datasets/corrections.json

{
  "slang": {
    "gonna": "going to",
    "wanna": "want to",
    "gotta": "got to",
    "lemme": "let me",
    "gimme": "give me",
    "kinda": "kind of",
    "sorta": "sort of",
    "dunno": "don't know",
    "ain't": "is not",
    "y'all": "you all",
    "bro": "brother",
    "sis": "sister",
    "dude": "friend",
    "nah": "no",
    "yep": "yes",
    "yup": "yes",
    "nope": "no",
    "ok": "okay",
    "coz": "because",
    "cuz": "because",
    "cos": "because",
    "sup": "what is up",
    "wassup": "what is up",
    "whatcha": "what are you",
    "imma": "I am going to",
    "fam": "family",
    "lit": "exciting",
    "lowkey": "slightly",
    "highkey": "very much",
    "salty": "upset",
    "shook": "shocked",
    "vibe": "feeling",
    "flex": "show off",
    "ghost": "disappear",
    "slay": "do excellently",
    "bet": "agreed",
    "cap": "lie",
    "no cap": "no lie",
    "fire": "excellent",
    "sus": "suspicious",
    "bruh": "brother",
    "hella": "very",
    "bougie": "fancy",
    "clap back": "respond sharply",
    "shade": "disrespect",
    "tea": "gossip",
    "mood": "relatable",
    "stan": "admire greatly",
    "simp": "overly devoted person",
    "yeet": "throw",
    "oof": "expressing discomfort",
    "periodt": "period",
    "deadass": "seriously",
    "idk": "I don't know",
    "tbh": "to be honest",
    "imo": "in my opinion",
    "smh": "shaking my head",
    "lol": "laughing out loud",
    "omg": "oh my god",
    "brb": "be right back",
    "btw": "by the way",
    "fyi": "for your information",
    "asap": "as soon as possible",
    "yolo": "you only live once"
  },
  "short_forms": {
    "govt": "government",
    "dept": "department",
    "mgr": "manager",
    "prof": "professor",
    "doc": "doctor",
    "info": "information",
    "tech": "technology",
    "app": "application",
    "bio": "biography",
    "exam": "examination",
    "lab": "laboratory",
    "math": "mathematics",
    "stats": "statistics",
    "ref": "reference",
    "temp": "temperature",
    "approx": "approximately",
    "etc": "etcetera",
    "vs": "versus",
    "min": "minute",
    "max": "maximum",
    "avg": "average",
    "qty": "quantity",
    "yr": "year",
    "mo": "month",
    "wk": "week",
    "hr": "hour",
    "sec": "second",
    "Mon": "Monday",
    "Tue": "Tuesday",
    "Wed": "Wednesday",
    "Thu": "Thursday",
    "Fri": "Friday",
    "Sat": "Saturday",
    "Sun": "Sunday",
    "Jan": "January",
    "Feb": "February",
    "Mar": "March",
    "Apr": "April",
    "Jun": "June",
    "Jul": "July",
    "Aug": "August",
    "Sep": "September",
    "Oct": "October",
    "Nov": "November",
    "Dec": "December"
  },
  "confused_words": {
    "their there": "there",
    "your you're": "you are",
    "its it's": "it is",
    "affect effect": "effect",
    "then than": "than",
    "loose lose": "lose",
    "weather whether": "whether",
    "to too": "too",
    "accept except": "except",
    "advise advice": "advice",
    "principal principle": "principle",
    "stationary stationery": "stationery",
    "complement compliment": "compliment",
    "desert dessert": "dessert",
    "alot": "a lot",
    "definately": "definitely",
    "seperate": "separate",
    "occured": "occurred",
    "recieve": "receive",
    "untill": "until",
    "writting": "writing",
    "calender": "calendar",
    "goverment": "government",
    "enviroment": "environment",
    "tommorrow": "tomorrow",
    "accomodate": "accommodate",
    "concious": "conscious",
    "necesary": "necessary",
    "occassion": "occasion",
    "priviledge": "privilege",
    "pronounciation": "pronunciation",
    "recomend": "recommend",
    "refered": "referred",
    "succesful": "successful",
    "wierd": "weird"
  },
  "domain_terms": {
    "medical": {
      "BP": "blood pressure",
      "HR": "heart rate",
      "temp": "temperature",
      "RX": "prescription",
      "OPD": "outpatient department",
      "ICU": "intensive care unit",
      "ER": "emergency room",
      "MRI": "magnetic resonance imaging",
      "CT scan": "computed tomography scan",
      "ECG": "electrocardiogram"
    },
    "education": {
      "GPA": "grade point average",
      "CGPA": "cumulative grade point average",
      "HOD": "head of department",
      "PhD": "Doctor of Philosophy",
      "MSc": "Master of Science",
      "BSc": "Bachelor of Science",
      "sem": "semester",
      "assign": "assignment",
      "ppt": "presentation"
    },
    "technology": {
      "API": "application programming interface",
      "UI": "user interface",
      "UX": "user experience",
      "DB": "database",
      "OS": "operating system",
      "CPU": "central processing unit",
      "GPU": "graphics processing unit",
      "RAM": "random access memory",
      "SSD": "solid state drive",
      "URL": "uniform resource locator"
    },
    "business": {
      "CEO": "Chief Executive Officer",
      "CFO": "Chief Financial Officer",
      "CTO": "Chief Technology Officer",
      "HR": "human resources",
      "KPI": "key performance indicator",
      "ROI": "return on investment",
      "B2B": "business to business",
      "B2C": "business to consumer",
      "ETA": "estimated time of arrival",
      "RSVP": "please respond"
    }
  },
  "pre_corrections": {
    "en": {
      "hii": "hi",
      "hiii": "hi",
      "hellooo": "hello",
      "byeee": "bye",
      "yesss": "yes",
      "nooo": "no",
      "pleaseee": "please",
      "thanksss": "thanks",
      "sorryyy": "sorry",
      "okayyyy": "okay",
      "wowww": "wow",
      "ummm": "",
      "uhhh": "",
      "ahhh": "",
      "hmmmm": "",
      "errr": "",
      "like like": "like",
      "you know you know": "you know",
      "I mean I mean": "I mean"
    }
  },
  "post_corrections": {}
}