Datasets:

BeIR
/

nq

_id stringlengths 4 10	title stringlengths 1 134	text stringlengths 0 17k
doc0	Minority interest	In accounting, minority interest (or non-controlling interest) is the portion of a subsidiary corporation's stock that is not owned by the parent corporation. The magnitude of the minority interest in the subsidiary company is generally less than 50% of outstanding shares, or the corporation would generally cease to be a subsidiary of the parent.[1]
doc1	Minority interest	It is, however, possible (such as through special voting rights) for a controlling interest requiring consolidation to be achieved without exceeding 50% ownership, depending on the accounting standards being employed. Minority interest belongs to other investors and is reported on the consolidated balance sheet of the owning company to reflect the claim on assets belonging to other, non-controlling shareholders. Also, minority interest is reported on the consolidated income statement as a share of profit belonging to minority shareholders.
doc2	Minority interest	The reporting of 'minority interest' is a consequence of the requirement by accounting standards to 'fully' consolidate partly owned subsidiaries. Full consolidation, as opposed to partial consolidation, results in financial statements that are constructed as if the parent corporation fully owns these partly owned subsidiaries; except for two line items that reflect partial ownership of subsidiaries: net income to common shareholders and common equity. The two minority interest line items are the net difference between what would have been the common equity and net income to common, if all subsidiaries were fully owned, and the actual ownership of the group. All the other line items in the financial statements assume a fictitious 100% ownership.
doc3	Minority interest	Some investors have expressed concern that the minority interest line items cause significant uncertainty for the assessment of value, leverage and liquidity.[2] A key concern of investors is that they cannot be sure what part of the reported cash position is owned by a 100% subsidiary and what part is owned by a 51% subsidiary.
doc4	Minority interest	Minority interest is an integral part of the enterprise value of a company. The converse concept is an associate company.
doc5	Minority interest	Under the International Financial Reporting Standards, the non-controlling interest is reported in accordance with IFRS 5 and is shown at the very bottom of the Equity section on the consolidated balance sheet and subsequently on the statement of changes in equity. Under US GAAP minority interest can be reported either in the liabilities section, the equity section, or the mezzanine section of the balance sheet. The Mezzanine section is located between liabilities and equity. FASB FAS 160 and FAS 141r significantly alter the way a parent company accounts for non-controlling interest (NCI) in a subsidiary. It is no longer acceptable to report minority interest in the mezzanine section of the balance sheet.
doc6	Chicago Fire (season 4)	The fourth season of Chicago Fire, an American drama television series with executive producer Dick Wolf, and producers Derek Haas, Michael Brandt, and Matt Olmstead, was ordered on February 5, 2015, by NBC,[1] and premiered on October 13, 2015 and concluded on May 17, 2016.[2] The season contained 23 episodes.[3]
doc7	Chicago Fire (season 4)	The show follows the lives of the firefighters and paramedics working at the Chicago Fire Department at the firehouse of Engine 51, Truck 81, Squad 3, Ambulance 61 and Battalion 25.
doc8	Chicago Fire (season 4)	Tensions only get worse between Patterson and Severide when Severide makes a call at a job without his consent causing Patterson to question Severide's being at 51. Meanwhile, Borelli and Chili take their romance to the next level, Cruz receives a visit from one of his brother Leon's old gang members. Also, Mouch receives tickets from one of the members of the rock band Rush and Boden and his wife Donna welcome a new neighbor who turns out to be overly friendly.
doc9	Chicago Fire (season 4)	Hermann is rushed to Chicago Med after being stabbed at Molly's. After losing a lot a blood, it is determined he needs emergency surgery. Feeling guilty about Hermann's present state, Cruz searches for Freddy to turn him in. Severide is reinstated as Lieutenant while Borelli grows more concerned about Chili's erratic behavior. Mouch considers finally proposing to Platt.
doc10	Love Will Keep Us Alive	"Love Will Keep Us Alive" is a song written by Jim Capaldi, Paul Carrack, and Peter Vale, and produced by the Eagles, Elliot Scheiner, and Rob Jacobs. It was first performed by the Eagles in 1994, during their "Hell Freezes Over" reunion tour, with lead vocals by bassist Timothy B. Schmit.
doc11	Love Will Keep Us Alive	Although the song was never formally released as a single in the US, and thus was not eligible to appear on the US Billboard Hot 100 under the rules then in place, it spent three weeks at number 1 on the Billboard adult contemporary chart in early 1995[1] and reached number 22 on Billboard's Hot 100 Airplay chart. In the United Kingdom, "Love Will Keep Us Alive" was issued as a single and peaked at number 52 on the UK Singles Chart.[2]
doc12	Love Will Keep Us Alive	Aside from being on the album Hell Freezes Over, the song appears on the Eagles' box set, Selected Works 1972-1999 and the 2003 compilation album, The Very Best Of.
doc13	Love Will Keep Us Alive	Paul Carrack recorded the song for his 1996 album, Blue Views; it also featured on his 2006 compilation album, Greatest Hits - The Story So Far.
doc14	Love Will Keep Us Alive	In 2011, Paul Carrack and Timothy B. Schmit recorded the song in London with the Royal Philharmonic Orchestra, and released it in the UK on the Carrack label.
doc15	Love Will Keep Us Alive	According to the liner notes that accompanied their 2003 greatest hits CD, this song was written when Carrack, Capaldi, and Schmit were planning to form a band with Don Felder and Max Carl during the late eighties or early nineties.[3] The band had the working name of Malibu Men's Choir.[4] This never materialized, so Schmit proposed the song for the Eagles' reunion album. According to Felder, they sent demo tapes to the Eagles manager, Irving Azoff, who rejected it as not good enough.[5] Felder thought it ironic that the Eagles would later record one of those rejected songs.[6]
doc16	Love Will Keep Us Alive	"Love Will Keep Us Alive" was also recorded by Capaldi and Dave Mason on their 40,000 Headman tour and live album, and by Carrack (duet with Lindsay Dracass) on his 2007 album Old, New, Borrowed and Blue. It was also covered by Canadian-Australian singer Wendy Matthews in 1995 as "Love Will Keep Me Alive" as a track from the album The Witness Tree (1994).
doc17	Patrick Brown (politician)	Patrick Walter Brown MPP (born May 26, 1978) is a Canadian politician who is the leader of the Progressive Conservative Party of Ontario and Ontario's Leader of the Official Opposition. Brown was a federal Conservative member of the House of Commons of Canada from 2006-15 representing the riding of Barrie.
doc18	Patrick Brown (politician)	In May 2015, Brown was elected leader of the Ontario PC Party, and stepped down as MP. He was elected Member of Provincial Parliament (MPP) for Simcoe North in a provincial by-election on September 3, 2015. Before being elected to federal office, Brown worked as a lawyer in Barrie.[1]
doc19	Patrick Brown (politician)	Brown was born in Toronto of Irish and Italian descent, and raised in the Roman Catholic faith. His father, Edmond Brown, a lawyer and former New Democratic Party candidate, was raised in England and Ireland before moving to Canada, and his mother, Judy (nÃ©e Tascona) Brown, is of partial Italian descent.[2]
doc20	Patrick Brown (politician)	Brown is the nephew of Joe Tascona, a Barrie Progressive Conservative MPP in the Mike Harris government.[3] He graduated from St. Michael's College School, a private Catholic school in Toronto.[4] He studied political science at the University of Toronto, and graduated with a law degree from the University of Windsor. During his second year at law school, he was one of 10 recipients of the As Prime Minister Award. He worked for Magna International in their legal department for four years.[citation needed]
doc21	Patrick Brown (politician)	Brown served two terms as President of the Progressive Conservative Youth Federation (PCYF),[5] a position he held from 1998 to 2002. He also served on the executive of the Progressive Conservative Party of Ontario, as a Vice President. As PCYF President, Brown was one of the early supporters of a united right and was criticized for his decision to support a united right from party leader Joe Clark and Member of Parliament Scott Brison. Brown was later re-elected as PCYF president with 81 percent of the vote against Jonathan Frate of Manitoba.
doc22	Patrick Brown (politician)	Brown was the Deputy Chairman of the International Young Democrat Union (IYDU).[when?] He has also represented Canada on a number of international assistance projects hosted by the IYDU.
doc23	Patrick Brown (politician)	Brown identifies himself as a "pragmatic conservative"[6] and since becoming leader he has tried to move the Ontario PC Party in a socially liberal and fiscally conservative direction.[7] At his first Ontario PC Convention as the new leader, Brown confirmed his belief in man-made climate change and announced his support for a revenue-neutral price on carbon.[8] Brown was also the first Ontario PC Leader to march in the Toronto Pride Parade.[9] Among his political mentors, Brown lists former Ontario Premier Bill Davis and former Canadian Prime Minister Brian Mulroney.[10]
doc24	Patrick Brown (politician)	Much of Brown's time at Queen's Park has been spent criticizing and debating the government's energy policies. He has promised to dismantle the Green Energy Act, rein in executive salaries at Hydro One, and place a moratorium on the signing of new energy contracts.[11][12][13]
doc25	Patrick Brown (politician)	Patrick Brown's first Private Member's Bill in the Ontario Legislature, Bill 151 the Estate Administration Tax Abolition Act, was an attempt to eliminate Ontario's estate administration, or probate tax.[14] His bill was voted down at Second Reading by the Liberal Government's majority.
