src/Gargantext/Text/Terms.hs

   1 {-|
   2 Module      : Gargantext.Text.Ngrams
   3 Description : Ngrams definition and tools
   4 Copyright   : (c) CNRS, 2017 - present
   5 License     : AGPL + CECILL v3
   6 Maintainer  : team@gargantext.org
   7 Stability   : experimental
   8 Portability : POSIX
   9
  10 An @n-gram@ is a contiguous sequence of n items from a given sample of
  11 text. In Gargantext application the items are words, n is a non negative
  12 integer.
  13
  14 Using Latin numerical prefixes, an n-gram of size 1 is referred to as a
  15 "unigram"; size 2 is a "bigram" (or, less commonly, a "digram"); size
  16 3 is a "trigram". English cardinal numbers are sometimes used, e.g.,
  17 "four-gram", "five-gram", and so on.
  18
  19 Source: https://en.wikipedia.org/wiki/Ngrams
  20
  21 TODO
  22 group Ngrams -> Tree
  23 compute occ by node of Tree
  24 group occs according groups
  25
  26 compute cooccurrences
  27 compute graph
  28
  29 -}
  30
  31 {-# LANGUAGE NoImplicitPrelude #-}
  32
  33 module Gargantext.Text.Terms
  34   where
  35
  36 import Data.Text (Text)
  37 import Data.Traversable
  38
  39 import Gargantext.Prelude
  40 import Gargantext.Core
  41 import Gargantext.Core.Types
  42 import Gargantext.Text.Terms.Multi (multiterms)
  43 import Gargantext.Text.Terms.Mono  (monoTerms)
  44
  45
  46 data TermType lang = Mono lang | Multi lang | MonoMulti lang
  47
  48
  49 --group :: [Text] -> [Text]
  50 --group = undefined
  51
  52 -- remove Stop Words
  53 -- map (filter (\t -> not . elem t)) $
  54 ------------------------------------------------------------------------
  55 -- | Sugar to extract terms from text (hiddeng mapM from end user).
  56 --extractTerms :: Traversable t => TermType Lang -> t Text -> IO (t [Terms])
  57 extractTerms :: TermType Lang -> [Text] -> IO [[Terms]]
  58 extractTerms termTypeLang = mapM (terms termTypeLang)
  59 ------------------------------------------------------------------------
  60 -- | Terms from Text
  61 -- Mono : mono terms
  62 -- Multi : multi terms
  63 -- MonoMulti : mono and multi
  64 -- TODO : multi terms should exclude mono (intersection is not empty yet)
  65 terms :: TermType Lang -> Text -> IO [Terms]
  66 terms (Mono      lang) txt = pure $ monoTerms lang txt
  67 terms (Multi     lang) txt = multiterms lang txt
  68 terms (MonoMulti lang) txt = terms (Multi lang) txt
  69 -- terms (WithList  list) txt = pure . concat $ extractTermsWithList list txt
  70 ------------------------------------------------------------------------
  71