Hugging Face's logo

Datasets:
UCSC-IRKM
/
RAGuard

Tasks:
Question Answering
Languages:
English
Size:
10K<n<100K
ArXiv:
License:
Dataset card Data Studio Files
xet
Community
Dataset Preview
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed because of a cast error
Error code:   DatasetGenerationCastError
Exception:    DatasetGenerationCastError
Message:      An error occurred while generating the dataset

All the data files must have the same columns, but at some point there are 5 new columns ({'Document ID', 'Link', 'Document Label', 'Title', 'Full Text'}) and 5 missing columns ({'Verdict', 'Original Verdict', 'Document IDs', 'Claim', 'Document Labels'}).

This happened while the csv dataset builder was generating data using

hf://datasets/UCSC-IRKM/RAGuard/documents.csv (at revision df60141119783e056c2b24abc3d4f78c33c5df68)

Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1871, in _prepare_split_single
                  writer.write_table(table)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/arrow_writer.py", line 643, in write_table
                  pa_table = table_cast(pa_table, self._schema)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/table.py", line 2293, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/table.py", line 2241, in cast_table_to_schema
                  raise CastError(
              datasets.table.CastError: Couldn't cast
              Document ID: int64
              Title: string
              Full Text: string
              Claim ID: int64
              Document Label: string
              Link: string
              -- schema metadata --
              pandas: '{"index_columns": [{"kind": "range", "name": null, "start": 0, "' + 959
              to
              {'Claim ID': Value(dtype='int64', id=None), 'Claim': Value(dtype='string', id=None), 'Verdict': Value(dtype='bool', id=None), 'Document IDs': Value(dtype='string', id=None), 'Document Labels': Value(dtype='string', id=None), 'Original Verdict': Value(dtype='string', id=None)}
              because column names don't match
              
              During handling of the above exception, another exception occurred:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1433, in compute_config_parquet_and_info_response
                  parquet_operations = convert_to_parquet(builder)
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1050, in convert_to_parquet
                  builder.download_and_prepare(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 925, in download_and_prepare
                  self._download_and_prepare(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1001, in _download_and_prepare
                  self._prepare_split(split_generator, **prepare_split_kwargs)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1742, in _prepare_split
                  for job_id, done, content in self._prepare_split_single(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1873, in _prepare_split_single
                  raise DatasetGenerationCastError.from_cast_error(
              datasets.exceptions.DatasetGenerationCastError: An error occurred while generating the dataset
              
              All the data files must have the same columns, but at some point there are 5 new columns ({'Document ID', 'Link', 'Document Label', 'Title', 'Full Text'}) and 5 missing columns ({'Verdict', 'Original Verdict', 'Document IDs', 'Claim', 'Document Labels'}).
              
              This happened while the csv dataset builder was generating data using
              
              hf://datasets/UCSC-IRKM/RAGuard/documents.csv (at revision df60141119783e056c2b24abc3d4f78c33c5df68)
              
              Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

Claim ID
int64
Claim
string
Verdict
bool
Document IDs
string
Document Labels
string
Original Verdict
string
1
" ... following World War II war crime trials were convened. The Japanese were tried and convicted and hung for war crimes committed against American POWs. Among those charges for which they were convicted was waterboarding."
true
[1, 2, 3, 4]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
true
2
"(After the auto bailout) General Motors is back on top as the world’s No. 1 automaker."
true
[5, 6, 7, 8, 9, 10, 11, 12, 13, 14]
['unrelated', 'unrelated', 'unrelated', 'misleading', 'unrelated', 'unrelated', 'unrelated', 'misleading', 'unrelated', 'misleading']
half-true
3
"(Bill) Ayers and Obama ran a radical education foundation together."
false
[15, 16, 17, 18, 19]
['supporting', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
pants-on-fire
4
"(GM and Chrysler) repaid their loans."
true
[20, 21, 22, 23]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
half-true
5
"(McCain) said he was opposed to overturning Roe v. Wade. Now he's for overturning Roe v. Wade."
true
[24, 25, 26, 27]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
half-true
6
"(McCain) was opposed to ethanol. Now he's for it."
true
[28, 29, 30, 31]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-true
7
"(My dad) used to campaign against the gas-guzzling dinosaurs."
true
[32, 33, 34, 35, 36, 37, 38, 39, 40]
['unrelated', 'unrelated', 'supporting', 'unrelated', 'misleading', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
true
8
"(Newt Gingrich) voted in favor of establishing the Department of Education, and yet he gets in a debate and says we should get rid of the Department of Education and send all the education issues back to the states."
true
[41, 42, 43, 44]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
half-true
9
"(Obama's) entire national security team, including his secretary of state, said we want to arm and train and equip (Syrian rebel forces), and he made the unilateral decision to turn them down."
true
[45, 46, 47, 48, 49]
['supporting', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-true
10
"(Romney's) record was that he raised taxes by $730-million."
true
[50, 51, 52, 53]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
half-true
11
"1.3-million people in America make their living off eBay."
false
[54, 55, 56, 57]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
false
12
"101 utilities cut rates, credit GOP tax cuts."
true
[58, 59, 60, 61, 62, 63, 64, 65, 66, 67]
['misleading', 'misleading', 'misleading', 'unrelated', 'unrelated', 'misleading', 'unrelated', 'misleading', 'unrelated', 'misleading']
mostly-true
13
"122 vicious prisoners, released by the Obama Administration from Gitmo, have returned to the battlefield."
false
[68, 69, 70, 71]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-false
14
"14 percent of noncitizens are registered to vote."
false
[72, 73, 74, 75, 76]
['supporting', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
false
15
"14,000 Americans . . . lose their health insurance every single day."
true
[77, 78, 79, 80]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-true
16
"160 million people like their private insurance."
true
[81, 82, 83, 84]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
half-true
17
"2008 was the deadliest year of the war (in Afghanistan) for American forces."
true
[85, 86, 87, 88]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
true
18
"22 veterans take their own lives each day."
true
[89, 90, 91, 92, 93, 94, 95, 96, 97]
['misleading', 'supporting', 'misleading', 'supporting', 'supporting', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-true
19
"30,000 Arizonans may have no choice" on Affordable Care Act coverage.
true
[98, 99, 100, 101, 102, 103]
['unrelated', 'misleading', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-true
20
"400 bills have my name on them either as a sponsor or a co-sponsor. You don’t get that done unless you work with folks on the other side."
true
[104, 105, 106]
['unrelated', 'unrelated', 'unrelated']
half-true
21
"45 nations had already moved” to enforce travel restrictions with China “before the president moved.”
true
[107, 108, 109, 110]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
half-true
22
"45.6 million people watched (the State of the Union), the highest number in history. @FoxNews beat every other Network, for the first time ever, with 11.7 million people tuning in."
false
[111]
['supporting']
mostly-false
23
"58,000 non-citizens voted in Texas, with 95,000 non-citizens registered to vote."
false
[112, 113, 114, 115, 116, 117, 118, 119]
['supporting', 'unrelated', 'unrelated', 'supporting', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
false
24
"6,400 Ohioans ... lost manufacturing jobs in the month of September."
true
