instruction
stringclasses
9 values
task
stringclasses
9 values
ds_name
stringclasses
35 values
content
stringlengths
1
2.3k
schema
stringclasses
61 values
anwser
stringlengths
14
3.65k
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
WikiNeural
Some actions against Orthodox priests and believers along with execution included torture , being sent to prison camps , labour camps or mental hospitals .
location, person, organization
{"entities": []}
Extract all relationships between entities from the given text and return them in JSON format with the key "relations". The output should be a JSON string containing a list of objects, each with "head", "relation", and "tail".
RE
semval-RE
the headmaster made the formal announcement .
component whole, NA, instrument agency, member collection, cause effect, entity destination, content container, message topic, product producer, entity origin
{"relations": [{"head": "headmaster", "relation": "product producer", "tail": "announcement"}]}
Extract all relationships between entities from the given text and return them in JSON format with the key "relations". The output should be a JSON string containing a list of objects, each with "head", "relation", and "tail".
RE
kbp37
International Business Machines Corporation or IBM is an American multinational technology and consulting corporation with headquarters in Armonk New York United States .
cities of residence, top members employees, members, alternate names, origin, state or provinces of residence, title of person, spouse, city of headquarters, NA, founded by, state or province of headquarters, employee of, countries of residence, country of birth, founded, subsidiaries, country of headquarters
{"relations": []}
Extract all arguments of the event from the given text and return them in JSON format with the key "arguments". The output should be a JSON string containing a list of objects, each with "name" and "role", based on the provided event type.
EEA
ACE05
They had planned to use the laser against Tokyo policemen , but the plan failed when the laser malfunctioned during the testing stage ."
victim, attacker, place, target, instrument
{"arguments": [{"name": "They", "role": "attacker"}, {"name": "Tokyo", "role": "place"}, {"name": "policemen", "role": "target"}, {"name": "laser", "role": "instrument"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
TweetNER7_sample_15000
watermelon sugar makes me feel like i belong on the beach in the 80s playing volleyball and jamming out with my friends . i love it S O much . thank u for this gem . {@Harry Styles.@} kiwi truly did walk . # HarryOnSNL
group, creative work, person, event, product, location, corporation
{"entities": [{"name": "watermelon sugar", "type": "corporation"}, {"name": "the beach", "type": "location"}, {"name": "playing volleyball and jamming out with my friends", "type": "event"}, {"name": "Harry Styles.", "type": "person"}]}
Extract all relationships between entities from the given text and return them in JSON format with the key "relations". The output should be a JSON string containing a list of objects, each with "head", "relation", and "tail".
RE
ADE_corpus_sample_15000
PURPOSE : To report two cases in which topical brimonidine resulted in apparent central nervous system depression and unresponsiveness in an infant .
adverse effect
{"relations": [{"head": "central nervous system depression", "relation": "adverse effect", "tail": "brimonidine"}]}
Extract all relationships between entities from the given text and return them in JSON format with the key "relations". The output should be a JSON string containing a list of objects, each with "head", "relation", and "tail".
RE
conll04_sample_5000
It seems to be more effective , ' ' said Dr. Alain Farley of McGill University in Montreal .
Organization based in, Located in, Live in, Work for, Kill
{"relations": [{"head": "Alain Farley", "relation": "Work for", "tail": "McGill University"}, {"head": "McGill University", "relation": "Organization based in", "tail": "Montreal"}]}
Extract all entity spans from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of entity names.
ES
ACE 2004
The Federal Aviation Administration ' s Web site , www . fly . faa . gov , has real - time information on operations at the biggest airports .
geographical social political, organization, person, location, facility, vehicle, weapon
{"entities": ["The Federal Aviation Administration", "the biggest airports"]}
Extract all relationships between entities from the given text and return them in JSON format with the key "relations". The output should be a JSON string containing a list of objects, each with "head", "relation", and "tail".
RE
ADE_corpus_sample_15000
Pemphigus vulgaris precipitated by glibenclamide therapy .
adverse effect
{"relations": [{"head": "Pemphigus vulgaris", "relation": "adverse effect", "tail": "glibenclamide"}]}
Extract all relationships between entities from the given text and return them in JSON format with the key "relations". The output should be a JSON string containing a list of objects, each with "head", "relation", and "tail".
RE
kbp37
Podravka 's origins are traced to the 1934 family business for fruit processing of the Wolf brothers in Koprivnica which was nationalised in 1947 .
cities of residence, top members employees, members, alternate names, origin, state or provinces of residence, title of person, spouse, city of headquarters, NA, founded by, state or province of headquarters, employee of, countries of residence, country of birth, founded, subsidiaries, country of headquarters
{"relations": [{"head": "Podravka", "relation": "city of headquarters", "tail": "Koprivnica"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
ncbi
About half of all missense mutations have been examined by in vitro expression analysis , and a significant correlation has been observed between residual PAH activity and disease phenotype .
Disease
{"entities": []}
Extract all entity spans from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of entity names.
ES
bc5cdr
Magnetic resonance imaging demonstrated bone marrow oedema in the painful bones .
Disease, Chemical
{"entities": ["bone marrow oedema"]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
WikiNeural
Although the ETFE is susceptible to punctures , these can be easily fixed with ETFE tape .
location, person, organization
{"entities": []}
Extract all entity pairs and their relationships from the given text and return them in JSON format with the key "relations". The output should be a JSON string containing a list of objects, each with "head", "relation", and "tail".
