionaire?	drugstores	category	[ "age", "source" ]	[ "number[UInt8]", "category" ]	fintech	001_Forbes
What is the title of the billionaire with the lowest rank?	null	category	[ "rank", "title" ]	[ "number[uint16]", "category" ]	null	001_Forbes
List the top 3 countries with the most billionaires.	['United States', 'China', 'India']	list[category]	[ "country" ]	[ "category" ]	['United States', 'China', 'Brazil']	001_Forbes
List the top 5 sources of wealth for billionaires.	['real estate', 'investments', 'pharmaceuticals', 'diversified', 'software']	list[category]	[ "source" ]	[ "category" ]	['diversified', 'media, automotive', 'Semiconductor materials', 'WeWork', 'beverages']	001_Forbes
List the top 4 cities where the youngest billionaires live.	[nan, 'Los Angeles', 'Jiaozuo', 'Oslo']	list[category]	[ "age", "city" ]	[ "number[UInt8]", "category" ]	['San Francisco', 'New York', 'Wuhan', 'Bangalore']	001_Forbes
List the bottom 3 categories with the fewest billionaires.	['Logistics', 'Sports', 'Gambling & Casinos']	list[category]	[ "category" ]	[ "category" ]	['Service', 'Fashion & Retail', 'Manufacturing']	001_Forbes
List the bottom 2 countries with the least number of billionaires.	['Colombia', 'Andorra']	list[category]	[ "country" ]	[ "category" ]	['Canada', 'Egypt']	001_Forbes
List the top 5 ranks of billionaires who are not self-made.	[3, 10, 14, 16, 18]	list[number]	[ "selfMade", "rank" ]	[ "boolean", "number[uint16]" ]	[288, 296, 509, 523, 601]	001_Forbes
List the bottom 3 ages of billionaires who have a philanthropy score of 5.	[48.0, 83.0, 83.0]	list[number]	[ "philanthropyScore", "age" ]	[ "number[UInt8]", "number[UInt8]" ]	[]	001_Forbes
List the top 6 final worth values of billionaires in the 'Technology' category.	[171000, 129000, 111000, 107000, 106000, 91400]	list[number]	[ "category", "finalWorth" ]	[ "category", "number[uint32]" ]	[]	001_Forbes
List the bottom 4 ranks of female billionaires.	[14, 18, 21, 30]	list[number]	[ "gender", "rank" ]	[ "category", "number[uint16]" ]	[]	001_Forbes
List the top 2 final worth values of billionaires in the 'Automotive' category.	[219000, 44800]	list[number]	[ "category", "finalWorth" ]	[ "category", "number[uint32]" ]	[]	001_Forbes
Did any children below the age of 18 survive?	True	boolean	[ "Age", "Survived" ]	[ "number[UInt8]", "boolean" ]	True	002_Titanic
Were there any passengers who paid a fare of more than $500?	True	boolean	[ "Fare" ]	[ "number[double]" ]	False	002_Titanic
Is every passenger's name unique?	True	boolean	[ "Name" ]	[ "text" ]	True	002_Titanic
Were there any female passengers in the 3rd class who survived?	True	boolean	[ "Sex", "Pclass", "Survived" ]	[ "category", "number[uint8]", "boolean" ]	True	002_Titanic
How many unique passenger classes are present in the dataset?	3	number	[ "Pclass" ]	[ "number[uint8]" ]	3	002_Titanic
What's the maximum age of the passengers?	80.0	number	[ "Age" ]	[ "number[UInt8]" ]	69.0	002_Titanic
How many passengers boarded without any siblings or spouses?	604	number	[ "Siblings_Spouses Aboard" ]	[ "number[uint8]" ]	12	002_Titanic
On average, how much fare did the passengers pay?	32.31	number	[ "Fare" ]	[ "number[double]" ]	23.096459999999997	002_Titanic