doc26	Patrick Brown (politician)	Brown has been noted for his close relationship with many of Ontario's diverse ethnic communities.[15] He has spoken in the Legislature in support of a motion condemning Islamophobia,[16][17] and was one of the first Canadian politicians to refer to the Tamil Genocide.[18] Brown has a personal relationship with Indian Prime Minister Narendra Modi, who refers to him as "Patrick Bhai" meaning brother and named him an honorary citizen of Gujarat.[19][20]
doc27	Patrick Brown (politician)	His critics have called him "policy-lite" since he made no policy statements during the Progressive Conservative leadership campaign.[21] Since winning the leadership race, he has focused his plan on four main issues which he suggests will lead to a more prosperous province: less red tape, improved transportation corridors, affordable energy, and addressing Ontario's growing skills gap.[22]
doc28	Patrick Brown (politician)	Brown's shift of the party to the political centre stands in contrast to his time as an MP where Brown had a socially conservative voting record.[23][24][25] As an MP, Brown voted to re-open the same-sex marriage and abortion debates, as well as voted against legalizing euthanasia and including gender expression in the Human Rights Act. He said those votes were to represent his constituents.[26][27] Brown said that he doesn't intend to revisit any of these issues in the provincial Legislature.[28][29]
doc29	Patrick Brown (politician)	Brown was elected to the Barrie City Council in 2000 at age 22, and was re-elected in 2003.[5]
doc30	Patrick Brown (politician)	Brown served on various Committees, including the Budget Committee. Brown's primary focus while on council was health care, despite it being a provincial responsibility. In response to a shortage of doctors, Brown founded the Physician Recruitment Task Force with the Royal Victoria Hospital to help attract more doctors to Barrie.[30]
doc31	Patrick Brown (politician)	In the 2004 federal election, Brown ran as the Conservative Party candidate in the riding of Barrie. He lost to incumbent Aileen Carroll by 1,295 votes.[31] Brown ran again in 2006, this time defeating Carroll by 1,523 votes.[32] He was re-elected in the 2008 election by 15,295 votes over Liberal candidate Rick Jones.[33]
doc32	Patrick Brown (politician)	In November 2010, the Canadian Taxpayers Federation expressed concern about how Patrick Brown used his Canadian House of Commons account. He sent flyers to his riding which included a letter of support and a flyer from Barrie City Councillor Michael Prowse. Brown used his House of Commons account to pay for the mailing because Michael Prowse could not afford to send the flyer out himself.[34]
doc33	Patrick Brown (politician)	In the 2011 election, Brown was elected to his third term in office.[35]
doc34	Patrick Brown (politician)	On September 28, 2014, he announced his intention to run in the 2015 Ontario party leadership election. He registered as a leadership candidate on November 20, 2014. He said that, unlike the other candidates, he was not involved in the four consecutive losses that have kept the Ontario PCs out of power since 2003.[36] At the time of his jump to provincial politics, he chaired the Conservative Party of Canada's Greater Toronto Area caucus and the Canada-India Parliamentary Association.[5]
doc35	Patrick Brown (politician)	In September 2014, Brown announced his intention to run in the contest to replace PC Party Leader, Tim Hudak. From the outset of his campaign, Brown positioned himself as an outsider, challenging the leadership of the PC Party, which had been defeated in the last four provincial elections. In the most recent election campaign, in 2014, the party election platform included a commitment to cut 100,000 public service jobs over 4 years through attrition.[37] As the only one of the original five leadership candidates who was not a member of the Ontario legislature, Brown was not involved in the promise, which he considered ill-advised,[36][38] [39] Brown's rivals attempted to use this same lack of previous involvement in provincial politics as an argument against his leadership bid.[40][41]
doc36	Patrick Brown (politician)	In March, Brown emerged as the front-runner in the leadership election, having sold over 40,000 of the 70,000 memberships in the party.[42][43][44][45] During the campaign, Brown was successful in bringing many new members to the party, many of whom came from ethnic communities.[46] The past four leadership contests had been won by those who sold the most memberships.[47]
doc37	Patrick Brown (politician)	Brown was endorsed by the Campaign Life Coalition and the Ontario Landowners Association.[48][49] During Brown's leadership bid both special interest groups actively supported him by selling Ontario PC Party memberships amongst their members.[50][51]
doc38	Patrick Brown (politician)	Brown was criticized by his main rival, Christine Elliott, for not resigning his federal seat during the leadership campaign.[52] Brown was absent from the House of Commons for some votes during the leadership campaign, attending 56% of the votes from September to December in 2014. However, his overall attendance for votes in 2014 was 83%. [53] A spokesperson for Prime Minister Stephen Harper confirmed that members are not expected to step down but are expected to "continue to fulfill their parliamentary responsibilities, including membership on committees and attendance at votes."[54]
doc39	Patrick Brown (politician)	The campaign started with five candidates including Vic Fedeli, Lisa MacLeod, and Monte McNaughton. All three withdrew in early 2015 citing membership recruitment or financial reasons. On May 9, 2015, Brown was elected leader, defeating his only remaining opponent, Christine Elliott, winning with 61.8% of the membership vote.[55][56]
doc40	Patrick Brown (politician)	Brown, who resigned his seat in the House of Commons on May 13, 2015, days after winning the provincial leadership, led the Progressive Conservative party from outside the legislature during most of the summer.[57] On July 22, 2015, Garfield Dunlop agreed to step down as MPP for Simcoe North on August 1 in order to open up a seat for Brown. A provincial by-election, called for September 3, 2015, was won by Brown.[58][59][60]
doc41	Patrick Brown (politician)	Under his leadership, the Ontario PC Party has won five by-elections, including two seats which had been previously held by the governing Liberals - Sault Ste. Marie and Scarborough-Rouge River.[61]
doc42	Fishin' in the Dark	"Fishin' in the Dark" is a song written by Wendy Waldman and Jim Photoglo and recorded by American country music group The Nitty Gritty Dirt Band. It was released in June 1987 as the second single from their album Hold On.[1] It reached number-one on the U.S. and Canadian country charts. It was the band's third number-one single on the U.S. country music charts and the second in Canada. After it became available for download, it has sold over a million digital copies by 2015.[2] It was certified Platinum by the RIAA on September 12, 2014.[3]
doc43	Fishin' in the Dark	The premise of the song is a couple contemplating a late-night fishing expedition. Specifically, the adventurers plan to make their way to an undisclosed river and chart constellations during an evening in which a full moon is present. Furthermore, the tentative date for this excursion is set in the late spring to early summer.
doc44	Fishin' in the Dark	The music video was directed by Bill Young and features the band playing in front of a live audience.
doc45	Fishin' in the Dark	*sales figures based on certification alone ^shipments figures based on certification alone
doc46	Three Rings	In Tolkien's mythology, the Three Rings are magical artifacts forged by the Elves of Eregion. After the One Ring, they are the most powerful of the twenty Rings of Power.[1]
doc47	Three Rings	The Three Rings were made by Celebrimbor after Sauron, in the guise of Annatar, had left Eregion. These were free of Sauron's influence, as he did not have a hand in their making. However, they were still forged by Celebrimbor with the arts taught to him by Sauron and thus were still bound to the One Ring. Upon perceiving Sauron's intent, the Elves hid the three from him. They were carried out of Middle-earth at the end of the Third Age, after the destruction of the One Ring.
doc48	Three Rings	The first ring, Narya, was adorned with a red gemstone, perhaps a ruby. It is seen in the final chapter of The Lord of the Rings, along with the other two Elven rings. But unlike them, it is not said what metal Narya was made of.
doc49	Three Rings	The name is derived from the Quenya nÃ¡r meaning fire. It was also called Narya the Great, Ring of Fire, Red Ring, and The Kindler.
doc50	Three Rings	According to Unfinished Tales, at the start of the War of the Elves and Sauron, Celebrimbor gave Narya together with the Ring Vilya to Gil-galad, High King of the Noldor. Gil-galad entrusted Narya to his lieutenant Círdan, Lord of the Havens of Mithlond, who kept it after Gil-galad's death. According to The Lord of the Rings, Gil-galad received only Vilya, while Círdan received Narya from the very beginning along with Galadriel receiving Nenya from the start.
doc51	Three Rings	In the Third Age, CÃrdan, recognizing Gandalf's true nature as one of the Maiar from Valinor, gave him the ring to aid him in his labours. It is described as having the power to inspire others to resist tyranny, domination, and despair (in other words, evoking hope in others around the wielder), as well as giving resistance to the weariness of time.
doc52	Three Rings	The second ring, Nenya, was made of mithril and adorned with a "white stone", presumably a diamond.[2] The name is derived from the Quenya nÃ©n meaning water. It is also called Ring of Adamant, Ring of Water and the White Ring.
doc53	Three Rings	The ring was wielded by Galadriel of Lothlórien, and possessed a radiance that matched that of the stars; while Frodo Baggins could see it by virtue of being a Ring-bearer, Samwise Gamgee tells Galadriel he only "saw a star through your fingers". (This appears in many editions as "finger"—which sounds more magical, since it suggests that her finger has somehow become transparent—but The Treason of Isengard, ch. 13, note 34, mentions it as an error.)