EPR
New-York-Times-RE
America is now Vietnam 's leading trading partner , and Intel has just announced the expansion of its factory near Ho Chi Minh City .
ethnicity, place lived, geographic distribution, company industry, country of administrative divisions, administrative division of country, location contains, person of company, profession, ethnicity of people, company shareholder among major shareholders, sports team of location, religion, neighborhood of, company major shareholders, place of death, nationality, children, company founders, company founded place, country of capital, company advisors, sports team location of teams, place of birth
{"relations": [{"head": "Vietnam", "relation": "location contains", "tail": "Ho Chi Minh City"}, {"head": "Ho Chi Minh City", "relation": "administrative division of country", "tail": "Vietnam"}, {"head": "Vietnam", "relation": "country of administrative divisions", "tail": "Ho Chi Minh City"}]}
Extract all entity spans from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of entity names.
ES
FindVehicle
Please find the red Sports car on the bottom-right of the picture and the pure white mini van on the upper-part
vehicle type, position of vehicle, color of vehicle, orientation of vehicle, vehicle velocity, vehicle range, brand of vehicle, vehicle model, SUV, hatchback, sports car, sedan, bus, truck, vintage car, coupe, roadster, estate car, motorcycle, MPV, van
{"entities": ["red", "Sports car", "bottom-right", "pure white", "mini van", "upper-part"]}
Extract all entity pairs and their relationships from the given text and return them in JSON format with the key "relations". The output should be a JSON string containing a list of objects, each with "head", "relation", and "tail".
EPR
conll04_sample_5000
The rig , leased by Chevron from Avis Bourg Co. of Folsom , was on its side in about 25 feet of water , a Coast Guard spokesman said.
Organization based in, Located in, Live in, Work for, Kill
{"relations": [{"head": "Avis Bourg Co.", "relation": "Organization based in", "tail": "Folsom"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
bc2gm
Mutational studies revealed that it was the homeodomain binding site II sequence that was required for this regulation .
GENE
{"entities": [{"name": "homeodomain binding site II sequence", "type": "GENE"}]}
Extract all entity spans from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of entity names.
ES
ncbi
We analyzed the HEXA gene of one pseudodeficient subject and identified both a C739 - to - T substitution that changes Arg247 - - - - Trp on one allele and a previously identified Tay - Sachs disease mutation on the second allele .
Disease
{"entities": ["Tay - Sachs disease"]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
WikiNeural
Before the creation of United Artists , Chaplin married for the first time .
location, person, organization
{"entities": [{"name": "United Artists", "type": "organization"}, {"name": "Chaplin", "type": "person"}]}
Extract all entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
ET
mit-restaurant
i need a good taco
Rating, Location, Amenity, Cuisine, Hours, Price, Dish, Restaurant Name
{"entities": [{"name": "good", "type": "Rating"}, {"name": "taco", "type": "Dish"}]}
Extract all entity spans from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of entity names.
ES
ACE 2004
President Fujimori appealed at the opening ceremony for the countries of the Andes to strive together to achieve local integration , in order to shake off the state of being undeveloped .
geographical social political, organization, person, location, facility, vehicle, weapon
{"entities": ["the countries of the Andes", "President", "President Fujimori", "the Andes"]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
Ontonotes_sample_30000
Irving from the UK 's Longman Publishing Company said that British publishers hope to further develop business with Chinese publishers on the basis of mutually respecting copyrights .
date, organization, person, geographical social political, national religious political, facility, cardinal, location, work of art, law, event, product, ordinal, percent, time, quantity, money, language
{"entities": [{"name": "Irving", "type": "person"}, {"name": "UK", "type": "geographical social political"}, {"name": "Longman Publishing Company", "type": "organization"}, {"name": "British", "type": "national religious political"}, {"name": "Chinese", "type": "national religious political"}]}
Extract all relationships between entities from the given text and return them in JSON format with the key "relations". The output should be a JSON string containing a list of objects, each with "head", "relation", and "tail".
RE
conll04_sample_5000
Stacy Lin Bickel , 25 , of Tempe pleaded guilty in November to negligent homicide in the death of Deborah Sue Tomazin , 24 , of Phoenix .
Organization based in, Located in, Live in, Work for, Kill
{"relations": [{"head": "Stacy Lin Bickel", "relation": "Live in", "tail": "Tempe"}, {"head": "Deborah Sue Tomazin", "relation": "Live in", "tail": "Phoenix"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
bc5cdr
Data on use of BZDs / RDs before admission , current medications and discharge diagnoses were collected from medical records .
Disease, Chemical
{"entities": [{"name": "BZDs", "type": "Chemical"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
ACE 2004_sample_15000
a bowl of cereal with some soy milk to put on it .
geographical social political, organization, person, location, facility, vehicle, weapon
{"entities": []}
Extract all entity pairs and their relationships from the given text and return them in JSON format with the key "relations". The output should be a JSON string containing a list of objects, each with "head", "relation", and "tail".
EPR
GIDS
Probably the Arthur Savage 1798-1852, naval surgeon, who as surgeon-intendent sailed with three convict ships to Australia between 1833 and 1836 and with an emigrant ship on which he and his family sailed in 1838. He had acted as an immigration agent in the UK before his departure. He served as health officer for Sydney from November 1839 until his death in New South Wales on 19/07/1852.
place of death, place of birth, education degree, education institution
{"relations": [{"head": "Arthur Savage ", "relation": "place of birth", "tail": "Sydney"}]}
Extract all events from the given text and return them in JSON format with the key "events". The output should be a JSON string containing a list of objects, each with "event_type", "trigger", and "arguments" (a list of objects with "name" and "role").
