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 data TermType = Mono | Multi | MonoMulti
  46
  47 -- remove Stop Words
  48 -- map (filter (\t -> not . elem t)) $
  49 ------------------------------------------------------------------------
  50 -- | Sugar to extract terms from text (hiddeng mapM from end user).
  51 extractTerms :: Traversable t => TermType -> Lang -> t Text -> IO (t [Terms])
  52 extractTerms termType lang = mapM (terms termType lang)
  53 ------------------------------------------------------------------------
  54 -- | Terms from Text
  55 -- Mono : mono terms
  56 -- Multi : multi terms
  57 -- MonoMulti : mono and multi
  58 -- TODO : multi terms should exclude mono (intersection is not empty yet)
  59 terms :: TermType -> Lang -> Text -> IO [Terms]
  60 terms Mono  lang txt     = pure $ monoterms' lang txt
  61 terms Multi lang txt     = multiterms lang txt
  62 terms MonoMulti lang txt = terms Multi lang txt
  63 ------------------------------------------------------------------------
  64