doc54	Three Rings	Nenya's power gave preservation, protection, and possibly concealment from evil because "there is a secret power here that holds evil from the land". However, the fact that Orcs from Moria entered L贸rien after The Fellowship of the Ring and L贸rien itself had suffered previous attacks from Sauron's Orcs sent from Dol Guldur suggests the power of the ring did not constitute military prowess. It was said that, protected as it was by Nenya, Lothl贸rien would not have fallen unless Sauron had personally come to attack it. Galadriel used these powers to create and sustain Lothl贸rien, but it also increased in her the longing for the Sea and her desire to return to the Undying Lands.
doc55	Three Rings	With the ring gone, the magic and beauty of Lórien also faded along with the extraordinary mallorn trees (save the one that Samwise Gamgee grew in Hobbiton) and it was gradually depopulated, until by the time Arwen came there to die in F.A. 121 it was deserted and in ruin.
doc56	Three Rings	The third ring, Vilya, was made of gold and adorned with a "great blue stone", probably a sapphire. The name is derived from the Quenya vilya meaning air. It is also called, Ring of Air, Ring of Firmament, or Blue Ring.
doc57	Three Rings	It is generally considered that Vilya was the mightiest of these three bands (as mentioned in the ending chapter in The Return of the King). The exact power of Vilya is not mentioned. However, it is reasonable to speculate that it also possesses the power to heal and to preserve (it is mentioned in The Silmarillion that Celebrimbor had forged the Three in order to heal and to preserve, rather than to enhance the strengths of each individual bearer as the Seven, Nine, and the lesser rings did). Its power of healing may be particularly strong, as Elrond seems to be the greatest healer in Middle-Earth at the time of the Quest.[3] There is some speculation that the ring controlled minor elements, considering the event where Elrond had summoned a torrent of water as the Nazgûl attempted to capture Frodo and the One Ring.
doc58	Three Rings	When Sauron laid waste to Eregion, Vilya was sent to the Elven-king Gil-galad far away in Lindon, where it was later given to Elrond, who bore it through the later years of the Second Age and all of the Third. As Gil-galad was the High King of the Noldor elves at the time of the rings' distribution it was thought that he was best fit to care for the most powerful of the three Elven rings.
doc59	Panning (audio)	Panning is the distribution of a sound signal (either monaural or stereophonic pairs) into a new stereo or multi-channel sound field determined by a pan control setting. A typical physical recording console has a pan control for each incoming source channel. A pan control or pan pot (short for "panoramic potentiometer") is an analog knob or slider with a position indicator which can range continuously from the 8 o'clock when fully left to the 4 o'clock position fully right. Audio mixing software replaces pan pots with on-screen virtual knobs or sliders which function identically to the physical counterparts.
doc60	Panning (audio)	A pan pot has an internal architecture which determines how much of a source signal is sent to the left and right buses. "Pan pots split audio signals into left and right channels, each equipped with its own discrete gain (volume) control."[1] This signal distribution is often called a taper or law.
doc61	Panning (audio)	When centered (at 12 o'clock), the law can be designed to send −3, −4.5 or −6 decibels (dB) equally to each bus. "Signal passes through both the channels at an equal volume while the pan pot points directly north."[1] If the two output buses are later recombined into a monaural signal, then a pan law of -6 dB is desirable. If the two output buses are to remain stereo then a law of -3 dB is desirable. A law of −4.5 dB at center is a compromise between the two. A pan control fully rotated to one side results in the source being sent at full strength (0 dB) to one bus (either the left or right channel) and zero strength (−∞ dB) to the other. Regardless of the pan setting, the overall sound power level remains (or appears to remain) constant.[2] Because of the phantom center phenomenon, sound panned to the center position is perceived as coming from between the left and right speakers, but not in the center unless listened to with headphones, because of head-related transfer function HRTF.[citation needed]
doc62	Panning (audio)	Panning in audio borrows its name from panning action in moving image technology. An audio pan pot can be used in a mix to create the impression that a source is moving from one side of the soundstage to the other, although ideally there would be timing (including phase and Doppler effects), filtering and reverberation differences present for a more complete picture of apparent movement within a defined space. Simple analog pan controls only change relative level; they don't add reverb to replace direct signal, phase changes, modify the spectrum, or change delay timing. "Tracks thus seem to move in the direction that [one] point[s] the pan pots on a mixer, even though [one] actually attenuate[s] those tracks on the opposite side of the horizontal plane."[3]
doc63	Panning (audio)	Panning can also be used in an audio mixer to reduce or reverse the stereo width of a stereo signal. For instance, the left and right channels of a stereo source can be panned straight up, that is sent equally to both the left output and the right output of the mixer, creating a dual mono signal.[citation needed]
doc64	Panning (audio)	An early panning process was used in the development of Fantasound, an early pioneering stereophonic sound reproduction system for Fantasia (1940).
doc65	Panning (audio)	Before pan pots were available, "a three-way switch was used to assign the track to the left output, right output, or both (the center)".[4] Ubiquitous in the Billboard charts throughout the middle and late 1960s, clear examples include the Beatles's "Strawberry Fields Forever" and Jimi Hendrix's "Purple Haze".[5] In the Beatles's "A Day In The Life" Lennon's vocals are switched to the extreme right on the first two strophes, on the third strophe they are switched center then extreme left, and switched left on the final strophe while during the bridge McCartney's vocals are switched extreme right.[6][7]
doc66	Disparate impact	Disparate impact in United States labor law refers to practices in employment, housing, and other areas that adversely affect one group of people of a protected characteristic more than another, even though rules applied by employers or landlords are formally neutral. Although the protected classes vary by statute, most federal civil rights laws protect based on race, color, religion, national origin, and sex as protected traits, and some laws include disability status and other traits as well.
doc67	Disparate impact	A violation of Title VII of the 1964 Civil Rights Act may be proven by showing that an employment practice or policy has a disproportionately adverse effect on members of the protected class as compared with non-members of the protected class.[1] Therefore, the disparate impact theory under Title VII prohibits employers "from using a facially neutral employment practice that has an unjustified adverse impact on members of a protected class. A facially neutral employment practice is one that does not appear to be discriminatory on its face; rather it is one that is discriminatory in its application or effect."[2] Where a disparate impact is shown, the plaintiff can prevail without the necessity of showing intentional discrimination unless the defendant employer demonstrates that the practice or policy in question has a demonstrable relationship to the requirements of the job in question.[3] This is the "business necessity" defense.[1]
doc68	Disparate impact	In addition to Title VII, other federal laws also have disparate impact provisions, including the Age Discrimination in Employment Act of 1967.[4] Some civil rights laws, such as Title VI of the Civil Rights Act of 1964, do not contain disparate impact provisions creating a private right of action,[5] although the federal government may still pursue disparate impact claims under these laws.[6] The U.S. Supreme Court has held that the Fair Housing Act of 1968 creates a cause of action for disparate impact.[7] Disparate impact contrasts with disparate treatment. A disparate impact is unintentional, whereas a disparate treatment is an intentional decision to treat people differently based on their race or other protected characteristics.
doc69	Disparate impact	While disparate impact is a legal theory of liability under Title VII, adverse impact is one element of that doctrine, which measures the effect an employment practice has on a class protected by Title VII. In the Uniform Guidelines on Employee Selection Procedures, an adverse impact is defined as a "substantially different rate of selection in hiring, promotion, or other employment decision which works to the disadvantage of members of a race, sex, or ethnic group".[8] A "substantially different" rate is typically defined in government enforcement or Title VII litigation settings using the 80% Rule, statistical significance tests, and/or practical significance tests. Adverse impact is often used interchangeably with "disparate impact," which was a legal term coined in one of the most significant U.S. Supreme Court rulings on disparate or adverse impact: Griggs v. Duke Power Co., 1971. Adverse Impact does not mean that an individual in a majority group is given preference over a minority group. However, having adverse impact does mean that there is the "potential" for discrimination in the hiring process and it could warrant investigation.[9]
doc70	Disparate impact	The 80% test was originally framed by a panel of 32 professionals (called the Technical Advisory Committee on Testing, or TACT) assembled by the State of California Fair Employment Practice Commission (FEPC) in 1971, which published the State of California Guidelines on Employee Selection Procedures in October, 1972. This was the first official government document that listed the 80% test in the context of adverse impact, and was later codified in the 1978 Uniform Guidelines on Employee Selection Procedures, a document used by the U.S. Equal Employment Opportunity Commission (EEOC), Department of Labor, and Department of Justice in Title VII enforcement.[10]