Which passenger class has the highest number of survivors?	1	category	[ "Pclass", "Survived" ]	[ "number[uint8]", "boolean" ]	3	002_Titanic
What's the most common gender among the survivors?	female	category	[ "Sex", "Survived" ]	[ "category", "boolean" ]	female	002_Titanic
Among those who survived, which fare range was the most common: (0-50, 50-100, 100-150, 150+)?	0-50	category	[ "Fare", "Survived" ]	[ "number[double]", "boolean" ]	0-50	002_Titanic
What's the most common age range among passengers: (0-18, 18-30, 30-50, 50+)?	18-30	category	[ "Age" ]	[ "number[UInt8]" ]	18-30	002_Titanic
Name the top 3 passenger classes by survival rate.	[1, 2, 3]	list[category]	[ "Pclass", "Survived" ]	[ "number[uint8]", "boolean" ]	[1, 3, 2]	002_Titanic
Could you list the bottom 3 fare ranges by number of survivors: (0-50, 50-100, 100-150, 150+)?	['50-100', '150+', '100-150']	list[category]	[ "Fare", "Survived" ]	[ "number[double]", "boolean" ]	[50-100, 150+, 100-150]	002_Titanic
What is the top 4 age ranges('30-50', '18-30', '0-18', '50+') with the highest number of survivors?	['30-50', '18-30', '0-18', '50+']	list[category]	[ "Age", "Survived" ]	[ "number[UInt8]", "boolean" ]	[30-50, 18-30, 0-18, 50+]	002_Titanic
What are the top 2 genders by average fare paid?	['female', 'male']	list[category]	[ "Sex", "Fare" ]	[ "category", "number[double]" ]	[female, male]	002_Titanic
What are the oldest 3 ages among the survivors?	[24.0, 22.0, 27.0]	list[number]	[ "Age", "Survived" ]	[ "number[UInt8]", "boolean" ]	[56.0, 47.0, 42.0]	002_Titanic
Which are the top 4 fares paid by survivors?	[13.0, 26.0, 7.75, 10.5]	list[number]	[ "Fare", "Survived" ]	[ "number[double]", "boolean" ]	[133.65, 39.0, 35.5, 30.5]	002_Titanic
Could you list the youngest 3 ages among the survivors?	[53.0, 55.0, 11.0]	list[number]	[ "Age", "Survived" ]	[ "number[UInt8]", "boolean" ]	[14.0, 24.0, 28.0]	002_Titanic
Which are the bottom 4 fares among those who didn't survive?	[90.0, 12.275, 9.35, 10.5167]	list[number]	[ "Fare", "Survived" ]	[ "number[double]", "boolean" ]	[13.0, 7.75, 11.5, 10.1708]	002_Titanic
Is the average age of the respondents above 30?	True	boolean	[ "[", "'", "W", "h", "a", "t", " ", "i", "s", " ", "y", "o", "u", "r", " ", "a", "g", "e", "?", " ", "👶", "🏻", "👵", "🏻", "'", "]" ]	[ "number[uint8]" ]	True	003_Love
Are there more single individuals than married ones in the dataset?	True	boolean	[ "[", "'", "W", "h", "a", "t", " ", "i", "s", " ", "y", "o", "u", "r", " ", "c", "i", "v", "i", "l", " ", "s", "t", "a", "t", "u", "s", "?", " ", "💍", "'", "]" ]	[ "category" ]	False	003_Love
Do the majority of respondents have a height greater than 170 cm?	True	boolean	[ "[", "W", "h", "a", "t", "'", "s", " ", "y", "o", "u", "r", " ", "h", "e", "i", "g", "h", "t", "?", " ", "i", "n", " ", "c", "m", " ", "📏", "]" ]	[ "number[uint8]" ]	True	003_Love
Is the most frequent hair color black?	False	boolean	[ "[", "'", "W", "h", "a", "t", " ", "i", "s", " ", "y", "o", "u", "r", " ", "h", "a", "i", "r", " ", "c", "o", "l", "o", "r", "?", " ", "👩", "🦰", "👱", "🏽", "'", "]" ]	[ "category" ]	False	003_Love