[120, 121, 122, 123]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-true
25
"70 percent of Florida is in drought today."
true
[124, 125, 126, 127]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-true
26
"8 years of the Clintons, major losses for Democrats across the nation."
true
[128, 129, 130, 131, 132, 133, 134, 135, 136, 137]
['misleading', 'unrelated', 'supporting', 'misleading', 'unrelated', 'misleading', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-true
27
"80% of the world’s population lives in countries where religious liberty is in significant danger or even completely outlawed."
true
[138, 139, 140, 141]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
half-true
28
"95 percent of the world’s customers are … beyond our shores."
true
[142, 143, 144, 145, 146, 147, 148, 149, 150, 151]
['unrelated', 'unrelated', 'unrelated', 'misleading', 'unrelated', 'unrelated', 'unrelated', 'misleading', 'unrelated', 'misleading']
half-true
29
"98 percent of American families" make "under $250,000 a year."
true
[152, 153, 154, 155, 156, 157]
['misleading', 'misleading', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
true
30
"98 percent of small businesses make less than $250,000" and would not see a tax increase under Barack Obama's plan.
true
[158, 159, 160, 161]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
true
31
"A construction worker who’s making $50 or $60 grand a year shouldn’t be paying higher tax rates than the guy who’s making $50 million a year. And that’s how it’s working right now."
true
[162, 163, 164, 165]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
half-true
32
"A day AFTER the election, Biden receives a dump of 143,379 votes at 3:42AM (in Wisconsin), when they learned he was losing badly. This is unbelievable!"
false
[166, 167, 168, 169, 170]
['misleading', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
pants-on-fire
33
"A few months ago when you were asked what's the biggest geopolitical threat facing America, you said Russia."
true
[171, 172, 173, 174, 175, 176, 177, 178, 179, 180]
['misleading', 'unrelated', 'misleading', 'unrelated', 'unrelated', 'unrelated', 'misleading', 'supporting', 'unrelated', 'unrelated']
half-true
34
"A few months ago, it hit where West Virginia is, on a per capita basis, one of the most successful (gross domestic product) states in our union."
false
[181, 182, 183, 184, 185]
['supporting', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-false
35
"A few months into office, (President Barack Obama) traveled around the globe to apologize for America."
false
[186, 187, 188, 189, 190, 191, 192, 193]
['supporting', 'supporting', 'unrelated', 'supporting', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
pants-on-fire
36
"A ham and cheese sandwich on one slice of bread is the responsibility of the USDA ... But a ham and cheese sandwich on two slices of bread is the responsibility of the Food and Drug Administration."
true
[194, 195, 196, 197, 198, 199, 200]
['unrelated', 'misleading', 'supporting', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
true
37
"A lot of people are talking about" the removal of a "very fair prosecutor" in Ukraine.
false
[201, 202, 203, 204]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
false
38
"A lot of people don't know that 50,000 Americans now make their living off eBay."
true
[205, 206, 207, 208, 209, 210, 211, 212]
['supporting', 'unrelated', 'unrelated', 'misleading', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
half-true
39
"A lot of the ideas in terms of the (health insurance) exchange, just being able to pool and improve the purchasing power of individuals in the insurance market, that originated from the Heritage Foundation."
true
[213, 214, 215, 216, 217]
['supporting', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-true
40
"A majority of Americans agree with" a ban on assault weapons. "And by the way, so did Ronald Reagan."
true
[218, 219, 220, 221, 222, 223, 224, 225, 226, 227]
['supporting', 'unrelated', 'unrelated', 'misleading', 'unrelated', 'unrelated', 'misleading', 'unrelated', 'unrelated', 'unrelated']
true
41
"A majority of President Trump's supporters and voters wanted (the United States) to stay in" the Paris Agreement.
false
[228, 229, 230, 231, 232, 233]
['supporting', 'supporting', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-false
42
"A million people … could get health insurance right away" if Texas expanded Medicaid under Obamacare.
true
[234, 235, 236, 237, 238, 239, 240]
['unrelated', 'unrelated', 'misleading', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
true
43
"A poll just came out on my inauguration speech. … People loved it. Loved and liked."
true
[241, 242, 243, 244]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-true
44
"A salesclerk at Hobby Lobby who needs contraception … is not going to get that service through her employer’s health care plan because her employer doesn’t think she should be using contraception."