doc71	Disparate impact	Originally, the Uniform Guidelines on Employee Selection Procedures provided a simple "80 percent" rule for determining that a company's selection system was having an "adverse impact" on a minority group. The rule was based on the rates at which job applicants were hired. For example, if XYZ Company hired 50 percent of the men applying for work in a predominantly male occupation while hiring only 20 percent of the female applicants, one could look at the ratio of those two hiring rates to judge whether there might be a discrimination problem. The ratio of 20:50 means that the rate of hiring for female applicants is only 40 percent of the rate of hiring for male applicants. That is, 20 divided by 50 equals 0.40, which is equivalent to 40 percent. Clearly, 40 percent is well below the 80 percent that was arbitrarily set as an acceptable difference in hiring rates. Therefore, in this example, XYZ Company could have been called upon to prove that there was a legitimate reason for hiring men at a rate so much higher than the rate of hiring women. Since the 1980s, courts in the U.S. have questioned the arbitrary nature of the 80 percent rule, making the rule less important than it was when the Uniform Guidelines were first published. A recent memorandum from the U.S. Equal Employment Opportunities Commission suggests that a more defensible standard would be based on comparing a company's hiring rate of a particular group with the rate that would occur if the company simply selected people at random.[11] In other words, if a company's selection system made it statistically more difficult than pure chance for a member of a certain group, such as women or African-Americans, to get a job, then this could be reasonably viewed as evidence that the selection system was systematically screening out members of that social group.
doc72	Disparate impact	The concept of practical significance for adverse impact was first introduced by Section 4D of the Uniform Guidelines,[12] which states "Smaller differences in selection rate may nevertheless constitute adverse impact, where they are significant in both statistical and practical terms ..." Several federal court cases have applied practical significance tests to adverse impact analyses to assess the "practicality" or "stability" of the results. This is typically done by evaluating the change to the statistical significance tests after hypothetically changing focal group members selection status from "failing" to "passing" (see for example, Contreras v. City of Los Angeles (656 F.2d 1267, 9th Cir. 1981); U.S. v. Commonwealth of Virginia (569 F.2d 1300, 4th Cir. 1978); and Waisome v. Port Authority (948 F.2d 1370, 1376, 2d Cir. 1991)).
doc73	Disparate impact	This form of discrimination occurs where an employer does not intend to discriminate; to the contrary, it occurs when identical standards or procedures are applied to everyone, despite the fact that they lead to a substantial difference in employment outcomes for the members of a particular group and they are unrelated to successful job performance. An important thing to note is that disparate impact is not, in and of itself, illegal.[13] This is because disparate impact only becomes illegal if the employer cannot justify the employment practice causing the adverse impact as a "job related for the position in question and consistent with business necessity" (called the "business necessity defense").[14]
doc74	Disparate impact	For example, a fire department requiring applicants to carry a 100 lb (50Â kg) pack up three flights of stairs. The upper-body strength required typically has an adverse impact on women. The fire department would have to show that this requirement is necessary and job-related. This typically requires employers to conduct validation studies that address both the Uniform Guidelines and professional standards. Accordingly, a fire department could be liable for "discriminating" against female job applicants solely because it failed to prove to a court's satisfaction that the 100-pound requirement was "necessary", even though the department never intended to hinder women's ability to become firefighters.
doc75	Disparate impact	Disparate impact is not the same as disparate treatment. Disparate treatment refers to the "intentional" discrimination of certain people groups during the hiring, promoting or placement process.
doc76	Disparate impact	The disparate impact theory has application also in the housing context under Title VIII of the Civil Rights Act of 1968, also known as The Fair Housing Act,. The ten federal appellate courts that have addressed the issue have all determined that one may establish a Fair Housing Act violation through the disparate impact theory of liability. The U.S. Department of Housing and Urban Development's Office of Fair Housing and Equal Opportunity, the federal government which administers the Fair Housing Act, issued a proposed regulation on November 16, 2011 setting forth how HUD applies disparate impact in Fair Housing Act cases. On February 8, 2013, HUD issued its Final Rule.[15]
doc77	Disparate impact	Until 2015, the U.S. Supreme Court had not yet determined whether the Fair Housing Act allowed for claims of disparate impact. This question reached the Supreme Court twice since 2012, first in Magner v. Gallagher and then in Township of Mount Holly v. Mount Holly Gardens Citizens. The Supreme Court seemed likely to rule that the Act does not contain a disparate impact provision, but both cases settled before the Court could issue a decision. The federal government appeared to pressure the settlement in one or both cases in an effort to preserve the disparate impact theory.[16][17][18]
doc78	Disparate impact	On June 25, 2015, by a 5â€“4 decision in Texas Department of Housing and Community Affairs v. Inclusive Communities Project, the Supreme Court held[7] that disparate-impact claims are cognizable under the Fair Housing Act. In an opinion by Justice Kennedy, "Recognition of disparate-impact claims is also consistent with the central purpose of the FHA, which, like Title VII and the ADEA, was enacted to eradicate discriminatory practices within a sector of the Nation's economy. Suits targeting unlawful zoning laws and other housing restrictions that unfairly exclude minorities from certain neighborhoods without sufficient justification are at the heartland of disparate-impact liability...Recognition of disparate impact liability under the FHA plays an important role in uncovering discriminatory intent: it permits plaintiffs to counteract unconscious prejudices and disguised animus that escape easy classification as disparate treatment." Under the Court's ruling in Inclusive Communities, in order to prove a case of disparate impact housing discrimination, the following must occur:
doc79	Disparate impact	The disparate impact theory of liability is controversial for several reasons. First, it labels certain unintended effects as "discriminatory", although discrimination is not an intentional act. Second, the theory is in tension with disparate treatment provisions under civil rights laws as well as the U.S. Constitution's guarantee of equal protection. For example, if the hypothetical fire department discussed above used the 100-pound requirement, that policy might disproportionately exclude female job applicants from employment. Under the 80% rule mentioned above, unsuccessful female job applicants would have a prima facie case of disparate impact "discrimination" against the department if they passed the 100-pound test at a rate less than 80% of the rate at which men passed the test. In order to avoid a lawsuit by the female job applicants, the department might refuse to hire anyone from its applicant poolâ€”in other words, the department may refuse to hire anyone because too many of the successful job applicants were male. Thus, the employer would have intentionally discriminated against the successful male job applicants because of their gender, and that likely amounts to illegal disparate treatment and a violation of the Constitution's right to equal protection. In the 2009 case Ricci v. DeStefano, the U.S. Supreme Court did rule that a fire department committed illegal disparate treatment by refusing to promote white firefighters, in an effort to avoid disparate impact liability in a potential lawsuit by black and Hispanic firefighters who disproportionately failed the required tests for promotion. Although the Court in that case did not reach the constitutional issue, Justice Scalia's concurring opinion suggested the fire department also violated the constitutional right to equal protection. Even before Ricci, lower federal courts have ruled that actions taken to avoid potential disparate impact liability violate the constitutional right to equal protection. One such case is Biondo v. City of Chicago, Illinois, from the Seventh Circuit.
doc80	Disparate impact	In 2013, the Equal Employment Opportunity Commission (EEOC) filed a suit, EEOC v. FREEMAN,[19] against the use of typical criminal-background and credit checks during the hiring process. While admitting that there are many legitimate and race-neutral reasons for employers to screen out convicted criminals and debtors, the EEOC presented the theory that this practice is discriminatory because minorities in the U.S. are more likely to be convicted criminals with bad credit histories than white Americans. Ergo, employers should have to include criminals and debtors in their hiring. In this instance U.S. District Judge Roger Titus ruled firmly against the disparate impact theory, stating that EEOC's action had been "a theory in search of facts to support it." "By bringing actions of this nature, the EEOC has placed many employers in the "Hobson's choice" of ignoring criminal history and credit background, thus exposing themselves to potential liability for criminal and fraudulent acts committed by employees, on the one hand, or incurring the wrath of the EEOC for having utilized information deemed fundamental by most employers. Something more... must be utilized to justify a disparate impact claim based upon criminal history and credit checks. To require less, would be to condemn the use of common sense, and this is simply not what the laws of this country require."
