src/Gargantext/Ngrams/Lang/En.hs

   1 {-# LANGUAGE OverloadedStrings #-}
   2 {-# LANGUAGE NoImplicitPrelude #-}
   3
   4 module Gargantext.Ngrams.Lang.En (selectNgrams, groupNgrams, textTest) where
   5
   6 import Gargantext.Prelude
   7 import Data.Text (Text)
   8 import Data.Monoid ((<>))
   9
  10 selectNgrams :: [(Text, Text, Text)] -> [(Text, Text, Text)]
  11 selectNgrams xs = filter selectNgrams' xs
  12     where
  13         selectNgrams' (_,"NN"   ,_             ) = True
  14         selectNgrams' (_,"NNS"  ,_             ) = True
  15         selectNgrams' (_,"NNP"  ,_             ) = True
  16         selectNgrams' (_,"NN+CC",_             ) = True
  17         selectNgrams' (_,_      ,"PERSON"      ) = True
  18         selectNgrams' (_,_      ,"ORGANIZATION") = True
  19         selectNgrams' (_,_      ,"LOCATION"    ) = True
  20         selectNgrams' (_,_      ,_             ) = False
  21
  22
  23 groupNgrams :: [(Text, Text, Text)] -> [(Text, Text, Text)]
  24 groupNgrams []       = []
  25
  26 groupNgrams ((j1,"JJ",j1'):(c1,"CC",c1'):(j2,"JJ",j2'):(j3,"JJ",_):xs)      = groupNgrams (jn1:cc:jn2:xs)
  27     where
  28         jn j' j'' jn' = (j' <> " " <> j'', "JJ", jn')
  29         cc  = (c1,"CC",c1')
  30         jn1 = (j1, "JJ", j1')
  31         jn2 = jn j2 j3 j2'
  32
  33 groupNgrams ((j1,"JJ",_):(_,"CC",_):(j2,"JJ",_):(n,"NN",nn):xs)      = groupNgrams (jn1:jn2:xs)
  34     where
  35         jn j m mm p = (j <> " " <> m, p, mm)
  36         jn1 = jn j1 n nn ("NN+CC" :: Text)
  37         jn2 = jn j2 n nn ("NN+CC" :: Text)
  38
  39 groupNgrams ((j1,"JJ",_):(_,"CC",_):(j2,"JJ",_):(n,"NNS",nn):xs)      = groupNgrams (jn1:jn2:xs)
  40     where
  41         jn j m mm p = (j <> " " <> m, p, mm)
  42         jn1 = jn j1 n nn ("NN+CC" :: Text)
  43         jn2 = jn j2 n nn ("NN+CC" :: Text)
  44
  45 groupNgrams ((x,"JJ",_):(y,"JJ",yy):xs)    = groupNgrams ((x <> " " <> y, "JJ", yy):xs)
  46 groupNgrams ((x,"JJ",_):(y,"NN",yy):xs)    = groupNgrams ((x <> " " <> y, "NN", yy):xs)
  47 groupNgrams ((x,"JJ",_):(y,"NNS",yy):xs)   = groupNgrams ((x <> " " <> y, "NN", yy):xs)
  48
  49 groupNgrams ((x,"NNP",_):(y,"NN",yy):xs)   = groupNgrams ((x <> " " <> y, "NN", yy):xs)
  50 groupNgrams ((x,"NN",_):(y,"NP",yy):xs)    = groupNgrams ((x <> " " <> y, "NN", yy):xs)
  51
  52 groupNgrams ((x,"NN",_):(y,"NNS",yy):xs)   = groupNgrams ((x <> " " <> y, "NN", yy):xs)
  53 groupNgrams ((x,"NP",_):(y,"NP",yy):xs)    = groupNgrams ((x <> " " <> y, "NN", yy):xs)
  54
  55 groupNgrams ((x,"NN",_):(y,"NN",yy):xs)    = groupNgrams ((x <> " " <> y, "NN", yy):xs)
  56
  57
  58 -- extractNgrams "Test the antiinflammatory or analgesic activity?"
  59 -- [[("``","``","O"),("Test","VB","O"),("the","DT","O"),("antiinflammatory activity analgesic activity","NN","O"),("?",".","O"),("''","''","O")]]
  60 -- > should be (antiinflammatory activity) <> (analgesic activity)
  61
  62 groupNgrams ((x,"NN",_):(o,"IN",_):(y,"NN",yy):xs)       = groupNgrams ((x <> " " <> o <> " " <> y, "NN", yy):xs)
  63 groupNgrams ((x,"NN",_):(o,"IN",_):(y,"NNP",yy):xs)       = groupNgrams ((x <> " " <> o <> " " <> y, "NN", yy):xs)
  64
  65 groupNgrams ((x,"NN",_):(o,"IN",_):(det,"DT",_):(y,"NN",yy):xs)       = groupNgrams ((x <> " " <> o <> " " <> det <> " " <> y, "NN", yy):xs)
  66 groupNgrams ((x,"NN",_):(o,"IN",_):(det,"DT",_):(y,"NNP",yy):xs)       = groupNgrams ((x <> " " <> o <> " " <> det <> " " <> y, "NN", yy):xs)
  67
  68 groupNgrams ((x,_,"PERSON"):(y,yy,"PERSON"):xs)             = groupNgrams ((x <> " " <> y,yy,"PERSON"):xs)
  69 groupNgrams ((x,_,"ORGANIZATION"):(y,yy,"ORGANIZATION"):xs) = groupNgrams ((x <> " " <> y,yy,"ORGANIZATION"):xs)
  70 groupNgrams ((x,_,"LOCATION"):(y,yy,"LOCATION"):xs) = groupNgrams ((x <> " " <> y,yy,"LOCATION"):xs)
  71
  72 groupNgrams (x:xs)                                          = (x:(groupNgrams xs))
  73
  74
  75 textTest :: [Text]
  76 textTest = [ "Alcoholic extract of Kaempferia galanga was tested for analgesic and antiinflammatory activities in animal models. "
  77            , "Three doses, 300 mg/kg, 600 mg/kg and 1200 mg/kg of the plant extract prepared as a suspension in 2 ml of 2% gum acacia were used. "
  78            , " Acute and sub acute inflammatory activities were studied in rats by carrageenan induced paw edema and cotton pellet induced granuloma models respectively. "
  79           , "In both models, the standard drug used was aspirin 100 mg/kg. "
  80           , "Two doses 600 mg/kg and 1200 mg/kg of plant extract exhibited significant (P<0.001) antiinflammatory activity in carrageenan model and cotton pellet granuloma model in comparison to control. "
  81          , "Analgesic activity was studied in rats using hot plate and tail-flick models. "
  82          , "Codeine 5 mg/kg and vehicle served as standard and control respectively. "
  83          , "The two doses of plant extract exhibited significant analgesic activity in tail flick model (P<0.001) and hot plate model (P<0.001) in comparison to control. "
  84          , "In conclusion K. galanga possesses antiinflammatory and analgesic activities. "]
  85
  86