false
[245, 246, 247, 248, 249]
['supporting', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-false
45
"A shocking 20 veterans are committing suicide each and every day, especially our older veterans."
true
[250, 251, 252, 253, 254, 255, 256, 257]
['misleading', 'supporting', 'supporting', 'supporting', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
true
46
"A single immigrant can bring in unlimited numbers of distant relatives."
false
[258, 259, 260, 261, 262, 263, 264, 265, 266, 267]
['supporting', 'unrelated', 'unrelated', 'unrelated', 'supporting', 'unrelated', 'unrelated', 'unrelated', 'supporting', 'unrelated']
mostly-false
47
"A small majority of Americans don’t think they like the Affordable Care Act, but a large majority of Americans don’t want to do away with the protections that are in the Affordable Care Act."
true
[268, 269, 270, 271]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-true
48
"A trade war is something very different (than curbing new trade agreements). We went down that road in the 1930s. It made the Great Depression longer and more painful."
true
[272, 273, 274, 275]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
true
49
"A violent felon can buy (a gun) over the Internet with no background check, no questions asked."
true
[276, 277, 278, 279, 280, 281, 282, 283, 284, 285]
['misleading', 'misleading', 'unrelated', 'misleading', 'misleading', 'unrelated', 'misleading', 'misleading', 'misleading', 'supporting']
mostly-true
50
"A year ago, in March 2007, I called for immediate action to address abuses in the subprime market, and I laid out detailed concrete proposals for how to do so."
true
[286, 287, 288, 289, 290, 291, 292, 293, 294]
['unrelated', 'misleading', 'unrelated', 'unrelated', 'misleading', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
true
51
"About 30 percent" of those released from Guantanamo "have re-entered the fight."
false
[295, 296, 297, 298]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-false
52
"About 40 percent of Hispanic homeowners have subprime mortgages."
true
[299, 300, 301, 302]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
true
53
"According to Pew Research, 99 percent of the people in Afghanistan support oppressive Sharia law."
true
[303, 304, 305, 306, 307, 308, 309]
['unrelated', 'unrelated', 'misleading', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-true
54
"According to data provided by the Department of Justice, the vast majority of individuals convicted of terrorism and terrorism-related offenses since 9/11 came here from outside of our country."
false
[310, 311, 312, 313, 314, 315]
['unrelated', 'supporting', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-false
55
"Across the country, $174 billion of the Recovery Act have been committed in its first 130 days."
true
[316, 317, 318, 319]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-true
56
"African American unemployment is the lowest ever recorded in our country. The Hispanic unemployment rate dropped a full point in the last year and is close to the lowest in recorded history. Dems did nothing for you but get your vote!"
true
[320, 321, 322, 323, 324, 325, 326, 327, 328]
['unrelated', 'misleading', 'unrelated', 'unrelated', 'misleading', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-true
57
"African-American children are 500 percent more likely to die from asthma than white kids."
true
[329, 330, 331, 332]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
true
58
"African-Americans are more likely to be arrested by police and sentenced to longer prison terms for doing the same thing that whites do."
true
[333, 334, 335, 336, 337, 338, 339, 340, 341, 342]
['supporting', 'misleading', 'supporting', 'supporting', 'unrelated', 'supporting', 'supporting', 'misleading', 'supporting', 'misleading']
true
59
"After 30 years of inaction, we raised fuel standards so that by the middle of the next decade, cars and trucks will go twice as far on a gallon of gas."
true
[343, 344, 345, 346, 347, 348, 349, 350, 351, 352]
['unrelated', 'unrelated', 'unrelated', 'supporting', 'misleading', 'misleading', 'misleading', 'unrelated', 'unrelated', 'unrelated']
half-true
60
"After Roe was dismantled, extremists evoked a law from 1849 to stop abortion” in Wisconsin.
false
[353, 354, 355, 356]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
false
61
"After a decade of decline, this country created over half a million manufacturing jobs in the last two and a half years."
true
[357, 358, 359, 360, 361]
['misleading', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
true
62
"After shedding jobs for more than 10 years, our manufacturers have added about 500,000 jobs over the past three."
true
[362, 363, 364, 365, 366, 367, 368, 369, 370, 371]
['misleading', 'unrelated', 'misleading', 'unrelated', 'unrelated', 'misleading', 'misleading', 'misleading', 'unrelated', 'unrelated']