EE
PHEE
Pulmonary fibrosis associated with nabumetone.
Event type: ( adverse event,potential therapeutic event ) . Arguments type: ( Treatment.Disorder,Treatment.Time_elapsed,Treatment.Dosage,Subject.Race,Treatment.Drug,Treatment.Route,Subject.Disorder,Treatment.Freq,Treatment,Effect,Subject.Population,Treatment.Duration,Combination.Drug,Subject,Subject.Age,Subject.Gender )
{"events": [{"event_type": "adverse event", "trigger": "associated ", "arguments": [{"name": "Pulmonary fibrosis", "role": "Effect"}, {"name": "nabumetone", "role": "Treatment"}, {"name": "nabumetone", "role": "Treatment.Drug"}]}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
Broad Tweet Corpus
British Ebola victim Will Pooley taking experimental drug ZMapp http://t.co/XoBaDDjYkl
location, person, organization
{"entities": [{"name": "British", "type": "location"}, {"name": "Will Pooley", "type": "person"}]}
Extract all arguments of the event from the given text and return them in JSON format with the key "arguments". The output should be a JSON string containing a list of objects, each with "name" and "role", based on the provided event type.
EEA
PHEE
BACKGROUND: Gemcitabine has mild renal toxicity, but cases of gemcitabine-associated hemolytic-uremic syndrome (HUS) have been reported.
Treatment.Freq, Subject.Population, Combination.Drug, Subject.Gender, Treatment.Duration, Subject, Effect, Treatment, Treatment.Disorder, Subject.Disorder, Treatment.Route, Subject.Race, Treatment.Time_elapsed, Treatment.Dosage, Subject.Age, Treatment.Drug
{"arguments": [{"name": "cases", "role": "Subject"}, {"name": "hemolytic-uremic syndrome (HUS)", "role": "Effect"}, {"name": "renal toxicity", "role": "Effect"}, {"name": "gemcitabine", "role": "Treatment"}, {"name": "gemcitabine", "role": "Treatment.Drug"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
WikiANN en
2 - Bill Shoemaker ( 1959 , 1966 )
location, person, organization
{"entities": [{"name": "Bill Shoemaker", "type": "person"}]}
Extract all entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
ET
AnatEM
Three distinct motor pools were observed , which confirms the functional identity of each muscle .
Anatomy
{"entities": [{"name": "muscle", "type": "Anatomy"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
mit-movie
what is a four star crime r rated movie from the 1970 s
genre, year, plot, average ratings, actor, title, song, character, rating, review, director, trailer
{"entities": [{"name": "four star", "type": "average ratings"}, {"name": "crime", "type": "genre"}, {"name": "r", "type": "rating"}, {"name": "1970 s", "type": "year"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
Ontonotes_sample_30000
Canal bank enterprises are spread all over the place .
date, organization, person, geographical social political, national religious political, facility, cardinal, location, work of art, law, event, product, ordinal, percent, time, quantity, money, language
{"entities": []}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
mit-movie
is there a documentary about the youth sports
genre, year, plot, average ratings, actor, title, song, character, rating, review, director, trailer
{"entities": [{"name": "documentary", "type": "genre"}, {"name": "the youth sports", "type": "plot"}]}
Extract all entity pairs from the given text and return them in JSON format with the key "entity_pairs". The output should be a JSON string containing a list of objects, each with "head" and "tail".
EP
GIDS
Severne was the son of John Edmund Severne Michael Severne of Wallop Hall, Shrewsbury and his wife Anna Maria Wigley daughter of Edmund Wigley of Shakenhurst Worcestershire and MP for Worcester. He was educated at Brasenose College, Oxford where he rowed for the Brasenose eight in the university eights in 1846 and graduated in 1848. He joined the 10th Hussars in 1849 and became a lieutenant in 1852. He transferred to the 16th Lancers and retired as a captain in 1857. He was a J.P. for Montgomeryshire and was High Sheriff of Montgomeryshire in 1854. He was also a JP and Deputy Lieutenant of Shropshire and Northamptonshire and was High Sheriff of Northamptonshire in 1861.
place of death, place of birth, education degree, education institution
{"entity_pairs": [{"head": "John Edmund Severne", "tail": "Brasenose College, Oxford"}]}
Extract all entity spans from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of entity names.
ES
GENIA_NER
Site-directed mutational studies demonstrate that Site I and Site IV together support ATF1- and CREB- induced trans-activation of the H4 promoter .
cell line, protein, RNA, DNA, cell type
{"entities": ["Site I", "Site IV", "H4 promoter"]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
ACE 2004_sample_15000
At least five people were reported killed in Alabama , thirty injured when a series of twisters struck parts of the rural State .
geographical social political, organization, person, location, facility, vehicle, weapon
{"entities": [{"name": "Alabama", "type": "geographical social political"}, {"name": "the rural State", "type": "geographical social political"}, {"name": "At least five people were reported killed in Alabama", "type": "person"}, {"name": "thirty", "type": "person"}, {"name": "parts of the rural State", "type": "location"}]}
Extract all relationships between entities from the given text and return them in JSON format with the key "relations". The output should be a JSON string containing a list of objects, each with "head", "relation", and "tail".
