own-pt / rte-sick

RTE Experiment
1 stars 3 forks source link

noun-nouns in SICK #27

Closed vcvpaiva closed 7 years ago

vcvpaiva commented 7 years ago

some noun noun compounds:

there are too many to verify all 1618 1168 straw hat shopping cart cardboard sign dirt ramp ? fishing rod bedroom slipper store counter pork chop football player holiday gift bag ice hockey goalkeeper climbing equipment soccer ball goal net cowboy hat rope bridge rain puddle tennis ball plastic sword rodeo bull park bench guitar case stone path bike rider jumper house toy girl riding car drag race video game bamboo structure bird cage baby rhino wind instrument Christmas reindeer headband toy car dirt bikers

vcvpaiva commented 7 years ago

we miss some NNs too e.g. skateboard trick, park porch? in A woman is doing a skateboard trick on an outdoor park porch

vcvpaiva commented 7 years ago

most of noun-nouns seem ok BUT:

  1. There is no deer jumping fences deer jumping fences considered nn, when jumping is verb
  2. dog fighting? (in a black dog and a tan dog fighting)
  3. ping pong as NN?
  4. back yard as NN?
  5. orange rider in The orange rider is not driving a motorcycle on one wheel
  6. panda dog??
  7. leafless tree ??
  8. german shepherd dog
vcvpaiva commented 7 years ago

query VERB <nn _ gives interesting results: wrong ones:

  1. deer jumping fences
  2. mowing grass
  3. sprinkling cheese
  4. There is no kitten drinking milk (kitten drinking milk is not noun-noun, but it could be)
  5. There is no one sitting in lawn chairs and reading books (reading books is a nn, not here)
  6. a soccer player sits on the field drinking water (field drinking water not nn here)

Correct ones: both mwes and not

  1. tire rolling race
  2. marching band
  3. vending machine
  4. paddling pool
  5. rock climbing wall
  6. ice skating rink
  7. ice skating park
  8. playing house

UPDATE: some of the nns from before disappeared, only kept the 24 from the comment below, e.g. paddling pool, vending machine, rock climbing wall, tire rolling race.

vcvpaiva commented 7 years ago

the converse query VERB <nn _ produces some 50 results, mostly wrong? 1.There is no man dancing (man dancing is not a nn)

  1. There is no young boy covered in grass jumping near a wooden fence . (grass jumping is not nn)
  2. There are no men sawing (men sawing not nn)
  3. a young girl in a bikini jumping on the beach (bikini jumping is not nn)
  4. A man is rock climbing , pausing and calculating the route (rock climbing should be a single word?)
  5. A man does bike jumps in the dark in an empty pool (bike jumps should be nn, but jumps is noun, not verb then)
  6. There is no band playing (band playing is not nn)

exception: There is no panda bear eating some bamboo (panda bear is a nn!)

UPDATE: only 24 cases now.

  1. "tire rolling race" in Two women are competing in a tire rolling race is a real triple noun compound, as is "tire rolling competition".
  2. "man sprinkling cheese" is a mistake, sprinkling is verb not noun.
  3. "paddling pool" is nn. 5
  4. "kick boxing" is not nn in the sentence
  5. "dancing people" is not nn, dancing is adj
  6. "man mowing grass" is not nn, mowing is verb
  7. "deer jumping fences" is not nn, jumping is verb
  8. "bike riding people" is not nn, riding is verb
  9. "jumping dirt ramps" is not nn, jumping is verb
  10. "vending machine" is nn 3
  11. "sprinkling seasoning" is not nn, sprinkling is verb
  12. "hiking area" is not nn in this sentence A distant person with a blue backpack is hiking area full of rocks bad sentence missing prep
  13. "rock climbing wall' is nn
  14. "playing guitar" is not a nn in A white man is staging a hat and a playing guitar sentence is ungrammatical
  15. "climbing boy" is not a nn, climbing is adj?
vcvpaiva commented 7 years ago

there are only 15 results to query ADJ <nn _

orange juice is a true nn. orange shirt is not. back pack should be a single word?

1.The man is doing a magic trick (magic is ADJ, not noun?)

  1. There is no rhino grazing in a field (rhino is not ADJ)
  2. Three people are driving four wheel vehicles in a field
  3. a boy is sitting in a room playing a piano by lamp light (lamp light is nn, but lamp is considered adj?) 5.The orange rider is driving a motorcycle on one wheel (orange rider, orange is adj?)
  4. A trick cyclist takes air (?trick cyclist??)

update: there are only 7 cases now.

  1. Number 2. above is still here (twice rhino grazing) There is no rhino grazing in a field
  2. Number 3. above is still here Three people are driving four wheel vehicles in a field
  3. Number 5. still here The orange rider is driving a motorcycle on one wheel
  4. 2 cases of "orange shirt" and one of "back pack"

only "back pack" is a nn, but it's a compound? the other cases are mistakes, orange is adjective, "grazing" is not a noun, but verb, four-wheel vehicles is also a compound?

vcvpaiva commented 7 years ago

there are only 9 results to query _ <nn ADJ

  1. spiked knuckle graphic (graphic is noun, not adj?) 3 times
  2. bike airborne is not a nn
  3. A man by a wall made of bricks is wearing a mask around his mouth and a hair net, (hair net is nn) 3 times
  4. roof top is nn
  5. A bubble blew , dyeing the girl 's shirt red (shirt red is not nn)

update: after sentence normalization have 6 sentences.

number 2. above is still here There is no cyclist on a yellow bike airborne "bike airborne" is not a noun-noun compound, airborne is adjective.

number 5. above is still here A bubble blew , dyeing the girl 's shirt red (shirt red is not nn)

the other 7 cases disappeared, but we now have 3 cases of trunk open as noun-noun: A man is standing at the wheel of a classic American car that has its door and trunk open and The girl is painting a coverall blue (coverall blue is not nn)

vcvpaiva commented 7 years ago

query:_ <nn ADV only a man stands at the wheel of a classic american car , door and trunk open (trunk open is not nn)

update: after normalization no results for the query, yay! also no matches for ADV <nn _

vcvpaiva commented 7 years ago

the numbers and examples above are before the normalized sentences, I believe. after normalization we have 1148 noun-nouns, instead of 1618. redone the work

vcvpaiva commented 7 years ago

closing this to keep working on #47