doc81	Disparate impact	The disparate impact theory is especially controversial under the Fair Housing Act because the Act regulates many activities relating to housing, insurance, and mortgage loansâ€”and some scholars have argued that the theory's use under the Fair Housing Act, combined with extensions of the Community Reinvestment Act, contributed to rise of sub-prime lending and the crash of the U.S. housing market and ensuing global economic recession.[20]
doc82	This Is Us (TV series)	This Is Us is an American family drama television series created by Dan Fogelman that premiered on NBC on September 20, 2016.[1] The series stars an ensemble cast featuring Milo Ventimiglia, Mandy Moore, Sterling K. Brown, Chrissy Metz, Justin Hartley, Susan Kelechi Watson, Chris Sullivan, Ron Cephas Jones, Jon Huertas, Alexandra Breckenridge, Niles Fitch, Logan Shroyer, Hannah Zeile, Mackenzie Hancsicsak, Parker Bates, Lonnie Chavis, Eris Baker, and Faithe Herman. The program details the lives and families of two parents, and their three children born on the same day as their father's birthday.[1] This Is Us is filmed in Los Angeles.[2]
doc83	This Is Us (TV series)	The series has received positive reviews and has been nominated for Best Television Series – Drama at the 74th Golden Globe Awards and Best Drama Series at the 7th Critics' Choice Awards, as well as being chosen as a Top Television Program by the American Film Institute. Sterling K. Brown has received an Emmy, a Golden Globe, a Critics' Choice Award, and an NAACP Image Award for his acting in the series. Mandy Moore and Chrissy Metz received Golden Globe nominations for Best Supporting Actress. In 2017, the series received ten Emmy nominations, including Outstanding Drama Series, with Brown winning for Outstanding Lead Actor in a Drama Series.
doc84	This Is Us (TV series)	On September 27, 2016, NBC picked up the series for a full season of 18 episodes.[3] In January 2017, NBC renewed the series for two additional seasons of 18 episodes each.[4] The second season premiered on September 26, 2017.
doc85	This Is Us (TV series)	The series follows the lives of siblings Kevin, Kate, and Randall (known as the "Big Three"), and their parents Jack and Rebecca Pearson. It takes place in the present and using flashbacks, at various times in the past. Kevin and Kate are the two surviving members of a triplet pregnancy, born six weeks premature on Jack's 36th birthday in 1980; their brother is stillborn. Believing they were meant to have three children, Jack and Rebecca, who are white, decide to adopt Randall, a black child born the same day and brought to the same hospital after his biological father abandoned him at a fire station. Jack dies when his children are 17.
doc86	This Is Us (TV series)	Most episodes feature a storyline taking place in the present (2016–2018, contemporaneous with airing) and a storyline taking place at a set time in the past; but some episodes are set in one time period or use multiple flashback time periods. Flashbacks often focus on Jack and Rebecca c.1980 both before and after their babies' birth, or on the family when the Big Three are children (at least ages 8–10) or adolescents; these scenes usually take place in Pittsburgh, where the Big Three are born and raised. Various other time periods and locations have also served a settings. As adults, Kate lives in Los Angeles, Randall and his family are in New Jersey, and Kevin relocates from Los Angeles to New York City.
doc87	This Is Us (TV series)	Fogelman intentionally recruited behind-the-scenes talent that would reflect the diversity of his cast, with the goal of bringing greater authenticity to the dialog and storylines. These include black directors Regina King and George Tillman, Jr. and black female writers Kay Oyegun and Jas Waters (part of a 30% black core writing staff that far outpaces the industry standard of 5%).[5] [6] In addition, Fogelman's sister Deborah, whose struggles with weight were one of the initial inspirations for the show, serves as a consultant.[7]
doc88	This Is Us (TV series)	In May 2017, Hulu acquired the SVOD rights to new and past episodes of the series to air exclusively on Hulu, in addition to NBC.com and the NBC app.[46]
doc89	This Is Us (TV series)	The review aggregation website Rotten Tomatoes reported an 89% approval rating for the first season with an average rating of 7.72/10 based on 63 reviews. The website's critical consensus reads, "Featuring full-tilt heartstring-tugging family drama, This Is Us will provide a suitable surrogate for those who have felt a void in their lives since Parenthood went off the air."[47] Metacritic, which uses a weighted average, assigned the season a score of 76 out of 100 based on 34 reviews, indicating "generally favorable reviews".[48] Season 2 received a 94% approval rating from Rotten Tomatoes based on 17 reviews.[49]
doc90	This Is Us (TV series)	Entertainment Weekly gave the first few episodes of This Is Us a rating of B, calling it "a refreshing respite from the relational violence and pessimism that marks the other buzz soaps that have bubbled forth from a culture of divisiveness". Moreover, they praised all the actors, specifically Sterling K. Brown, for being able to navigate "his scenes with such intelligence, authenticity, and charisma".[50]
doc91	Market economy	A market economy is an economic system where decisions regarding investment, production, and distribution are based on the interplay of supply and demand,[1] which determines the prices of goods and services.[2] The major defining characteristic of a market economy is that investment decisions, or the allocation of producer good, are primarily made through capital and financial markets.[3] This is contrasted with a planned economy, where investment and production decisions are embodied in an integrated plan of production established by a state or other organizational body that controls the factors of production.
doc92	Market economy	Market economies can range from free market systems to regulated markets and various forms of interventionist variants. In reality, free markets do not exist in pure form, since societies and governments all regulate them to varying degrees.[4][5] Different perspectives exist as to how strong a role the government should have in both guiding and regulating market economies and addressing the inequalities the market naturally produces. Most existing market economies include a degree of state-directed activity or economic planning, and are thus classified as mixed economies. The term free-market economy is sometimes used synonymously with market economy.[6]
doc93	Market economy	Market economies do not logically presuppose the existence of private ownership of the means of production. A market economy can and often does include various types of cooperatives, collectives, or autonomous state agencies that acquire and exchange capital goods in capital markets. These all utilize a market-determined free price system to allocate capital goods and labor.[3] In addition, there are many variations of market socialism, some of which involve employee-owned enterprises based on self-management; as well as models that involve the combination of public ownership of the means of production with factor markets.[7]
doc94	Market economy	Capitalism generally refers to an economic system where the means of production are largely or entirely privately owned and operated for a profit, structured on the process of capital accumulation. In general, in capitalist systems investment, distribution, income, and prices are determined by markets, whether regulated or unregulated.
doc95	Market economy	There are different variations of capitalism with different relationships to markets. In Laissez-faire and free market variations of capitalism, markets are utilized most extensively with minimal or no state intervention and regulation over prices and the supply of goods and services. In interventionist, welfare capitalism and mixed economies, markets continue to play a dominant role but are regulated to some extent by government in order to correct market failures or to promote social welfare. In state capitalist systems, markets are relied upon the least, with the state relying heavily on either indirect economic planning and/or state-owned enterprises to accumulate capital.
doc96	Market economy	Capitalism has been dominant in the Western world since the end of feudalism, but most feel[who?] that the term "mixed economies" more precisely describes most contemporary economies, due to their containing both private-owned and state-owned enterprises. In capitalism, prices determine the demand-supply scale. For example, higher demand for certain goods and services lead to higher prices and lower demand for certain goods lead to lower prices.
doc97	Market economy	Laissez-faire is synonymous with what was referred to as strict capitalist free market economy during the early and mid-19th century[citation needed] as a classical liberal (right-libertarian) ideal to achieve. It is generally understood that the necessary components for the functioning of an idealized free market include the complete absence of government regulation, subsidies, artificial price pressures, and government-granted monopolies (usually classified as coercive monopoly by free market advocates) and no taxes or tariffs other than what is necessary for the government to provide protection from coercion and theft, maintaining peace and property rights, and providing for basic public goods. Right-libertarian advocates of anarcho-capitalism see the state as morally illegitimate and economically unnecessary and destructive.
doc98	Market economy	Free-market economy refers to an economic system where prices for goods and services are set freely by the forces of supply and demand and are allowed to reach their point of equilibrium without intervention by government policy. It typically entails support for highly competitive markets, private ownership of productive enterprises. Laissez-faire is a more extensive form of free-market economy where the role of the state is limited to protecting property rights.
doc99	Market economy	Welfare capitalism refers to a capitalist economy that includes public policies favoring extensive provisions for social welfare services. The economic mechanism involves a free market and the predominance of privately owned enterprises in the economy, but public provision of universal welfare services aimed at enhancing individual autonomy and maximizing equality. Examples of contemporary welfare capitalism include the Nordic model of capitalism predominant in Northern Europe.[8]