RE
NYT11_sample_30000
How much can the system support increased autonomy and authority of school leaders without making the commensurate changes with respect to the external demands , '' asked Anthony Lombardi , the principal of P.S. 49 in Middle Village , Queens . ''
nationality, country capital, place of death, children, location contains, place of birth, place lived, administrative division of country, country of administrative divisions, company, neighborhood of, company founders
{"relations": [{"head": "Middle Village", "relation": "neighborhood of", "tail": "Queens"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
WikiNeural
This industry appears to have spread throughout the coastal regions of the Maghreb between 15,000 and 10,000 BC .
location, person, organization
{"entities": [{"name": "Maghreb", "type": "location"}]}
Extract all entity pairs from the given text and return them in JSON format with the key "entity_pairs". The output should be a JSON string containing a list of objects, each with "head" and "tail".
EP
New-York-Times-RE
Boxing regulators have criticized the safety panel because every member has links to the boxing commissioners in Nevada or has political ties to the state 's Republican governor , Kenny Guinn , who appoints the commissioners .
ethnicity, place lived, geographic distribution, company industry, country of administrative divisions, administrative division of country, location contains, person of company, profession, ethnicity of people, company shareholder among major shareholders, sports team of location, religion, neighborhood of, company major shareholders, place of death, nationality, children, company founders, company founded place, country of capital, company advisors, sports team location of teams, place of birth
{"entity_pairs": [{"head": "Kenny Guinn", "tail": "Nevada"}]}
Extract all event types and their triggers from the given text and return them in JSON format with the key "events". The output should be a JSON string containing a list of objects, each with "event_type" and "trigger".
EET
CASIE
" Anyone who views a specially crafted profile gets popped , " a white hat hacker who has found several bugs in Steam in the past , and asked to remain anonymous , told me in a Twitter DM .
phishing, data breach, ransom, discover vulnerability, patch vulnerability
{"events": [{"event_type": "discover vulnerability", "trigger": "has found"}]}
Extract all events from the given text and return them in JSON format with the key "events". The output should be a JSON string containing a list of objects, each with "event_type", "trigger", and "arguments" (a list of objects with "name" and "role").
EE
GENIA
Northern analysis using a probe encompassing a common region of both isoforms showed two mRNA species of 2.7 and 4.5 kilobase pairs , while an NFATc.beta -specific probe detected only the 4.5-kilobase pair mRNA which was preferentially expressed in the spleen.
Event type: ( RNA, cell type, protein, cell line, DNA ) . Arguments type: ( None )
{"events": [{"event_type": "RNA", "trigger": "mRNA species", "arguments": []}, {"event_type": "RNA", "trigger": "4.5-kilobase pair mRNA", "arguments": []}, {"event_type": "DNA", "trigger": "NFATc.beta -specific probe", "arguments": []}, {"event_type": "protein", "trigger": "NFATc.beta", "arguments": []}]}
Extract all event types and their triggers from the given text and return them in JSON format with the key "events". The output should be a JSON string containing a list of objects, each with "event_type" and "trigger".
EET
PHEE
He was diagnosed with possible serotonin syndrome; his symptoms resolved after clomipramine was stopped but before clozapine was restarted eight days later.
adverse event, potential therapeutic event
{"events": [{"event_type": "adverse event", "trigger": "resolved "}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
bc5cdr
We therefore systematically analyzed the functional impact of the most prevalent immune escape variants , the sG145R and sP120T mutants , on the viral replication efficacy and antiviral drug susceptibility of common treatment - associated mutants with resistance to lamivudine ( LAM ) and / or HBeAg negativity .
Disease, Chemical
{"entities": [{"name": "lamivudine", "type": "Chemical"}, {"name": "LAM", "type": "Chemical"}, {"name": "HBeAg", "type": "Chemical"}]}
Extract all relationships between entities from the given text and return them in JSON format with the key "relations". The output should be a JSON string containing a list of objects, each with "head", "relation", and "tail".
RE
ADE_corpus_sample_15000
Gold - induced aplastic anemia .
adverse effect
{"relations": [{"head": "aplastic anemia", "relation": "adverse effect", "tail": "Gold"}]}
Extract all event types and their triggers from the given text and return them in JSON format with the key "events". The output should be a JSON string containing a list of objects, each with "event_type" and "trigger".
EET
ACE05
there ' s a concern with the huge series there this weekend with the yankees coming to town that sammy would be suspended during that .
declare bankruptcy, transfer ownership, marry, die, phone write, arrest jail, convict, sue, transfer money, end organization, start organization, end position, merge organization, start position, transport, charge indict, elect, attack, injure, born, extradite, release parole, sentence, demonstrate, meet, trial hearing, divorce, nominate, fine, appeal, pardon, execute, acquit
{"events": [{"event_type": "transport", "trigger": "coming"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
FabNER
A qualitative analysis of fiber orientation conducted on the SCF filament indicated that the SCFs maintain their orientation from filament to part except in the intersection zone of rasters .
manufacturing standard, machanical property, manufacturing process, material, engineering feature, process parameter, machine or equipment, application, process characterization, enabling technology, biomedical, concept or principle
{"entities": [{"name": "qualitative", "type": "concept or principle"}, {"name": "fiber orientation", "type": "engineering feature"}, {"name": "filament", "type": "material"}, {"name": "orientation", "type": "concept or principle"}, {"name": "filament", "type": "material"}]}
Extract all events from the given text and return them in JSON format with the key "events". The output should be a JSON string containing a list of objects, each with "event_type", "trigger", and "arguments" (a list of objects with "name" and "role").
