src/Gargantext/Text/Metrics/TFICF.hs

   1 {-|
   2 Module      : Gargantext.Text.Metrics.TFICF
   3 Description : TFICF Ngrams tools
   4 Copyright   : (c) CNRS, 2017
   5 License     : AGPL + CECILL v3
   6 Maintainer  : team@gargantext.org
   7 Stability   : experimental
   8 Portability : POSIX
   9
  10 Definition of TFICF : Term Frequency - Inverse of Context Frequency
  11
  12 TFICF is a generalization of [TFIDF](https://en.wikipedia.org/wiki/Tf%E2%80%93idf).
  13
  14 -}
  15
  16
  17 module Gargantext.Text.Metrics.TFICF ( TFICF
  18                                      , TficfContext(..)
  19                                      , Total(..)
  20                                      , Count(..)
  21                                      , tficf
  22                                      )
  23   where
  24
  25 import Data.Text (Text)
  26 import Gargantext.Prelude
  27
  28 path :: Text
  29 path = "Gargantext.Text.Metrics.TFICF"
  30
  31 type TFICF = Double
  32
  33 data TficfContext n m = TficfInfra n m
  34                       | TficfSupra n m
  35   deriving (Show)
  36
  37 data Total = Total {unTotal :: !Double}
  38 data Count = Count {unCount :: !Double}
  39
  40 tficf :: TficfContext Count Total
  41       -> TficfContext Count Total
  42       -> TFICF
  43 tficf (TficfInfra (Count ic) (Total it) )
  44       (TficfSupra (Count sc) (Total st) )
  45             | it >= ic && st >= sc = (ic/it) / log (sc/st)
  46             | otherwise            = panic $ "[ERR]" <> path <>" Frequency impossible"
  47 tficf _ _ = panic $ "[ERR]" <> path <> "Undefined for these contexts"
  48
  49