true
63
"After the last administration nearly killed the US auto industry, I saved the U.S. auto industry ... We brought you a lot of car plants… (and Japan) “announced five car companies are coming to Michigan.”
false
[372, 373]
['unrelated', 'unrelated']
false
64
"After years of wage stagnation, we are finally seeing rising wages."
false
[374, 375, 376, 377, 378, 379, 380, 381, 382, 383]
['supporting', 'unrelated', 'supporting', 'supporting', 'unrelated', 'supporting', 'supporting', 'misleading', 'supporting', 'unrelated']
mostly-false
65
"Again today (Ahmadinejad) made light of 9/11, and said that he's not even sure it happened and that people actually died."
true
[384, 385, 386, 387]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
true
66
"All across Ohio, steel mills are reopening."
true
[388, 389, 390, 391, 392, 393, 394, 395, 396, 397]
['misleading', 'misleading', 'misleading', 'unrelated', 'unrelated', 'unrelated', 'misleading', 'unrelated', 'unrelated', 'unrelated']
half-true
67
"All legal scholars, both sides, wanted, and in fact demanded” that Roe v. Wade be overturned.
false
[398, 399, 400, 401, 402, 403, 404, 405, 406, 407]
['unrelated', 'supporting', 'supporting', 'unrelated', 'supporting', 'unrelated', 'supporting', 'unrelated', 'unrelated', 'unrelated']
false
68
"All of the records, as far as I know, about what we did with health care, those are already available."
false
[408, 409, 410, 411, 412, 413, 414, 415, 416, 417]
['unrelated', 'supporting', 'supporting', 'supporting', 'misleading', 'supporting', 'unrelated', 'supporting', 'supporting', 'supporting']
mostly-false
69
"All pipelines that are coming into this country from now on has to be American steel."
false
[418, 419, 420, 421, 422]
['supporting', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
false
70
"Almost every country on Earth sees America as stronger and more respected today than they did eight years ago when I took office."
true
[423, 424, 425, 426]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-true
71
"Almost half a million people are still eligible (for DACA), particularly in the Asian-American and Pacific Islander community."
true
[427, 428, 429, 430]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
true
72
"Already we've identified $2 trillion in deficit reductions over the next decade."
false
[431, 432, 433, 434, 435, 436, 437]
['unrelated', 'supporting', 'supporting', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-false
73
"Already, the Affordable Care Act is helping to slow the growth of (Medicare’s) health care costs."
true
[438, 439, 440, 441, 442, 443, 444, 445]
['unrelated', 'misleading', 'unrelated', 'misleading', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
half-true
74
"America has lost nearly one third of its manufacturing jobs since 1997, following the enactment of disastrous trade deals supported by Bill and Hillary Clinton."
true
[446, 447, 448, 449, 450, 451, 452, 453, 454]
['misleading', 'unrelated', 'unrelated', 'unrelated', 'misleading', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
half-true
75
"America is No. 1 in wind power."
true
[455, 456, 457, 458]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-true
76
"America is No. 1 one in oil and gas."
true
[459, 460, 461, 462, 463, 464, 465, 466, 467, 468]
['misleading', 'misleading', 'unrelated', 'supporting', 'misleading', 'supporting', 'unrelated', 'unrelated', 'unrelated', 'misleading']
true
77
"America is a nation of builders. We built the Empire State Building in just one year. Isn’t it a disgrace that it can now take 10 years just to get a minor permit approved for the building of a simple road?"
true
[469, 470, 471, 472, 473, 474, 475, 476, 477, 478]
['unrelated', 'supporting', 'supporting', 'unrelated', 'unrelated', 'supporting', 'misleading', 'supporting', 'unrelated', 'unrelated']
half-true
78
"America is one of the highest-taxed nations in the world."
false
[479, 480, 481, 482, 483, 484]
['supporting', 'supporting', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-false
79
"American homeownership rate in Q2 2016 was 62.9% - lowest rate in 51yrs."
true
[485, 486, 487, 488]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-true
80
"Americans haven’t had a raise in 15 years."
true
[489, 490, 491, 492, 493, 494, 495, 496, 497, 498, 499, 500, 501, 502, 503, 504, 505, 506, 507, 508]
['misleading', 'unrelated', 'unrelated', 'supporting', 'unrelated', 'supporting', 'unrelated', 'supporting', 'misleading', 'supporting', 'misleading', 'unrelated', 'misleading', 'supporting', 'unrelated', 'supporting', 'unrelated', 'supporting', 'misleading', 'supporting']
mostly-true
81
"Americans spend 100-million hours a year filling out these forms (FAFSA). That is the equivalent of 55,500 full-time jobs."
true
[509, 510, 511, 512]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
half-true
82