EE
GENIA
In AML , chromosomal aberrations result in oncoproteins such as AML1/ETO , PML/RARalpha , or activated Ras , which can deregulate genes important for normal myelopoiesis .
Event type: ( RNA, cell type, protein, cell line, DNA ) . Arguments type: ( None )
{"events": [{"event_type": "protein", "trigger": "activated Ras", "arguments": []}, {"event_type": "protein", "trigger": "PML/RARalpha", "arguments": []}, {"event_type": "protein", "trigger": "AML1/ETO", "arguments": []}, {"event_type": "protein", "trigger": "oncoproteins", "arguments": []}]}
Extract all events from the given text and return them in JSON format with the key "events". The output should be a JSON string containing a list of objects, each with "event_type", "trigger", and "arguments" (a list of objects with "name" and "role").
EE
GENIA
In vitro studies on hematopoietic control mechanisms have been hampered by the heterogeneity of the analyzed cell populations, ie, lack of lineage specificity and developmental stage homogeneity of progenitor/precursor cells growing in culture.
Event type: ( RNA, cell type, protein, cell line, DNA ) . Arguments type: ( None )
{"events": [{"event_type": "cell type", "trigger": "progenitor/precursor cells", "arguments": []}]}
Extract all entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
ET
MultiNERD
They would lose to Atlético Nacional on penalties in the third place game .
media, plant, vehicle, astronomical object, time, disease, animal, location, person, organization, mythical figure, food, biology, geographical phenomenon, instrument, event
{"entities": [{"name": "Atl\u00e9tico Nacional", "type": "organization"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
ACE 2005_sample_15000
No matter what the MSM tries to tell you , there are Iraqi ' s who will help the allies , who will fight alongside them and put their lives at risk to make sure Democracy succeeds in their country .
organization, person, geographical social political, vehicle, location, weapon, facility
{"entities": [{"name": "their country", "type": "geographical social political"}, {"name": "the MSM", "type": "organization"}, {"name": "you", "type": "person"}, {"name": "Iraqi ' s who will help the allies", "type": "person"}, {"name": "who", "type": "person"}, {"name": "the allies", "type": "person"}, {"name": "who", "type": "person"}, {"name": "them", "type": "person"}, {"name": "their", "type": "person"}, {"name": "their", "type": "person"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
Broad Tweet Corpus
I support # wikipediablackout ! Show your support here http://t.co/j1n0Fn9d
location, person, organization
{"entities": []}
Extract all entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
ET
mit-restaurant
what is a good family restaurant in my area
Rating, Location, Amenity, Cuisine, Hours, Price, Dish, Restaurant Name
{"entities": [{"name": "good", "type": "Rating"}, {"name": "family", "type": "Cuisine"}, {"name": "restaurant", "type": "Amenity"}, {"name": "in my area", "type": "Location"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
AnatEM
Isotype - specific inhibitors of the glycolytic key regulator pyruvate kinase subtype M2 moderately decelerate tumor cell proliferation .
Anatomy
{"entities": [{"name": "tumor cell", "type": "Anatomy"}]}
Extract all entity pairs from the given text and return them in JSON format with the key "entity_pairs". The output should be a JSON string containing a list of objects, each with "head" and "tail".
EP
New-York-Times-RE
It 's vital that New Jersey and New York , with the help of the federal government and the City of New York , focus on moving highly skilled employees into and out of Manhattan and maximizing the region 's economic potential .
ethnicity, place lived, geographic distribution, company industry, country of administrative divisions, administrative division of country, location contains, person of company, profession, ethnicity of people, company shareholder among major shareholders, sports team of location, religion, neighborhood of, company major shareholders, place of death, nationality, children, company founders, company founded place, country of capital, company advisors, sports team location of teams, place of birth
{"entity_pairs": [{"head": "New York", "tail": "Manhattan"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
ncbi
Genetic instability in human ovarian cancer cell lines .
Disease
{"entities": [{"name": "ovarian cancer", "type": "Disease"}]}
Extract all event types and their triggers from the given text and return them in JSON format with the key "events". The output should be a JSON string containing a list of objects, each with "event_type" and "trigger".
EET
CASIE
The bug was found by researchers Filippo Cavallarin and Dawid Golunski , independently of one another , and affects SquirrelMail versions 1.4.22 and below .
phishing, data breach, ransom, discover vulnerability, patch vulnerability
{"events": [{"event_type": "discover vulnerability", "trigger": "was found"}]}
Extract all entity spans from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of entity names.
ES
bc5cdr
In both instances , suspicion of the presence of carcinoma was aroused by the palpation of a small nodule in the vaginal fornix .
Disease, Chemical
{"entities": ["carcinoma"]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
Ontonotes_sample_30000
From Asia , he has mink jackets with floral patterns made by using different colored furs .
date, organization, person, geographical social political, national religious political, facility, cardinal, location, work of art, law, event, product, ordinal, percent, time, quantity, money, language
{"entities": [{"name": "Asia", "type": "location"}]}
Extract all relationships between entities from the given text and return them in JSON format with the key "relations". The output should be a JSON string containing a list of objects, each with "head", "relation", and "tail".
RE
conll04_sample_5000
` ` From the onset , we at the Texas Department of Agriculture felt that more was at stake than steak , and a whole lot more was at stake than growth hormones , ' ' said Texas Agriculture Commissioner Jim Hightower at a news conference Monday.