End of preview. Expand in Dataset Viewer.

YAML Metadata Error: "size_categories" must be a string

YAML Metadata Warning: The task_categories "zero-shot-retrieval" is not in the official list: text-classification, token-classification, table-question-answering, question-answering, zero-shot-classification, translation, summarization, feature-extraction, text-generation, text2text-generation, fill-mask, sentence-similarity, text-to-speech, text-to-audio, automatic-speech-recognition, audio-to-audio, audio-classification, voice-activity-detection, depth-estimation, image-classification, object-detection, image-segmentation, text-to-image, image-to-text, image-to-image, image-to-video, unconditional-image-generation, video-classification, reinforcement-learning, robotics, tabular-classification, tabular-regression, tabular-to-text, table-to-text, multiple-choice, text-retrieval, time-series-forecasting, text-to-video, image-text-to-text, visual-question-answering, document-question-answering, zero-shot-image-classification, graph-ml, mask-generation, zero-shot-object-detection, text-to-3d, image-to-3d, image-feature-extraction, other

YAML Metadata Warning: The task_categories "information-retrieval" is not in the official list: text-classification, token-classification, table-question-answering, question-answering, zero-shot-classification, translation, summarization, feature-extraction, text-generation, text2text-generation, fill-mask, sentence-similarity, text-to-speech, text-to-audio, automatic-speech-recognition, audio-to-audio, audio-classification, voice-activity-detection, depth-estimation, image-classification, object-detection, image-segmentation, text-to-image, image-to-text, image-to-image, image-to-video, unconditional-image-generation, video-classification, reinforcement-learning, robotics, tabular-classification, tabular-regression, tabular-to-text, table-to-text, multiple-choice, text-retrieval, time-series-forecasting, text-to-video, image-text-to-text, visual-question-answering, document-question-answering, zero-shot-image-classification, graph-ml, mask-generation, zero-shot-object-detection, text-to-3d, image-to-3d, image-feature-extraction, other

YAML Metadata Warning: The task_categories "zero-shot-information-retrieval" is not in the official list: text-classification, token-classification, table-question-answering, question-answering, zero-shot-classification, translation, summarization, feature-extraction, text-generation, text2text-generation, fill-mask, sentence-similarity, text-to-speech, text-to-audio, automatic-speech-recognition, audio-to-audio, audio-classification, voice-activity-detection, depth-estimation, image-classification, object-detection, image-segmentation, text-to-image, image-to-text, image-to-image, image-to-video, unconditional-image-generation, video-classification, reinforcement-learning, robotics, tabular-classification, tabular-regression, tabular-to-text, table-to-text, multiple-choice, text-retrieval, time-series-forecasting, text-to-video, image-text-to-text, visual-question-answering, document-question-answering, zero-shot-image-classification, graph-ml, mask-generation, zero-shot-object-detection, text-to-3d, image-to-3d, image-feature-extraction, other

YAML Metadata Warning: The task_ids "passage-retrieval" is not in the official list: acceptability-classification, entity-linking-classification, fact-checking, intent-classification, language-identification, multi-class-classification, multi-label-classification, multi-input-text-classification, natural-language-inference, semantic-similarity-classification, sentiment-classification, topic-classification, semantic-similarity-scoring, sentiment-scoring, sentiment-analysis, hate-speech-detection, text-scoring, named-entity-recognition, part-of-speech, parsing, lemmatization, word-sense-disambiguation, coreference-resolution, extractive-qa, open-domain-qa, closed-domain-qa, news-articles-summarization, news-articles-headline-generation, dialogue-modeling, dialogue-generation, conversational, language-modeling, text-simplification, explanation-generation, abstractive-qa, open-domain-abstractive-qa, closed-domain-qa, open-book-qa, closed-book-qa, slot-filling, masked-language-modeling, keyword-spotting, speaker-identification, audio-intent-classification, audio-emotion-recognition, audio-language-identification, multi-label-image-classification, multi-class-image-classification, face-detection, vehicle-detection, instance-segmentation, semantic-segmentation, panoptic-segmentation, image-captioning, image-inpainting, image-colorization, super-resolution, grasping, task-planning, tabular-multi-class-classification, tabular-multi-label-classification, tabular-single-column-regression, rdf-to-text, multiple-choice-qa, multiple-choice-coreference-resolution, document-retrieval, utterance-retrieval, entity-linking-retrieval, fact-checking-retrieval, univariate-time-series-forecasting, multivariate-time-series-forecasting, visual-question-answering, document-question-answering

YAML Metadata Warning: The task_ids "tweet-retrieval" is not in the official list: acceptability-classification, entity-linking-classification, fact-checking, intent-classification, language-identification, multi-class-classification, multi-label-classification, multi-input-text-classification, natural-language-inference, semantic-similarity-classification, sentiment-classification, topic-classification, semantic-similarity-scoring, sentiment-scoring, sentiment-analysis, hate-speech-detection, text-scoring, named-entity-recognition, part-of-speech, parsing, lemmatization, word-sense-disambiguation, coreference-resolution, extractive-qa, open-domain-qa, closed-domain-qa, news-articles-summarization, news-articles-headline-generation, dialogue-modeling, dialogue-generation, conversational, language-modeling, text-simplification, explanation-generation, abstractive-qa, open-domain-abstractive-qa, closed-domain-qa, open-book-qa, closed-book-qa, slot-filling, masked-language-modeling, keyword-spotting, speaker-identification, audio-intent-classification, audio-emotion-recognition, audio-language-identification, multi-label-image-classification, multi-class-image-classification, face-detection, vehicle-detection, instance-segmentation, semantic-segmentation, panoptic-segmentation, image-captioning, image-inpainting, image-colorization, super-resolution, grasping, task-planning, tabular-multi-class-classification, tabular-multi-label-classification, tabular-single-column-regression, rdf-to-text, multiple-choice-qa, multiple-choice-coreference-resolution, document-retrieval, utterance-retrieval, entity-linking-retrieval, fact-checking-retrieval, univariate-time-series-forecasting, multivariate-time-series-forecasting, visual-question-answering, document-question-answering

