Disturbance

Disturbance Object

Spell-checking object. Each token is associated with an object and is checked against our dictionary. If the token does not exist, a list of candidates will be suggested along with the most likely replacement. Gender and plural are not taken into account as the lemma will be the same regardless.

KEY	TYPE	DESCRIPTION	CONSTRAINTS
original	string	token as found in the sentence	-
predict	string	predicted token after spell-checking, same as original for valid words	-
probabilities	dictionary	candidates for correction with correspond rank (0 being the most likely candidate)	Key only exists if word should be corrected
to_correct	int	1 if token should be corrected but no replacement has been found	0 or 1

Python

[
  {
  "original":"pierre",
  "predict":"pierre",
  "to_correct":0
  },
  {
  "original":"aime",
  "predict":"aime",
  "to_correct":0
  },
  {
  "original":"les",
  "predict":"les",
  "to_correct":0
  },
  {
  "original":"banans",
  "predict":"banane",
  "probabilities":{
      "banals":2,
      "banane":0,
      "bananes":1,
      "banats":3,
      "banians":3,
      "bannans":3,
      "bannas":3,
      "bavans":3,
      "nanans":3
    },
  "to_correct":0
  }
]

JSON

[
  {
  "original":"pierre",
  "predict":"pierre",
  "to_correct":0
  },
  {
  "original":"aime",
  "predict":"aime",
  "to_correct":0
  },
  {
  "original":"les",
  "predict":"les",
  "to_correct":0
  },
  {
  "original":"banans",
  "predict":"banane",
  "probabilities":{
      "banals":2,
      "banane":0,
      "bananes":1,
      "banats":3,
      "banians":3,
      "bannans":3,
      "bannas":3,
      "bavans":3,
      "nanans":3
    },
  "to_correct":0
  }
]