Organization based in, Located in, Live in, Work for, Kill
{"relations": [{"head": "Texas Department of Agriculture", "relation": "Organization based in", "tail": "Texas"}, {"head": "Jim Hightower", "relation": "Work for", "tail": "Texas Department of Agriculture"}, {"head": "Jim Hightower", "relation": "Live in", "tail": "Texas"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
TweetNER7_sample_15000
someone said enhypen is running , not walking . seriously the power they hold URGHH # BORDER _ DAY _ ONE # Intro _ Walk _ the _ Line {@ENHYPEN@}
group, creative work, person, event, product, location, corporation
{"entities": [{"name": "the", "type": "creative work"}, {"name": "BORDER", "type": "creative work"}, {"name": "DAY", "type": "creative work"}, {"name": "ONE", "type": "creative work"}, {"name": "Intro", "type": "creative work"}, {"name": "Walk", "type": "creative work"}, {"name": "the", "type": "creative work"}, {"name": "Line", "type": "creative work"}, {"name": "ENHYPEN", "type": "group"}]}
Extract all events from the given text and return them in JSON format with the key "events". The output should be a JSON string containing a list of objects, each with "event_type", "trigger", and "arguments" (a list of objects with "name" and "role").
EE
ACE05
I ' d like to -- I know that , you know , there ' s been problems , a lot of conflict over there , and it seems like no matter what ,
Event type: ( declare bankruptcy, transfer ownership, transfer money, marry, transport, die, phone write, arrest jail, convict, sentence, sue, end organization, start organization, end position, start position, meet, elect, attack, injure, born, fine, release parole, charge indict, extradite, trial hearing, demonstrate, divorce, nominate, appeal, pardon, execute, acquit, merge organization ) . Arguments type: ( beneficiary, prosecutor, plaintiff, seller, victim, buyer, agent, destination, artifact, organization, recipient, person, place, origin, target, giver, adjudicator, attacker, vehicle, defendant, entity, instrument )
{"events": [{"event_type": "attack", "trigger": "conflict", "arguments": [{"name": "there", "role": "place"}]}]}
Extract all relationships between entities from the given text and return them in JSON format with the key "relations". The output should be a JSON string containing a list of objects, each with "head", "relation", and "tail".
RE
SciERC_sample_10000
We describe a new method for the representation of NLP structures within reranking approaches .
conjunction, feature of, hyponym of, used for, part of, compare, evaluate for
{"relations": [{"head": "method", "relation": "used for", "tail": "NLP structures"}, {"head": "reranking approaches", "relation": "feature of", "tail": "NLP structures"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
mit-restaurant
try to find a kid friendly coffee shop that opens before 8 am
Rating, Location, Amenity, Cuisine, Hours, Price, Dish, Restaurant Name
{"entities": [{"name": "kid friendly", "type": "Amenity"}, {"name": "coffee", "type": "Cuisine"}, {"name": "opens before 8 am", "type": "Hours"}]}
Extract all entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
ET
WikiANN en
At this time , ''Finland '' referred to the south-western part of what today is Finland .
location, person, organization
{"entities": [{"name": "Finland", "type": "location"}]}
Extract all entity spans from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of entity names.
ES
MultiNERD
He attended the University of New Orleans ( 1976 – 1979 ) .
media, plant, vehicle, astronomical object, time, disease, animal, location, person, organization, mythical figure, food, biology, geographical phenomenon, instrument, event
{"entities": ["University of New Orleans"]}
Extract all entity spans from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of entity names.
ES
bc5cdr
The maximal NGF contents obtained by PG - 9 were 17 . 6 - fold of the control value .
Disease, Chemical
{"entities": ["PG - 9"]}
Extract all entity spans from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of entity names.
ES
CoNLL 2003_sample_15000
LONDON 1996-08-22
location, else, organization, person
{"entities": ["LONDON"]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
bc4chemd
Magnetic resonance imaging ( MRI ) of both tumors showed that PPTT was capable of improving the tumor mass penetration of HPMA copolymers .
Chemical
{"entities": [{"name": "HPMA", "type": "Chemical"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
ACE 2005_sample_15000
it says - - is that possible ?
organization, person, geographical social political, vehicle, location, weapon, facility
{"entities": []}
Extract all entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
ET
bc5cdr
5 - azacytidine potentiates initiation induced by carcinogens in rat liver .
Disease, Chemical
{"entities": [{"name": "5 - azacytidine", "type": "Chemical"}, {"name": "initiation induced by carcinogens", "type": "Disease"}]}
Extract all entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
ET
bc5cdr
Further observations on the electrophysiologic effects of oral amiodarone therapy .
Disease, Chemical
{"entities": [{"name": "amiodarone", "type": "Chemical"}]}
Extract all entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
ET
PolyglotNER
Stromberg 's credits include films such as Avatar , Alice in Wonderland and Oz the Great and Powerful , the first two earning him Academy Award wins for Best Art Direction .
location, person, organization
{"entities": [{"name": "Stromberg", "type": "person"}]}
Extract all arguments of the event from the given text and return them in JSON format with the key "arguments". The output should be a JSON string containing a list of objects, each with "name" and "role", based on the provided event type.