How many unique nationalities are present in the dataset?	13	number	[ "[", "W", "h", "a", "t", "'", "s", " ", "y", "o", "u", "r", " ", "n", "a", "t", "i", "o", "n", "a", "l", "i", "t", "y", "?", "\"", "]", "\"" ]	[ "category" ]	1	003_Love
What is the average gross annual salary?	56332.81720430108	number	[ "[", "'", "G", "r", "o", "s", "s", " ", "a", "n", "n", "u", "a", "l", " ", "s", "a", "l", "a", "r", "y", " ", "(", "i", "n", " ", "e", "u", "r", "o", "s", ")", " ", "💸", "'", "]" ]	[ "number[UInt32]" ]	62710.0	003_Love
How many respondents wear glasses all the time?	0	number	[ "[", "'", "H", "o", "w", " ", "o", "f", "t", "e", "n", " ", "d", "o", " ", "y", "o", "u", " ", "w", "e", "a", "r", " ", "g", "l", "a", "s", "s", "e", "s", "?", " ", "👓", "'", "]" ]	[ "category" ]	0	003_Love
What's the median age of the respondents?	33.0	number	[ "[", "'", "W", "h", "a", "t", " ", "i", "s", " ", "y", "o", "u", "r", " ", "a", "g", "e", "?", " ", "👶", "🏻", "👵", "🏻", "'", "]" ]	[ "number[uint8]" ]	32.5	003_Love
What is the most common level of studies achieved?	Master	category	[ "[", "'", "W", "h", "a", "t", " ", "i", "s", " ", "t", "h", "e", " ", "m", "a", "x", "i", "m", "u", "m", " ", "l", "e", "v", "e", "l", " ", "o", "f", " ", "s", "t", "u", "d", "i", "e", "s", " ", "y", "o", "u", " ", "h", "a", "v", "e", " ", "a", "c", "h", "i", "e", "v", "e", "d", "?", " ", "🎓", "'", "]" ]	[ "category" ]	Master	003_Love
Which body complexity has the least number of respondents?	Very thin	category	[ "[", "'", "W", "h", "a", "t", " ", "i", "s", " ", "y", "o", "u", "r", " ", "b", "o", "d", "y", " ", "c", "o", "m", "p", "l", "e", "x", "i", "t", "y", "?", " ", "🏋", "️", "'", "]" ]	[ "category" ]	Obese	003_Love
What's the most frequent eye color?	Brown	category	[ "[", "'", "W", "h", "a", "t", " ", "i", "s", " ", "y", "o", "u", "r", " ", "e", "y", "e", " ", "c", "o", "l", "o", "r", "?", " ", "👁", "️", "'", "]" ]	[ "category" ]	Brown	003_Love
Which sexual orientation has the highest representation?	Heterosexual	category	[ "[", "W", "h", "a", "t", "'", "s", " ", "y", "o", "u", "r", " ", "s", "e", "x", "u", "a", "l", " ", "o", "r", "i", "e", "n", "t", "a", "t", "i", "o", "n", "?", "\"", "]", "\"" ]	[ "category" ]	Heterosexual	003_Love
List the top 3 most common areas of knowledge.	['[Computer Science]', '[Business]', '[Enginering & Architecture]']	list[category]	[ "[", "'", "W", "h", "a", "t", " ", "a", "r", "e", "a", " ", "o", "f", " ", "k", "n", "o", "w", "l", "e", "d", "g", "e", " ", "i", "s", " ", "c", "l", "o", "s", "e", "r", " ", "t", "o", " ", "y", "o", "u", "?", "'", "]" ]	[ "list[category]" ]	['[Computer Science]', '[Business]', '[Enginering & Architecture]']	003_Love
List the bottom 3 hair lengths in terms of frequency.	['Medium', 'Long', 'Bald']	list[category]	[ "[", "'", "H", "o", "w", " ", "l", "o", "n", "g", " ", "i", "s", " ", "y", "o", "u", "r", " ", "h", "a", "i", "r", "?", " ", "💇", "🏻", "♀", "️", "💇", "🏽", "♂", "️", "'", "]" ]	[ "category" ]	['Short', 'Medium', 'Long']	003_Love