"Americans who get their insurance through the workplace, cost savings could be as much as $3,000 less per employee than if we do nothing."
true
[513, 514, 515, 516, 517, 518]
['unrelated', 'misleading', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
half-true
83
"An amendment, pushed by Lyndon Johnson many years ago, threatens religious institutions with a loss of their tax-exempt status if they openly advocate their political views."
true
[519, 520, 521, 522, 523, 524, 525, 526, 527]
['supporting', 'supporting', 'supporting', 'misleading', 'supporting', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
true
84
"An analysis showed that Bernie Sanders would have won the Democratic nomination if it were not for the Super Delegates."
false
[528, 529, 530, 531, 532, 533, 534, 535, 536]
['unrelated', 'unrelated', 'supporting', 'unrelated', 'supporting', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
false
85
"And 100 percent, John, of your ads . . . 100 percent of them have been negative."
false
[537, 538, 539, 540]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
pants-on-fire
86
"And it's going to be difficult for us to do this as long as we're spending $275 million a day on a war that should have never been authorized..."
true
[541, 542, 543, 544]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-true
87
"Another one he should veto is the SCHIP program, which he should say 'Take the C out of, because now it's for everybody, like every other entitlement program.'"
false
[545, 546, 547, 548, 549, 550, 551, 552]
['unrelated', 'unrelated', 'unrelated', 'unrelated', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
false
88
"Anybody can buy any weapon, any time without much, if any, regulation. They buy it over the internet. They can buy machine guns."
false
[553, 554, 555, 556, 557, 558, 559, 560, 561]
['supporting', 'supporting', 'supporting', 'supporting', 'supporting', 'unrelated', 'supporting', 'supporting', 'supporting']
mostly-false
89
"Apple has just announced it plans to invest a total of $350 billion in America, and hire another 20,000 workers."
true
[562, 563, 564, 565, 566, 567, 568]
['misleading', 'unrelated', 'misleading', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
half-true
90
"Arizona schools rank 45th in the nation, dead last in funding per student."
true
[569, 570, 571, 572, 573, 574]
['misleading', 'misleading', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-true
91
"As Bernie Sanders said, Hillary Clinton ‘voted for virtually every trade agreement that has cost the workers of this country millions, millions of jobs.’ "
true
[575, 576, 577, 578, 579, 580, 581, 582, 583, 584]
['supporting', 'supporting', 'unrelated', 'unrelated', 'misleading', 'unrelated', 'unrelated', 'unrelated', 'misleading', 'unrelated']
half-true
92
"As Secretary of State, Hillary Clinton laundered money to Bill Clinton through Laureate Education, while Bill Clinton was an honorary chairman of the group."
false
[585]
['unrelated']
false
93
"As a legislator, (Marco Rubio) flipped on a key vote after making a quick $200,000 from selling the house to the mother of the bill's lobbyist."
false
[586, 587]
['unrelated', 'unrelated']
mostly-false
94
"As a percentage of gross domestic product, we are reducing nondefense discretionary spending to its lowest level since the '60s, lower than it was under Reagan, lower than it was under Clinton, lower than it was under Bush, or both Bushes."
false
[588, 589, 590, 591]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-false
95
"As governor, in the face of partisan attacks, (Charlie Crist) had the courage to save jobs and lead his state into economic recovery."
true
[592, 593, 594, 595]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
half-true
96
"As has been noted by many observers, including Bill Clinton's former secretary of labor, my plan does more than anybody to reduce costs."
false
[596, 597, 598, 599]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
false
97
"As long as I have served . . . This is the first time every single solitary decision has required 60 senators."
false
[600, 601, 602, 603, 604, 605, 606, 607, 608, 609]
['supporting', 'unrelated', 'misleading', 'unrelated', 'supporting', 'unrelated', 'unrelated', 'supporting', 'unrelated', 'unrelated']
false
98
"As many as a quarter of our students aren’t even finishing high school."
true
[610, 611, 612, 613]
['unrelated', 'unrelated', 'unrelated', 'unrelated']
mostly-true
99
"As president, Barack Obama has never visited Israel."
true
[614, 615, 616, 617, 618, 619]
['unrelated', 'misleading', 'unrelated', 'unrelated', 'unrelated', 'unrelated']
true
100
"As usual, Hillary & the Dems are trying to rig the debates so 2 are up against major NFL games."
false
[620, 621, 622, 623, 624, 625, 626, 627, 628]
['unrelated', 'unrelated', 'supporting', 'unrelated', 'unrelated', 'supporting', 'unrelated', 'unrelated', 'supporting']
pants-on-fire
End of preview.