EEA
CASIE
V2 of the same router contains the same vulnerability , but can not be exploited without authentication , he said .
vulnerable system owner, vulnerability, capabilities, time, vulnerable system version, discoverer, common vulnerabilities and exposures, supported platform, vulnerable system
{"arguments": [{"name": "router", "role": "vulnerable system"}, {"name": "the same vulnerability", "role": "vulnerability"}, {"name": "V2", "role": "vulnerable system version"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
mit-restaurant
hii want today just a salad
Rating, Location, Amenity, Cuisine, Hours, Price, Dish, Restaurant Name
{"entities": [{"name": "salad", "type": "Dish"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
mit-movie
what michael dorn r rated adventure movie was liked a lot
genre, year, plot, average ratings, actor, title, song, character, rating, review, director, trailer
{"entities": [{"name": "michael dorn", "type": "actor"}, {"name": "r", "type": "rating"}, {"name": "adventure", "type": "genre"}, {"name": "was liked a lot", "type": "average ratings"}]}
Extract all entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
ET
bc2gm
Recombinant I - kappa B alpha inhibited kappa B motif binding by nuclear factor - kappa B1 , RelA , and c - Rel as indicated by studies using UV radiation - induced covalent cross - linking to a bromodeoxyuridine - substituted kappa B oligonucleotide .
GENE
{"entities": [{"name": "Recombinant I - kappa B alpha", "type": "GENE"}, {"name": "kappa B motif", "type": "GENE"}, {"name": "nuclear factor - kappa B1", "type": "GENE"}, {"name": "RelA", "type": "GENE"}, {"name": "c - Rel", "type": "GENE"}, {"name": "kappa B oligonucleotide", "type": "GENE"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
WikiNeural
Today it is owned by the Italian government , and operated by the municipality .
location, person, organization
{"entities": []}
Extract all entity spans from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of entity names.
ES
Broad Tweet Corpus
Taking heart attack victims to hospital ‘ wastes valuable time ’ http://t.co/YBpYRcgtFN ( Getty ) http://t.co/NeKvF7ygHT
location, person, organization
{"entities": ["Getty"]}
Extract all entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
ET
bc2gm
This may result in more reabsorption and hence reduced renal clearance .
GENE
{"entities": []}
Extract all arguments of the event from the given text and return them in JSON format with the key "arguments". The output should be a JSON string containing a list of objects, each with "name" and "role", based on the provided event type.
EEA
PHEE
The toxic effects of methotrexate included elevated liver transaminases (3/4), nausea (2/4), abdominal pain (2/4), bone pain (2/4), mild neutropenia (1/4), and mild pruritus (1/4).
Treatment.Freq, Subject.Population, Combination.Drug, Subject.Gender, Treatment.Duration, Subject, Effect, Treatment, Treatment.Disorder, Subject.Disorder, Treatment.Route, Subject.Race, Treatment.Time_elapsed, Treatment.Dosage, Subject.Age, Treatment.Drug
{"arguments": [{"name": "toxic effects", "role": "Effect"}, {"name": "elevated liver transaminases (3/4), nausea (2/4), abdominal pain (2/4), bone pain (2/4), mild neutropenia (1/4), and mild pruritus (1/4)", "role": "Effect"}, {"name": "methotrexate", "role": "Treatment"}, {"name": "methotrexate", "role": "Treatment.Drug"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
bc2gm
Thirty Class I and Class II recessions in 30 patients were treated with a modified subepithelial connective tissue graft procedure .
GENE
{"entities": []}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
bc5cdr
During the first day of hospitalization ( while intubated ) , intravenous metoprolol was given , resulting in severe angioedema .
Disease, Chemical
{"entities": [{"name": "metoprolol", "type": "Chemical"}, {"name": "angioedema", "type": "Disease"}]}
Extract all relationships between entities from the given text and return them in JSON format with the key "relations". The output should be a JSON string containing a list of objects, each with "head", "relation", and "tail".
RE
conll04_sample_5000
ROCK HILL , S.C .
Organization based in, Located in, Live in, Work for, Kill
{"relations": [{"head": "ROCK HILL", "relation": "Located in", "tail": "S.C"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
Ontonotes_sample_30000
The talks were proposed by the Palestinians , who say they 're ready to explore all possibilities .
date, organization, person, geographical social political, national religious political, facility, cardinal, location, work of art, law, event, product, ordinal, percent, time, quantity, money, language
{"entities": [{"name": "Palestinians", "type": "national religious political"}]}
Extract all relationships between entities from the given text and return them in JSON format with the key "relations". The output should be a JSON string containing a list of objects, each with "head", "relation", and "tail".
RE
NYT11_sample_30000
Steve Levy , the Democratic county executive in Suffolk County , sent police officers to help shut down single-family houses crammed with 40 and 50 immigrant workers in Farmingville , a hamlet that has become synonymous with the conflict between day laborers and suburban homeowners .
nationality, country capital, place of death, children, location contains, place of birth, place lived, administrative division of country, country of administrative divisions, company, neighborhood of, company founders
{"relations": [{"head": "Suffolk County", "relation": "location contains", "tail": "Farmingville"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
GENIA_NER
In the present studies, rat lungs were depleted of alveolar macrophages by airway instillation of liposome-encapsulated dichloromethylene diphosphonate .
cell line, protein, RNA, DNA, cell type
{"entities": [{"name": "alveolar macrophages", "type": "cell type"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
ncbi
By use of chemical cleavage of mismatches adapted to fluorescence - based detection systems , we have characterized the mutations underlying alpha - Gal A deficiency in 16 individuals from six unrelated families with FD .