Name the top 5 civil statuses represented in the dataset.	['Single', 'Married', 'In a Relationship', 'In a Relationship Cohabiting', 'Divorced']	list[category]	[ "[", "'", "W", "h", "a", "t", " ", "i", "s", " ", "y", "o", "u", "r", " ", "c", "i", "v", "i", "l", " ", "s", "t", "a", "t", "u", "s", "?", " ", "💍", "'", "]" ]	[ "category" ]	['Married', 'In a Relationship', 'In a Relationship Cohabiting', 'Single', 'Divorced']	003_Love
What are the 4 least common hair colors?	['Red', 'Other', 'White', 'Blue']	list[category]	[ "[", "'", "W", "h", "a", "t", " ", "i", "s", " ", "y", "o", "u", "r", " ", "h", "a", "i", "r", " ", "c", "o", "l", "o", "r", "?", " ", "👩", "🦰", "👱", "🏽", "'", "]" ]	[ "category" ]	['Brown', 'Black']	003_Love
What are the top 4 maximum gross annual salaries?	[500000.0, 360000.0, 300000.0, 300000.0]	list[number]	[ "[", "'", "G", "r", "o", "s", "s", " ", "a", "n", "n", "u", "a", "l", " ", "s", "a", "l", "a", "r", "y", " ", "(", "i", "n", " ", "e", "u", "r", "o", "s", ")", " ", "💸", "'", "]" ]	[ "number[UInt32]" ]	[150000.0, 130000.0, 125000.0, 120000.0]	003_Love
Name the bottom 3 values for the happiness scale.	[2, 2, 2]	list[number]	[ "[", "'", "H", "a", "p", "p", "i", "n", "e", "s", "s", " ", "s", "c", "a", "l", "e", "'", "]" ]	[ "number[uint8]" ]	[7, 10, 6]	003_Love
What are the 5 highest ages present in the dataset?	[65, 62, 60, 60, 59]	list[number]	[ "[", "'", "W", "h", "a", "t", " ", "i", "s", " ", "y", "o", "u", "r", " ", "a", "g", "e", "?", " ", "👶", "🏻", "👵", "🏻", "'", "]" ]	[ "number[uint8]" ]	[65, 60, 51, 50, 50]	003_Love
List the bottom 6 skin tone values based on frequency.	[2, 1, 6, 0, 7, 8]	list[number]	[ "[", "'", "W", "h", "a", "t", " ", "i", "s", " ", "y", "o", "u", "r", " ", "s", "k", "i", "n", " ", "t", "o", "n", "e", "?", "'", "]" ]	[ "number[uint8]" ]	[3, 1, 6, 2, 7, 0]	003_Love
Are there any trips with a total distance greater than 30 miles?	False	boolean	[ "trip_distance" ]	[ "number[double]" ]	False	004_Taxi
Were there any trips that cost more than $100 in total?	False	boolean	[ "total_amount" ]	[ "number[double]" ]	False	004_Taxi
Is there any trip with more than 6 passengers?	False	boolean	[ "passenger_count" ]	[ "number[uint8]" ]	False	004_Taxi
Did all the trips use a payment type of either 1 or 2?	False	boolean	[ "payment_type" ]	[ "number[uint8]" ]	True	004_Taxi
What is the maximum fare amount charged for a trip?	75.25	number	[ "fare_amount" ]	[ "number[double]" ]	85.0	004_Taxi
How many unique pickup locations are in the dataset?	96	number	[ "PULocationID" ]	[ "number[uint16]" ]	193	004_Taxi
What is the average tip amount given by passengers?	2.74	number	[ "tip_amount" ]	[ "number[double]" ]	1.5	004_Taxi
How many trips took place in the airport area?	99807	number	[ "Airport_fee" ]	[ "number[UInt8]" ]	194	004_Taxi
Which payment type is the most common in the dataset?	1	category	[ "payment_type" ]	[ "number[uint8]" ]	1	004_Taxi
Which vendor has the most trips recorded?	2	category	[ "VendorID" ]	[ "number[uint8]" ]	2	004_Taxi
What is the most common drop-off location?	236	category	[ "DOLocationID" ]	[ "number[uint16]" ]	161	004_Taxi
On which date did the first recorded trip occur?	2023-01-31	category	[ "tpep_pickup_datetime" ]	[ "date[ns", "UTC]" ]	2019-01-01 00:46:40	004_Taxi