Dataset Card for RAGuard

Dataset Details

RAGuard is a fact-checking dataset designed to evaluate the robustness of RAG systems against misleading retrievals. It consists of 2,648 political claims made by U.S. presidential candidates (2000–2024), each labeled as either true or false, and a knowledge base comprising 16,331 documents. Each claim is linked to a set of associated documents, categorized as supporting, misleading, or unrelated

Dataset Description

claims.csv contains the following fields for each claim, scraped from PolitiFact.

  1. Claim ID
  2. Claim: Full text of the claim
  3. Verdict: Binary fact-checking verdict, True or False
  4. Document IDs: List IDs of documents corresponding to this claim
  5. Document Labels: List of labels for the associated documents, either supporting, misleading, or unrelated, generated by an LLM.

documents.csv contains the following fields for each document in our knowledge base, scraped from Reddit.

  1. Document ID
  2. Title: Reddit post title
  3. Full Text: Content of the document
  4. Claim ID: ID of the corresponding claim
  5. Document Label: Label for the document's label to the claim, either supporting, misleading, or unrelated
  6. Link: URL to the original document

Dataset Source and Usage

  • See Section 2.4 of our paper for supported tasks

Citation

If you use this dataset, please cite our paper:

@misc{zeng2025worsezeroshotfactcheckingdataset,
      title={Worse than Zero-shot? A Fact-Checking Dataset for Evaluating the Robustness of RAG Against Misleading Retrievals}, 
      author={Linda Zeng and Rithwik Gupta and Divij Motwani and Diji Yang and Yi Zhang},
      year={2025},
      eprint={2502.16101},
      archivePrefix={arXiv},
      primaryClass={cs.AI},
      url={https://arxiv.org/abs/2502.16101}, 
}

Disclaimer

This dataset has been compiled from publicly available sources on the Internet. It may contain discussions on sensitive political topics, including viewpoints that some individuals may find controversial or offensive. The inclusion of any content does not imply endorsement of any views expressed. Users are advised to exercise discretion and ensure compliance with applicable ethical guidelines and legal frameworks when using this dataset.

Downloads last month
60