YAML Metadata Warning: The task_ids "citation-prediction-retrieval" is not in the official list: acceptability-classification, entity-linking-classification, fact-checking, intent-classification, language-identification, multi-class-classification, multi-label-classification, multi-input-text-classification, natural-language-inference, semantic-similarity-classification, sentiment-classification, topic-classification, semantic-similarity-scoring, sentiment-scoring, sentiment-analysis, hate-speech-detection, text-scoring, named-entity-recognition, part-of-speech, parsing, lemmatization, word-sense-disambiguation, coreference-resolution, extractive-qa, open-domain-qa, closed-domain-qa, news-articles-summarization, news-articles-headline-generation, dialogue-modeling, dialogue-generation, conversational, language-modeling, text-simplification, explanation-generation, abstractive-qa, open-domain-abstractive-qa, closed-domain-qa, open-book-qa, closed-book-qa, slot-filling, masked-language-modeling, keyword-spotting, speaker-identification, audio-intent-classification, audio-emotion-recognition, audio-language-identification, multi-label-image-classification, multi-class-image-classification, face-detection, vehicle-detection, instance-segmentation, semantic-segmentation, panoptic-segmentation, image-captioning, image-inpainting, image-colorization, super-resolution, grasping, task-planning, tabular-multi-class-classification, tabular-multi-label-classification, tabular-single-column-regression, rdf-to-text, multiple-choice-qa, multiple-choice-coreference-resolution, document-retrieval, utterance-retrieval, entity-linking-retrieval, fact-checking-retrieval, univariate-time-series-forecasting, multivariate-time-series-forecasting, visual-question-answering, document-question-answering

YAML Metadata Warning: The task_ids "duplication-question-retrieval" is not in the official list: acceptability-classification, entity-linking-classification, fact-checking, intent-classification, language-identification, multi-class-classification, multi-label-classification, multi-input-text-classification, natural-language-inference, semantic-similarity-classification, sentiment-classification, topic-classification, semantic-similarity-scoring, sentiment-scoring, sentiment-analysis, hate-speech-detection, text-scoring, named-entity-recognition, part-of-speech, parsing, lemmatization, word-sense-disambiguation, coreference-resolution, extractive-qa, open-domain-qa, closed-domain-qa, news-articles-summarization, news-articles-headline-generation, dialogue-modeling, dialogue-generation, conversational, language-modeling, text-simplification, explanation-generation, abstractive-qa, open-domain-abstractive-qa, closed-domain-qa, open-book-qa, closed-book-qa, slot-filling, masked-language-modeling, keyword-spotting, speaker-identification, audio-intent-classification, audio-emotion-recognition, audio-language-identification, multi-label-image-classification, multi-class-image-classification, face-detection, vehicle-detection, instance-segmentation, semantic-segmentation, panoptic-segmentation, image-captioning, image-inpainting, image-colorization, super-resolution, grasping, task-planning, tabular-multi-class-classification, tabular-multi-label-classification, tabular-single-column-regression, rdf-to-text, multiple-choice-qa, multiple-choice-coreference-resolution, document-retrieval, utterance-retrieval, entity-linking-retrieval, fact-checking-retrieval, univariate-time-series-forecasting, multivariate-time-series-forecasting, visual-question-answering, document-question-answering

YAML Metadata Warning: The task_ids "argument-retrieval" is not in the official list: acceptability-classification, entity-linking-classification, fact-checking, intent-classification, language-identification, multi-class-classification, multi-label-classification, multi-input-text-classification, natural-language-inference, semantic-similarity-classification, sentiment-classification, topic-classification, semantic-similarity-scoring, sentiment-scoring, sentiment-analysis, hate-speech-detection, text-scoring, named-entity-recognition, part-of-speech, parsing, lemmatization, word-sense-disambiguation, coreference-resolution, extractive-qa, open-domain-qa, closed-domain-qa, news-articles-summarization, news-articles-headline-generation, dialogue-modeling, dialogue-generation, conversational, language-modeling, text-simplification, explanation-generation, abstractive-qa, open-domain-abstractive-qa, closed-domain-qa, open-book-qa, closed-book-qa, slot-filling, masked-language-modeling, keyword-spotting, speaker-identification, audio-intent-classification, audio-emotion-recognition, audio-language-identification, multi-label-image-classification, multi-class-image-classification, face-detection, vehicle-detection, instance-segmentation, semantic-segmentation, panoptic-segmentation, image-captioning, image-inpainting, image-colorization, super-resolution, grasping, task-planning, tabular-multi-class-classification, tabular-multi-label-classification, tabular-single-column-regression, rdf-to-text, multiple-choice-qa, multiple-choice-coreference-resolution, document-retrieval, utterance-retrieval, entity-linking-retrieval, fact-checking-retrieval, univariate-time-series-forecasting, multivariate-time-series-forecasting, visual-question-answering, document-question-answering

YAML Metadata Warning: The task_ids "news-retrieval" is not in the official list: acceptability-classification, entity-linking-classification, fact-checking, intent-classification, language-identification, multi-class-classification, multi-label-classification, multi-input-text-classification, natural-language-inference, semantic-similarity-classification, sentiment-classification, topic-classification, semantic-similarity-scoring, sentiment-scoring, sentiment-analysis, hate-speech-detection, text-scoring, named-entity-recognition, part-of-speech, parsing, lemmatization, word-sense-disambiguation, coreference-resolution, extractive-qa, open-domain-qa, closed-domain-qa, news-articles-summarization, news-articles-headline-generation, dialogue-modeling, dialogue-generation, conversational, language-modeling, text-simplification, explanation-generation, abstractive-qa, open-domain-abstractive-qa, closed-domain-qa, open-book-qa, closed-book-qa, slot-filling, masked-language-modeling, keyword-spotting, speaker-identification, audio-intent-classification, audio-emotion-recognition, audio-language-identification, multi-label-image-classification, multi-class-image-classification, face-detection, vehicle-detection, instance-segmentation, semantic-segmentation, panoptic-segmentation, image-captioning, image-inpainting, image-colorization, super-resolution, grasping, task-planning, tabular-multi-class-classification, tabular-multi-label-classification, tabular-single-column-regression, rdf-to-text, multiple-choice-qa, multiple-choice-coreference-resolution, document-retrieval, utterance-retrieval, entity-linking-retrieval, fact-checking-retrieval, univariate-time-series-forecasting, multivariate-time-series-forecasting, visual-question-answering, document-question-answering

YAML Metadata Warning: The task_ids "biomedical-information-retrieval" is not in the official list: acceptability-classification, entity-linking-classification, fact-checking, intent-classification, language-identification, multi-class-classification, multi-label-classification, multi-input-text-classification, natural-language-inference, semantic-similarity-classification, sentiment-classification, topic-classification, semantic-similarity-scoring, sentiment-scoring, sentiment-analysis, hate-speech-detection, text-scoring, named-entity-recognition, part-of-speech, parsing, lemmatization, word-sense-disambiguation, coreference-resolution, extractive-qa, open-domain-qa, closed-domain-qa, news-articles-summarization, news-articles-headline-generation, dialogue-modeling, dialogue-generation, conversational, language-modeling, text-simplification, explanation-generation, abstractive-qa, open-domain-abstractive-qa, closed-domain-qa, open-book-qa, closed-book-qa, slot-filling, masked-language-modeling, keyword-spotting, speaker-identification, audio-intent-classification, audio-emotion-recognition, audio-language-identification, multi-label-image-classification, multi-class-image-classification, face-detection, vehicle-detection, instance-segmentation, semantic-segmentation, panoptic-segmentation, image-captioning, image-inpainting, image-colorization, super-resolution, grasping, task-planning, tabular-multi-class-classification, tabular-multi-label-classification, tabular-single-column-regression, rdf-to-text, multiple-choice-qa, multiple-choice-coreference-resolution, document-retrieval, utterance-retrieval, entity-linking-retrieval, fact-checking-retrieval, univariate-time-series-forecasting, multivariate-time-series-forecasting, visual-question-answering, document-question-answering

YAML Metadata Warning: The task_ids "question-answering-retrieval" is not in the official list: acceptability-classification, entity-linking-classification, fact-checking, intent-classification, language-identification, multi-class-classification, multi-label-classification, multi-input-text-classification, natural-language-inference, semantic-similarity-classification, sentiment-classification, topic-classification, semantic-similarity-scoring, sentiment-scoring, sentiment-analysis, hate-speech-detection, text-scoring, named-entity-recognition, part-of-speech, parsing, lemmatization, word-sense-disambiguation, coreference-resolution, extractive-qa, open-domain-qa, closed-domain-qa, news-articles-summarization, news-articles-headline-generation, dialogue-modeling, dialogue-generation, conversational, language-modeling, text-simplification, explanation-generation, abstractive-qa, open-domain-abstractive-qa, closed-domain-qa, open-book-qa, closed-book-qa, slot-filling, masked-language-modeling, keyword-spotting, speaker-identification, audio-intent-classification, audio-emotion-recognition, audio-language-identification, multi-label-image-classification, multi-class-image-classification, face-detection, vehicle-detection, instance-segmentation, semantic-segmentation, panoptic-segmentation, image-captioning, image-inpainting, image-colorization, super-resolution, grasping, task-planning, tabular-multi-class-classification, tabular-multi-label-classification, tabular-single-column-regression, rdf-to-text, multiple-choice-qa, multiple-choice-coreference-resolution, document-retrieval, utterance-retrieval, entity-linking-retrieval, fact-checking-retrieval, univariate-time-series-forecasting, multivariate-time-series-forecasting, visual-question-answering, document-question-answering