Which are the top 3 most frequent pickup locations?	[161, 237, 236]	list[category]	[ "PULocationID" ]	[ "number[uint16]" ]	[237, 236, 161]	004_Taxi
Name the 4 most common rate codes used.	[1, 2, 5, 4]	list[category]	[ "RatecodeID" ]	[ "number[uint8]" ]	[1, 2, 5, 3]	004_Taxi
list the 2 most frequent store and forward flags.	['N', 'Y']	list[category]	[ "store_and_fwd_flag" ]	[ "category" ]	['N', 'Y']	004_Taxi
Identify the top 4 payment types used by frequency	[1, 2, 4, 3]	list[category]	[ "payment_type" ]	[ "number[uint8]" ]	[1, 2, 3]	004_Taxi
Report the 4 highest toll amounts paid.	[0, 0, 0, 0]	list[number]	[ "tolls_amount" ]	[ "number[uint8]" ]	[0, 0, 0, 0]	004_Taxi
list the top 3 longest trip distances	[19.83, 19.74, 19.68]	list[number]	[ "trip_distance" ]	[ "number[double]" ]	[8.32, 5.93, 2.8]	004_Taxi
Identify the 5 largest total amounts paid for trips.	[80.0, 80.0, 80.0, 80.0, 79.55]	list[number]	[ "total_amount" ]	[ "number[double]" ]	[45.8, 39.9, 33.2, 25.2, 24.87]	004_Taxi
Report the 6 highest fare amounts charged.	[75.25, 74.4, 73.0, 73.0, 73.0, 73.0]	list[number]	[ "fare_amount" ]	[ "number[double]" ]	[40.8, 28.9, 21.2, 17.0, 14.9, 13.5]	004_Taxi
Are there any complaints made in Brooklyn?	True	boolean	[ "borough" ]	[ "category" ]	True	005_NYC
Do any complaints have 'Dog' as a descriptor?	True	boolean	[ "descriptor" ]	[ "category" ]	False	005_NYC
Were there any complaints raised in April?	True	boolean	[ "month_name" ]	[ "category" ]	True	005_NYC
Is the Mayor's office of special enforcement one of the agencies handling complaints?	True	boolean	[ "agency" ]	[ "category" ]	False	005_NYC
How many complaints have been made in Queens?	23110	number	[ "borough" ]	[ "category" ]	0	005_NYC
What's the total number of unique agencies handling complaints?	22	number	[ "agency" ]	[ "category" ]	7	005_NYC
How many complaints were raised at midnight?	14811	number	[ "hour" ]	[ "number[uint8]" ]	2	005_NYC
How many unique descriptors are present in the dataset?	1131	number	[ "descriptor" ]	[ "category" ]	16	005_NYC
Which borough has the most complaints?	BROOKLYN	category	[ "borough" ]	[ "category" ]	QUEENS	005_NYC
Which month sees the highest number of complaints?	July	category	[ "month_name" ]	[ "category" ]	January	005_NYC
Which weekday has the least complaints?	Sunday	category	[ "weekday_name" ]	[ "category" ]	Thursday	005_NYC
Which agency is least frequently handling complaints?	ACS	category	[ "agency" ]	[ "category" ]	DOHMH	005_NYC
List the top 5 most frequent complaint types.	['Noise - Residential', 'HEAT/HOT WATER', 'Illegal Parking', 'Blocked Driveway', 'Street Condition']	list[category]	[ "complaint_type" ]	[ "category" ]	[HEAT/HOT WATER, Building/Use, Noise - Residential, General Construction/Plumbing, Air Quality]	005_NYC
Which 4 agencies handle the most complaints?	['NYPD', 'HPD', 'DOT', 'DSNY']	list[category]	[ "agency" ]	[ "category" ]	[NYPD, HPD, DOB, DSNY]	005_NYC
Name the 3 least frequent descriptors for complaints.	['Booting Company', 'Ready NY - Businesses', 'Animal']	list[category]	[ "descriptor" ]	[ "category" ]	[Structure - Outdoors, Air: Odor/Fumes, Restaurant (AD2), 12 Dead Animals]	005_NYC