Disease
{"entities": [{"name": "alpha - Gal A deficiency", "type": "Disease"}, {"name": "FD", "type": "Disease"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
ACE 2005_sample_15000
CARLSON From the right , I ' m Tucker Carlson .
organization, person, geographical social political, vehicle, location, weapon, facility
{"entities": [{"name": "I", "type": "person"}, {"name": "Tucker Carlson", "type": "person"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
PolyglotNER
Leiblin Park , Nova Scotia is a residential neighbourhood in Halifax on the Mainland Halifax within the Halifax Regional Municipality Nova Scotia .
location, person, organization
{"entities": [{"name": "Nova Scotia", "type": "location"}, {"name": "Halifax", "type": "location"}, {"name": "Mainland Halifax", "type": "location"}, {"name": "Halifax Regional Municipality Nova Scotia", "type": "location"}]}
Extract all relationships between entities from the given text and return them in JSON format with the key "relations". The output should be a JSON string containing a list of objects, each with "head", "relation", and "tail".
RE
NYT11_sample_30000
Though use of natural gas is growing Germany and Russia are jointly building a pipeline under the Baltic Sea last winter 's cut-off left a bad taste in Berlin . ''
nationality, country capital, place of death, children, location contains, place of birth, place lived, administrative division of country, country of administrative divisions, company, neighborhood of, company founders
{"relations": [{"head": "Berlin", "relation": "administrative division of country", "tail": "Germany"}, {"head": "Germany", "relation": "country of administrative divisions", "tail": "Berlin"}]}
Extract all relationships between entities from the given text and return them in JSON format with the key "relations". The output should be a JSON string containing a list of objects, each with "head", "relation", and "tail".
RE
NYT11_sample_30000
Ms. Rice , who had previously declared that Belarus was the last dictatorship in Europe , warned bluntly on Thursday that the country should not conduct a '' sham election '' next year because its conduct would be '' watched by the international community , '' much as Ukraine 's election last year was watched and deemed fraudulent , helping to lead to a revolution .
nationality, country capital, place of death, children, location contains, place of birth, place lived, administrative division of country, country of administrative divisions, company, neighborhood of, company founders
{"relations": [{"head": "Europe", "relation": "location contains", "tail": "Belarus"}]}
Extract all entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
ET
FabNER
Additive manufacturing methods present opportunities for structures to have tailored mechanical anisotropy by integrating controlled lattice structures into their design .
manufacturing standard, machanical property, manufacturing process, material, engineering feature, process parameter, machine or equipment, application, process characterization, enabling technology, biomedical, concept or principle
{"entities": [{"name": "Additive manufacturing", "type": "manufacturing process"}, {"name": "mechanical anisotropy", "type": "machanical property"}, {"name": "lattice structures", "type": "engineering feature"}, {"name": "design", "type": "engineering feature"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
bc5cdr
Effects of tetrandrine and fangchinoline on experimental thrombosis in mice and human platelet aggregation .
Disease, Chemical
{"entities": [{"name": "tetrandrine", "type": "Chemical"}, {"name": "fangchinoline", "type": "Chemical"}, {"name": "thrombosis", "type": "Disease"}, {"name": "platelet aggregation", "type": "Disease"}]}
Extract all relationships between entities from the given text and return them in JSON format with the key "relations". The output should be a JSON string containing a list of objects, each with "head", "relation", and "tail".
RE
SciERC_sample_10000
The TAP-XL Automated Analyst 's Assistant is an application designed to help an English - speaking analyst write a topical report , culling information from a large inflow of multilingual , multimedia data .
conjunction, feature of, hyponym of, used for, part of, compare, evaluate for
{"relations": [{"head": "multilingual , multimedia data", "relation": "used for", "tail": "TAP-XL Automated Analyst 's Assistant"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
mit-restaurant
where is the chucky cheese in topeka
Rating, Location, Amenity, Cuisine, Hours, Price, Dish, Restaurant Name
{"entities": [{"name": "chucky cheese", "type": "Restaurant Name"}, {"name": "topeka", "type": "Location"}]}
Extract all entity spans from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of entity names.
ES
WikiNeural
Andorran gastronomy is mainly Catalan , although it has also adopted other elements of French and Italian cuisines .
location, person, organization
{"entities": []}
Extract all entity pairs and their relationships from the given text and return them in JSON format with the key "relations". The output should be a JSON string containing a list of objects, each with "head", "relation", and "tail".
EPR
NYT11_sample_30000
Problems with Chicago 's and Cook County 's new voting system installed to solve problems with the old punch-card method also left a hard-fought county race up in the air for nearly a day after the polls closed .
nationality, country capital, place of death, children, location contains, place of birth, place lived, administrative division of country, country of administrative divisions, company, neighborhood of, company founders
{"relations": [{"head": "Cook County", "relation": "location contains", "tail": "Chicago"}]}
Extract all named entities and their types from the given text and return them in JSON format with the key "entities". The output should be a JSON string containing a list of objects, each with "name" and "type".
NER
WikiANN en
A Few Days with Me ''
location, person, organization
{"entities": [{"name": "A Few Days with Me", "type": "organization"}]}
Extract all relationships between entities from the given text and return them in JSON format with the key "relations". The output should be a JSON string containing a list of objects, each with "head", "relation", and "tail".
RE
semval-RE
in particular , a ratpack of colonels in arrested development are the worst possible people to handle relations with latin countries .
component whole, NA, instrument agency, member collection, cause effect, entity destination, content container, message topic, product producer, entity origin
{"relations": [{"head": "ratpack", "relation": "member collection", "tail": "colonels"}]}