Dataset Card for BEIR Benchmark

Dataset Summary

BEIR is a heterogeneous benchmark that has been built from 18 diverse datasets representing 9 information retrieval tasks:

Fact-checking: FEVER, Climate-FEVER, SciFact
Question-Answering: NQ, HotpotQA, FiQA-2018
Bio-Medical IR: TREC-COVID, BioASQ, NFCorpus
News Retrieval: TREC-NEWS, Robust04
Argument Retrieval: Touche-2020, ArguAna
Duplicate Question Retrieval: Quora, CqaDupstack
Citation-Prediction: SCIDOCS
Tweet Retrieval: Signal-1M
Entity Retrieval: DBPedia

All these datasets have been preprocessed and can be used for your experiments.

Supported Tasks and Leaderboards

The dataset supports a leaderboard that evaluates models against task-specific metrics such as F1 or EM, as well as their ability to retrieve supporting information from Wikipedia.

The current best performing models can be found here.

Languages

All tasks are in English (en).

Dataset Structure

All BEIR datasets must contain a corpus, queries and qrels (relevance judgments file). They must be in the following format:

corpus file: a .jsonl file (jsonlines) that contains a list of dictionaries, each with three fields _id with unique document identifier, title with document title (optional) and text with document paragraph or passage. For example: {"_id": "doc1", "title": "Albert Einstein", "text": "Albert Einstein was a German-born...."}
queries file: a .jsonl file (jsonlines) that contains a list of dictionaries, each with two fields _id with unique query identifier and text with query text. For example: {"_id": "q1", "text": "Who developed the mass-energy equivalence formula?"}
qrels file: a .tsv file (tab-seperated) that contains three columns, i.e. the query-id, corpus-id and score in this order. Keep 1st row as header. For example: q1 doc1 1

Data Instances

A high level example of any beir dataset:

corpus = {
    "doc1" : {
        "title": "Albert Einstein", 
        "text": "Albert Einstein was a German-born theoretical physicist. who developed the theory of relativity, \
                 one of the two pillars of modern physics (alongside quantum mechanics). His work is also known for \
                 its influence on the philosophy of science. He is best known to the general public for his massâ€“energy \
                 equivalence formula E = mc2, which has been dubbed 'the world's most famous equation'. He received the 1921 \
                 Nobel Prize in Physics 'for his services to theoretical physics, and especially for his discovery of the law \
                 of the photoelectric effect', a pivotal step in the development of quantum theory."
        },
    "doc2" : {
        "title": "", # Keep title an empty string if not present
        "text": "Wheat beer is a top-fermented beer which is brewed with a large proportion of wheat relative to the amount of \
                 malted barley. The two main varieties are German WeiÃŸbier and Belgian witbier; other types include Lambic (made\
                 with wild yeast), Berliner Weisse (a cloudy, sour beer), and Gose (a sour, salty beer)."
    },
}

queries = {
    "q1" : "Who developed the mass-energy equivalence formula?",
    "q2" : "Which beer is brewed with a large proportion of wheat?"
}

qrels = {
    "q1" : {"doc1": 1},
    "q2" : {"doc2": 1},
}

Data Fields

Examples from all configurations have the following features:

Corpus

corpus: a dict feature representing the document title and passage text, made up of:
- _id: a string feature representing the unique document id
  - title: a string feature, denoting the title of the document.
  - text: a string feature, denoting the text of the document.

Queries

queries: a dict feature representing the query, made up of:
- _id: a string feature representing the unique query id
- text: a string feature, denoting the text of the query.

Qrels

qrels: a dict feature representing the query document relevance judgements, made up of:
- _id: a string feature representing the query id
  - _id: a string feature, denoting the document id.
  - score: a int32 feature, denoting the relevance judgement between query and document.

Data Splits

Dataset	Website	BEIR-Name	Type	Queries	Corpus	Rel D/Q	Down-load	md5
MSMARCO	Homepage	`msmarco`	`train` `dev` `test`	6,980	8.84M	1.1	Link	`444067daf65d982533ea17ebd59501e4`
TREC-COVID	Homepage	`trec-covid`	`test`	50	171K	493.5	Link	`ce62140cb23feb9becf6270d0d1fe6d1`
NFCorpus	Homepage	`nfcorpus`	`train` `dev` `test`	323	3.6K	38.2	Link	`a89dba18a62ef92f7d323ec890a0d38d`
BioASQ	Homepage	`bioasq`	`train` `test`	500	14.91M	8.05	No	How to Reproduce?
NQ	Homepage	`nq`	`train` `test`	3,452	2.68M	1.2	Link	`d4d3d2e48787a744b6f6e691ff534307`
HotpotQA	Homepage	`hotpotqa`	`train` `dev` `test`	7,405	5.23M	2.0	Link	`f412724f78b0d91183a0e86805e16114`
FiQA-2018	Homepage	`fiqa`	`train` `dev` `test`	648	57K	2.6	Link	`17918ed23cd04fb15047f73e6c3bd9d9`
Signal-1M(RT)	Homepage	`signal1m`	`test`	97	2.86M	19.6	No	How to Reproduce?
TREC-NEWS	Homepage	`trec-news`	`test`	57	595K	19.6	No	How to Reproduce?
ArguAna	Homepage	`arguana`	`test`	1,406	8.67K	1.0	Link	`8ad3e3c2a5867cdced806d6503f29b99`
Touche-2020	Homepage	`webis-touche2020`	`test`	49	382K	19.0	Link	`46f650ba5a527fc69e0a6521c5a23563`
CQADupstack	Homepage	`cqadupstack`	`test`	13,145	457K	1.4	Link	`4e41456d7df8ee7760a7f866133bda78`
Quora	Homepage	`quora`	`dev` `test`	10,000	523K	1.6	Link	`18fb154900ba42a600f84b839c173167`
DBPedia	Homepage	`dbpedia-entity`	`dev` `test`	400	4.63M	38.2	Link	`c2a39eb420a3164af735795df012ac2c`
SCIDOCS	Homepage	`scidocs`	`test`	1,000	25K	4.9	Link	`38121350fc3a4d2f48850f6aff52e4a9`
FEVER	Homepage	`fever`	`train` `dev` `test`	6,666	5.42M	1.2	Link	`5a818580227bfb4b35bb6fa46d9b6c03`
Climate-FEVER	Homepage	`climate-fever`	`test`	1,535	5.42M	3.0	Link	`8b66f0a9126c521bae2bde127b4dc99d`
SciFact	Homepage	`scifact`	`train` `test`	300	5K	1.1	Link	`5f7d1de60b170fc8027bb7898e2efca1`
Robust04	Homepage	`robust04`	`test`	249	528K	69.9	No	How to Reproduce?

Dataset Creation

Curation Rationale

[Needs More Information]

Source Data

Initial Data Collection and Normalization

[Needs More Information]

Who are the source language producers?

[Needs More Information]

Annotations

Annotation process

[Needs More Information]

Who are the annotators?

[Needs More Information]

Personal and Sensitive Information

[Needs More Information]

Considerations for Using the Data

Social Impact of Dataset

[Needs More Information]

Discussion of Biases

[Needs More Information]

Other Known Limitations

[Needs More Information]

Additional Information

Dataset Curators

[Needs More Information]

Licensing Information

[Needs More Information]

Citation Information

Cite as:

@inproceedings{
thakur2021beir,
title={{BEIR}: A Heterogeneous Benchmark for Zero-shot Evaluation of Information Retrieval Models},
author={Nandan Thakur and Nils Reimers and Andreas R{\"u}ckl{\'e} and Abhishek Srivastava and Iryna Gurevych},
booktitle={Thirty-fifth Conference on Neural Information Processing Systems Datasets and Benchmarks Track (Round 2)},
year={2021},
url={https://openreview.net/forum?id=wCu6T5xFjeJ}
}