End of preview. Expand in Dataset Viewer.

💾🏋️💾 DataBench 💾🏋️💾

This repository contains the original 65 datasets used for the paper Question Answering over Tabular Data with DataBench: A Large-Scale Empirical Evaluation of LLMs which appeared in LREC-COLING 2024.

Large Language Models (LLMs) are showing emerging abilities, and one of the latest recognized ones is tabular reasoning in question answering on tabular data. Although there are some available datasets to assess question answering systems on tabular data, they are not large and diverse enough to evaluate this new ability of LLMs. To this end, we provide a corpus of 65 real world datasets, with 3,269,975 and 1615 columns in total, and 1300 questions to evaluate your models for the task of QA over Tabular Data.

Usage

from datasets import load_dataset

# Load all QA pairs
all_qa = load_dataset("cardiffnlp/databench", name="qa", split="train")

# Load SemEval 2025 task 8 Question-Answer splits
semeval_train_qa = load_dataset("cardiffnlp/databench", name="semeval", split="train")
semeval_dev_qa = load_dataset("cardiffnlp/databench", name="semeval", split="dev")

You can use any of the individual integrated libraries to load the actual data where the answer is to be retrieved.

For example, using pandas in Python:

import pandas as pd

# "001_Forbes", the id of the dataset
ds_id = all_qa['dataset'][0] 

# full dataset
df = pd.read_parquet(f"hf://datasets/cardiffnlp/databench/data/{ds_id}/all.parquet")

# sample dataset
df = pd.read_parquet(f"hf://datasets/cardiffnlp/databench/data/{ds_id}/sample.parquet")

📚 Datasets

By clicking on each name in the table below, you will be able to explore each dataset.

	Name	Rows	Cols	Domain	Source (Reference)
1	Forbes	2668	17	Business	Forbes
2	Titanic	887	8	Travel and Locations	Kaggle
3	Love	373	35	Social Networks and Surveys	Graphext
4	Taxi	100000	20	Travel and Locations	Kaggle
5	NYC Calls	100000	46	Business	City of New York
6	London Airbnbs	75241	74	Travel and Locations	Kaggle
7	Fifa	14620	59	Sports and Entertainment	Kaggle
8	Tornados	67558	14	Health	Kaggle
9	Central Park	56245	6	Travel and Locations	Kaggle
10	ECommerce Reviews	23486	10	Business	Kaggle
11	SF Police	713107	35	Social Networks and Surveys	US Gov
12	Heart Failure	918	12	Health	Kaggle
13	Roller Coasters	1087	56	Sports and Entertainment	Kaggle
14	Madrid Airbnbs	20776	75	Travel and Locations	Inside Airbnb
15	Food Names	906	4	Business	Data World
16	Holiday Package Sales	4888	20	Travel and Locations	Kaggle
17	Hacker News	9429	20	Social Networks and Surveys	Kaggle
18	Staff Satisfaction	14999	11	Business	Kaggle
19	Aircraft Accidents	23519	23	Health	Kaggle
20	Real Estate Madrid	26026	59	Business	Idealista
21	Telco Customer Churn	7043	21	Business	Kaggle
22	Airbnbs Listings NY	37012	33	Travel and Locations	Kaggle
23	Climate in Madrid	36858	26	Travel and Locations	AEMET
24	Salary Survey Spain 2018	216726	29	Business	INE
25	Data Driven SEO	62	5	Business	Graphext
26	Predicting Wine Quality	1599	12	Business	Kaggle
27	Supermarket Sales	1000	17	Business	Kaggle
28	Predict Diabetes	768	9	Health	Kaggle
29	NYTimes World In 2021	52588	5	Travel and Locations	New York Times
30	Professionals Kaggle Survey	19169	64	Business	Kaggle
31	Trustpilot Reviews	8020	6	Business	TrustPilot
32	Delicatessen Customers	2240	29	Business	Kaggle
33	Employee Attrition	14999	11	Business	Kaggle(modified)
34	World Happiness Report 2020	153	20	Social Networks and Surveys	World Happiness
35	Billboard Lyrics	5100	6	Sports and Entertainment	Brown University
36	US Migrations 2012-2016	288300	9	Social Networks and Surveys	US Census
37	Ted Talks	4005	19	Social Networks and Surveys	Kaggle
38	Stroke Likelihood	5110	12	Health	Kaggle
39	Happy Moments	100535	11	Social Networks and Surveys	Kaggle
40	Speed Dating	8378	123	Social Networks and Surveys	Kaggle
41	Airline Mentions X (former Twitter)	14640	15	Social Networks and Surveys	X (former Twitter)
42	Predict Student Performance	395	33	Business	Kaggle
43	Loan Defaults	83656	20	Business	SBA
44	IMDb Movies	85855	22	Sports and Entertainment	Kaggle
45	Spotify Song Popularity	21000	19	Sports and Entertainment	Spotify
46	120 Years Olympics	271116	15	Sports and Entertainment	Kaggle
47	Bank Customer Churn	7088	15	Business	Kaggle
48	Data Science Salary Data	742	28	Business	Kaggle
49	Boris Johnson UK PM Tweets	3220	34	Social Networks and Surveys	X (former Twitter)
50	ING 2019 X Mentions	7244	22	Social Networks and Surveys	X (former Twitter)
51	Pokemon Features	1072	13	Business	Kaggle
52	Professional Map	1227	12	Business	Kern et al, PNAS'20
53	Google Patents	9999	20	Business	BigQuery
54	Joe Biden Tweets	491	34	Social Networks and Surveys	X (former Twitter)
55	German Loans	1000	18	Business	Kaggle
56	Emoji Diet	58	35	Health	Kaggle
57	Spain Survey 2015	20000	45	Social Networks and Surveys	CIS
58	US Polls 2020	3523	52	Social Networks and Surveys	Brandwatch
59	Second Hand Cars	50000	21	Business	DataMarket
60	Bakery Purchases	20507	5	Business	Kaggle
61	Disneyland Customer Reviews	42656	6	Travel and Locations	Kaggle
62	Trump Tweets	15039	20	Social Networks and Surveys	X (former Twitter)
63	Influencers	1039	14	Social Networks and Surveys	X (former Twitter)
64	Clustering Zoo Animals	101	18	Health	Kaggle
65	RFM Analysis	541909	8	Business	UCI ML

🏗️ Folder structure

Each folder represents one dataset. You will find the following files within:

all.parquet: the processed data, with each column tagged with our typing system, in parquet.
qa.parquet: contains the human-made set of questions, tagged by type and columns used, for the dataset (sample_answer indicates the answers for DataBench lite)
sample.parquet: sample containing 20 rows of the original dataset (DataBench lite)
info.yml: additional information about the dataset

🗂️ Column typing system

In an effort to map the stage for later analysis, we have categorized the columns by type. This information allows us to segment different kinds of data so that we can subsequently analyze the model's behavior on each column type separately. All parquet files have been casted to their smallest viable data type using the open source Lector reader.

What this means is that in the data types we have more granular information that allows us to know if the column contains NaNs or not (following panda’s convention of Int vs int), as well as whether small numerical values contain negatives (Uint vs int) and their range. We also have dates with potential timezone information (although for now they’re all UTC), as well as information about categories’ cardinality coming from the arrow types.

In the table below you can see all the data types assigned to each column, as well as the number of columns for each type. The most common data types are numbers and categories with 1336 columns of the total of 1615 included in DataBench. These are followed by some other more rare types as urls, booleans, dates or lists of elements.

Type	Columns	Example
number	788	55
category	548	apple
date	50	1970-01-01
text	46	A red fox ran...
url	31	google.com
boolean	18	True
list[number]	14	[1,2,3]
list[category]	112	[apple, orange, banana]
list[url]	8	[google.com, apple.com]

🔗 Reference

You can download the paper here.

If you use this resource, please use the following reference:

@inproceedings{oses-etal-2024-databench,
    title = "Question Answering over Tabular Data with DataBench: A Large-Scale Empirical Evaluation of LLMs",
    author = "Jorge Osés Grijalba and Luis Alfonso Ureña-López and
    Eugenio Martínez Cámara and Jose Camacho-Collados",
    booktitle = "Proceedings of LREC-COLING 2024",
    year = "2024",
    address = "Turin, Italy"
}

Downloads last month: 380

Edit dataset card

Size of downloaded dataset files:

1.03 MB

Size of the auto-converted Parquet files:

1.03 MB

Number of rows:

2,616